dense elastic deformation batch normalization.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np \n",
    "import pandas as pd \n",
    "import skimage, os\n",
    "from skimage.morphology import ball, disk, dilation, binary_erosion, remove_small_objects, erosion, closing, reconstruction, binary_closing\n",
    "from skimage.measure import label,regionprops, perimeter\n",
    "from skimage.morphology import binary_dilation, binary_opening\n",
    "from skimage.filters import roberts, sobel\n",
    "from skimage import measure, feature\n",
    "from skimage.segmentation import clear_border\n",
    "from skimage import data\n",
    "from scipy import ndimage as ndi\n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d.art3d import Poly3DCollection\n",
    "import scipy.misc\n",
    "import numpy as np\n",
    "from glob import glob\n",
    "from skimage.io import imread\n",
    "from livelossplot import PlotLossesKerasTF\n",
    "from functools import *\n",
    "import elasticdeform\n",
    "\n",
    "import random\n",
    "os.environ[\"CUDA_VISIBLE_DEVICES\"]='1'\n",
    "BASE_IMG_PATH=os.path.join('hecktor-master','data','resampled')\n",
    "BATCH_SIZE = 3\n",
    "INPUT_SHAPE = (144,144,144,2)\n",
    "PROJECT_NAME = 'hector'\n",
    "AUG_RATE = 0.25\n",
    "\n",
    "\n",
    "random.seed(42)\n",
    "random_seed = 42\n",
    "\n",
    "AUGMENTATION_OPS = [\n",
    "    lambda t: np.rot90(t, axes=(1, 2), k=1),  # 90 degree rotation\n",
    "    lambda t: np.rot90(t, axes=(1, 2), k=2),  # 180 degree rotation\n",
    "    lambda t: np.rot90(t, axes=(1, 2), k=3),  # 270 degree rotation\n",
    "    lambda t: np.flip(t, axis=0),  # Axial flip\n",
    "    lambda t: np.flip(t, axis=1),  # Sagittal flip\n",
    "    lambda t: np.flip(t, axis=2)  # Coronal flip\n",
    "]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['hecktor-master/data/resampled/CHGJ073_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ035_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ082_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS069_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ083_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS052_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM047_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ025_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS041_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS097_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS046_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS055_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR013_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS015_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM027_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS015_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ053_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS098_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ086_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM001_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ090_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS087_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS065_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS068_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS056_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ065_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM026_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ032_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ030_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM006_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM016_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ070_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ032_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ058_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS016_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR024_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS039_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM064_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ008_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ067_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS101_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS016_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ074_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM060_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS078_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS090_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ057_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ030_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS053_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS051_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ090_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS090_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM049_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ069_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM040_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM017_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ048_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ029_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM041_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ036_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM036_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS006_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM046_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS068_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM030_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM023_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR016_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS009_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS048_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ065_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM039_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ062_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS068_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM059_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM027_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR025_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM044_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR016_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ038_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS008_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ091_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM065_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS067_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR029_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM053_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM054_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS041_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS019_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ017_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM033_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR034_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR001_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM015_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ007_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ007_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM024_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS081_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM058_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM016_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM012_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS100_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ050_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM006_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ016_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ039_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ088_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS027_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM019_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS094_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ026_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS036_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ034_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM045_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM013_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ052_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ074_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS005_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM029_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS055_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS060_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM035_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM048_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ034_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM047_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS042_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ031_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM035_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ072_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS049_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS013_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ026_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ034_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM041_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ046_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS069_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR025_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS026_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS096_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS039_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS008_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS046_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS005_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ057_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS049_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ081_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS047_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR021_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ088_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM060_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ081_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ082_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ092_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS007_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS080_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS009_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ073_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR029_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM038_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM029_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR024_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ080_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS031_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ015_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS052_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS056_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ035_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS076_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ017_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS033_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ074_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM038_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS098_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS077_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ070_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS078_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS081_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ088_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM032_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM021_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM014_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM014_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR030_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM051_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM017_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS065_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM007_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM063_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS061_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS038_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM021_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ066_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM055_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS047_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS019_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM056_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM007_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM062_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS003_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS087_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR011_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS007_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS031_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS086_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS026_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ058_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR030_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ092_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ043_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS020_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ087_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS085_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ052_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS028_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ015_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM050_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS058_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ037_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS100_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM056_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM058_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ077_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM019_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM008_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM023_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ071_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS040_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ046_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM022_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS015_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS030_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ038_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ055_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ008_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS020_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ066_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR040_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ071_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS010_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS058_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS066_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS039_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM029_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM056_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM024_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM026_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS064_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM034_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS073_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS031_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS022_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM047_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS101_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ076_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM045_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ090_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR028_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR021_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS043_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ010_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM010_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM022_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS045_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS035_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM042_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS022_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS088_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR020_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR005_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR004_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS045_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS085_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS009_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM037_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ066_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ055_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM039_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM049_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ080_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS091_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM001_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS088_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS036_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM036_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM001_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR013_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS091_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM008_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM064_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM040_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS060_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR025_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS067_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ050_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM044_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS095_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM044_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS043_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM018_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ035_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM043_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS064_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ018_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS033_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS013_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS050_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM032_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ069_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ013_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS046_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS008_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM034_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ069_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM060_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS074_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR023_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM048_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR005_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS005_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM051_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ077_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ078_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS021_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM018_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM055_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ036_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR024_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM011_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS010_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS049_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ013_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS035_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS057_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS097_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ086_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ072_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM065_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM014_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR012_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM015_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS003_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS065_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM017_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS074_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS048_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR020_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ017_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS095_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS048_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS027_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ083_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS074_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS033_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM059_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR023_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM046_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ031_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ037_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS030_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR040_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS096_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS050_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS013_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ016_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ078_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM049_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM038_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ050_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM048_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS057_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM062_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS077_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ062_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS030_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR021_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM045_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS083_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ036_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR005_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS052_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM010_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM019_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM061_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS021_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS088_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR012_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS026_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS083_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ087_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR040_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM053_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ029_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM037_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ026_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ070_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR028_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS042_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ089_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS041_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM036_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM055_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM051_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS040_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS021_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ081_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM027_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ053_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM022_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS066_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS004_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ008_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS004_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS060_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS100_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ086_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS096_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM058_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM050_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM030_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS077_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM057_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS098_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR011_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM053_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS095_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS051_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR012_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ092_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM012_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ048_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ046_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS101_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS006_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS094_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM024_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS073_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS020_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS007_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS087_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM042_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS089_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS078_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS043_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM057_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR014_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS085_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS089_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM034_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR004_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ085_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ032_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM023_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM007_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS056_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ058_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS076_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM035_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM062_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ028_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR004_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS091_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ089_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM011_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ018_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS038_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ076_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ048_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM061_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS081_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR011_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM011_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM030_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS058_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS019_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR013_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS076_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM059_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM039_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ083_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR030_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM037_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ030_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM043_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ055_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR014_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM016_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ052_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS036_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS010_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS055_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR034_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ037_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ025_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS003_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM012_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ053_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM015_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ091_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM046_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR001_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS069_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM013_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ031_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS080_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS086_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ076_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM063_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM033_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ015_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ085_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS073_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS089_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM013_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS066_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ016_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS006_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS040_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS038_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS035_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM010_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR034_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM064_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ039_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS053_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS045_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS004_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM006_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM033_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ071_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ085_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM041_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM057_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR016_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR029_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM043_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM021_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ025_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR014_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ018_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ077_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS064_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS042_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS090_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ057_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS016_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR020_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ029_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS083_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ087_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS027_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM002_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ039_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ080_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS097_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM040_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM054_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS086_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM050_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS061_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS053_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS050_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS028_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ007_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM061_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR023_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ067_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS028_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ043_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ028_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM008_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ082_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM063_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS051_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS094_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM042_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS061_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS047_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ091_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ065_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ028_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ089_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ067_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS057_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ038_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM026_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS067_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ010_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM002_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM018_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR001_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ013_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHMR028_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ078_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ072_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS080_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM002_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM065_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM054_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUS022_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHUM032_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ043_ct.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ062_ct_gtvt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ010_pt.nii.gz',\n",
       " 'hecktor-master/data/resampled/CHGJ073_ct_gtvt.nii.gz']"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "glob(os.path.join(BASE_IMG_PATH,'*'))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "201  matching files found: hecktor-master/data/resampled/CHUS074_ct.nii.gz hecktor-master/data/resampled/CHUS074_pt.nii.gz\n",
      "201  matching files found: hecktor-master/data/resampled/CHUS074_ct.nii.gz hecktor-master/data/resampled/CHUS074_ct_gtvt.nii.gz\n"
     ]
    }
   ],
   "source": [
    "# show some of the files\n",
    "all_ct_images= glob(os.path.join(BASE_IMG_PATH,'*ct.nii.gz'))\n",
    "all_pt_images = [x.replace('ct.nii.gz', 'pt.nii.gz') for x in all_ct_images]\n",
    "all_mask_images = [x.replace('ct.nii.gz', 'ct_gtvt.nii.gz') for x in all_ct_images]\n",
    "np.random.seed(42)\n",
    "rand_index = np.random.randint(0, len(all_ct_images))\n",
    "print(len(all_ct_images),' matching files found:',all_ct_images[rand_index], all_pt_images[rand_index])\n",
    "print(len(all_mask_images),' matching files found:',all_ct_images[rand_index], all_mask_images[rand_index])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 2\n",
      "1.694209098815918\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Mask')"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "try:\n",
    "    import nibabel as nib\n",
    "except:\n",
    "    raise ImportError('Install NIBABEL')\n",
    "    \n",
    "import time\n",
    "\n",
    "start = time.time()\n",
    "test_ct=np.asarray(nib.load(all_ct_images[rand_index]).dataobj)\n",
    "test_pt=np.asarray(nib.load(all_pt_images[rand_index]).dataobj)\n",
    "test_mask=np.asarray(nib.load(all_mask_images[rand_index]).dataobj)\n",
    "\n",
    "import random\n",
    "num1, num2 = sorted(random.sample(set([0, 1, 2]), 2))\n",
    "print(num1, num2)\n",
    "[test_ct, test_pt, test_mask ] = elasticdeform.deform_random_grid([test_ct,test_pt,test_mask ] , sigma=3, points=4, \n",
    "                                                                  axis=[(num1, num2), (num1, num2), (num1, num2)])\n",
    "\n",
    "print(time.time() - start )\n",
    "fig, (ax1, ax2, ax3) = plt.subplots(1,3, figsize = (12, 6))\n",
    "\n",
    "ax1.imshow(test_ct[test_ct.shape[0]//2], cmap = 'gray')\n",
    "ax1.set_title('CT')\n",
    "ax2.imshow(test_pt[test_pt.shape[0]//2], cmap = 'gray')\n",
    "ax2.set_title('PT')\n",
    "ax3.imshow(test_mask[test_mask.shape[0]//2], cmap = 'gray')\n",
    "ax3.set_title('Mask')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0 0.0\n",
      "(144, 144, 144)\n",
      "10.825079917907715\n",
      "1.3599677 -0.3217075\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "try:\n",
    "    import nibabel as nib\n",
    "except:\n",
    "    raise ImportError('Install NIBABEL')\n",
    "    \n",
    "start = time.time()\n",
    "test_ct=np.asarray(nib.load(all_ct_images[rand_index]).dataobj)\n",
    "test_pt=np.asarray(nib.load(all_pt_images[rand_index]).dataobj)\n",
    "test_mask=np.asarray(nib.load(all_mask_images[rand_index]).dataobj)\n",
    "\n",
    "print(test_mask.max(), test_mask.min())\n",
    "print(test_ct.shape)\n",
    "\n",
    "[test_ct, test_pt, test_mask ] = elasticdeform.deform_random_grid([test_ct,test_pt,test_mask ] , sigma=3, points=5)\n",
    "\n",
    "print(time.time() - start )\n",
    "fig, (ax1, ax2, ax3) = plt.subplots(1,3, figsize = (12, 6))\n",
    "\n",
    "ax1.imshow(test_ct[test_ct.shape[0]//2], cmap = 'gray')\n",
    "ax1.set_title('CT')\n",
    "ax2.imshow(test_pt[test_pt.shape[0]//2], cmap = 'gray')\n",
    "ax2.set_title('PT')\n",
    "ax3.imshow(test_mask[test_mask.shape[0]//2], cmap = 'gray')\n",
    "ax3.set_title('Mask')\n",
    "\n",
    "print(test_mask.max(), test_mask.min())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([ 129.,  140.,  150.,  147.,  198., 2359.,  693.,  277.,   72.,\n",
       "          14.]),\n",
       " array([0.5007677 , 0.5866877 , 0.6726077 , 0.7585277 , 0.84444773,\n",
       "        0.9303677 , 1.0162877 , 1.1022077 , 1.1881278 , 1.2740477 ,\n",
       "        1.3599677 ], dtype=float32),\n",
       " <a list of 10 Patch objects>)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(test_mask[test_mask>0.5].ravel())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fce120b9c90>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x864 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy, imageio, elasticdeform\n",
    "X = numpy.zeros((200, 300))\n",
    "X[::10, ::10] = 1\n",
    "\n",
    "# apply deformation with a random 3 x 3 grid\n",
    "X_deformed = elasticdeform.deform_random_grid(X, sigma=5, points=5)\n",
    "\n",
    "f = plt.figure(figsize=(10,12))\n",
    "\n",
    "ax1 = plt.subplot(211)\n",
    "ax1.imshow(X)\n",
    "ax2 = plt.subplot(212)\n",
    "ax2.imshow(X_deformed)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "12.502632 -0.005205076 (144, 144, 144)\n",
      "10.030021 -0.94009787 (144, 144, 144)\n"
     ]
    }
   ],
   "source": [
    "def make144(image):\n",
    "    new_image = skimage.transform.resize(image, (144,144,144), mode= 'edge')\n",
    "    return new_image\n",
    "\n",
    "def preprocess(image):\n",
    "    if image.shape != (144,144,144):\n",
    "        image = make144(image)\n",
    "    new_image = (image - image.mean())/image.std()\n",
    "    \n",
    "    return new_image\n",
    "print(test_pt.max(), test_pt.min(),test_pt.shape)\n",
    "new_img = preprocess(test_pt)\n",
    "print(new_img.max(), new_img.min(),new_img.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(144, 144, 288)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = preprocess(test_pt)\n",
    "data = np.concatenate((data, preprocess(test_pt)), -1)\n",
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "#check if ct, pt, mask have same numbers.\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "if not len(all_ct_images) == len(all_pt_images) == len(all_mask_images):\n",
    "    print(len(all_ct_images), len(all_pt_images), len(all_mask_images))\n",
    "    raise ValueError('ct,pt,mask number not match')\n",
    "\n",
    "data_list_index = range(len(all_ct_images))\n",
    "train_data_list, val_data_list = train_test_split(data_list_index, test_size = 0.2, random_state = 42)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "from functools import *\n",
    "\n",
    "def get_augment_operations(random_state, probability=.25):\n",
    "    \"\"\"\n",
    "    Return a function that can be applied for augmenting a tensor.\n",
    "    Included operations are:\n",
    "        - flipping (axial, sagittal and coronal);\n",
    "        - rotation (90, 180 and 720).\n",
    "    Each of the six operations has a certain probability to be included for \n",
    "    augmentation.\n",
    "    :param random_state: random state.\n",
    "    :param probability: probability for operation inclusion.\n",
    "    :return: function to apply to tensor.\n",
    "    \"\"\"\n",
    "    probabilities = random_state.uniform(size=len(AUGMENTATION_OPS))\n",
    "    operations = [\n",
    "        f for p, f in zip(probabilities, AUGMENTATION_OPS) if p > probability\n",
    "    ]\n",
    "\n",
    "    def _wrapped(x):\n",
    "        augmented = reduce(lambda t, f: f(t), operations, x)\n",
    "        assert augmented.shape == x.shape\n",
    "        return augmented\n",
    "\n",
    "    return _wrapped\n",
    "\n",
    "# Augment\n",
    "#f = get_augment_operations(random_state, augment_probability)\n",
    "#x = f(x)\n",
    "#y = f(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-0.3217075\n"
     ]
    }
   ],
   "source": [
    "# Augment test\n",
    "\n",
    "random_state = np.random.RandomState()\n",
    "f = get_augment_operations(random_state)\n",
    "\n",
    "_aug_test_ct = f(test_ct)\n",
    "_aug_test_pt = f(test_pt)\n",
    "_aug_test_mask = f(test_mask)\n",
    "fig, (ax1, ax2, ax3) = plt.subplots(1,3, figsize = (12, 6))\n",
    "\n",
    "ax1.imshow(_aug_test_ct[_aug_test_ct.shape[0]//2], cmap = 'gray')\n",
    "ax1.set_title('CT')\n",
    "ax2.imshow(_aug_test_pt[_aug_test_pt.shape[0]//2], cmap = 'gray')\n",
    "ax2.set_title('PT')\n",
    "ax3.imshow(_aug_test_mask[_aug_test_mask.shape[0]//2], cmap = 'gray')\n",
    "ax3.set_title('Mask')\n",
    "plt.show()\n",
    "print(_aug_test_mask.min())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3, 144, 144, 144, 2) (3, 144, 144, 144)\n",
      "(3, 144, 144, 144)\n",
      "(3, 144, 144, 144, 2) (3, 144, 144, 144)\n",
      "(3, 144, 144, 144)\n",
      "(3, 144, 144, 144, 2) (3, 144, 144, 144)\n",
      "(3, 144, 144, 144)\n",
      "(3, 144, 144, 144, 2) (3, 144, 144, 144)\n",
      "(3, 144, 144, 144)\n",
      "(3, 144, 144, 144, 2) (3, 144, 144, 144)\n",
      "(3, 144, 144, 144)\n",
      "(3, 144, 144, 144, 2) (3, 144, 144, 144)\n",
      "(3, 144, 144, 144)\n",
      "(3, 144, 144, 144, 2) (3, 144, 144, 144)\n",
      "(3, 144, 144, 144)\n",
      "(3, 144, 144, 144, 2) (3, 144, 144, 144)\n",
      "(3, 144, 144, 144)\n",
      "testing valid pipeline\n",
      "(3, 144, 144, 144, 2) (3, 144, 144, 144)\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import random\n",
    "\n",
    "def tf_data_generator(file_list, batch_size = 2, augment_rate = 0):\n",
    "\n",
    "    i = 0\n",
    "    j = 0\n",
    "    if augment_rate!=0:\n",
    "        #if we want to augmenation\n",
    "        k = int(len(file_list) * augment_rate)\n",
    "        indicies = random.sample(range(len(file_list)), k)\n",
    "        aug_list = [file_list[i] for i in indicies]\n",
    "            \n",
    "    while True:\n",
    "        if i*batch_size >= (len(file_list) * (1 + augment_rate)): \n",
    "            #start again\n",
    "            i = 0\n",
    "            j = 0\n",
    "            np.random.shuffle(file_list)\n",
    "            \n",
    "        if augment_rate != 0 and i*batch_size >= len(file_list):  \n",
    "            #augmentation\n",
    "            # if data in augmentation list \n",
    "            file_chunk = aug_list[j*batch_size:(j+1)*batch_size] \n",
    "            data = []\n",
    "            labels = []\n",
    "            for file in file_chunk:   \n",
    "                _ct=np.asarray(nib.load(all_ct_images[file]).dataobj)\n",
    "                _pt=np.asarray(nib.load(all_pt_images[file]).dataobj)\n",
    "                _mask=np.asarray(nib.load(all_mask_images[file]).dataobj)\n",
    "                \n",
    "                _ct = preprocess(_ct)\n",
    "                _pt = preprocess(_pt)\n",
    "                _mask = make144(_mask)\n",
    "                \n",
    "                if i % 2 == 0 :\n",
    "                    #elastic deformation\n",
    "                    num1, num2 = sorted(random.sample(set([0, 1, 2]), 2))\n",
    "                    [_ct, _pt, _mask] = elasticdeform.deform_random_grid([_ct, _pt, _mask] , sigma=3, points=4, \n",
    "                                                                      axis=[(num1, num2), (num1, num2), (num1, num2)])            \n",
    "                #if i % 3 == 0: \n",
    "                #    #elastic deformation\n",
    "                #    #print(_ct.shape, _pt.shape, _mask.shape)\n",
    "                #    num1, num2 = sorted(random.sample(set([0, 1, 2]), 2))\n",
    "                #    [_ct, _pt, _mask] = elasticdeform.deform_random_grid([_ct, _pt, _mask] , sigma=3, points=4, \n",
    "                #                                                      axis=[(num1, num2), (num1, num2), (num1, num2)]) \n",
    "                \n",
    "                _mask[_mask<0.5] = 0 \n",
    "                _mask[_mask >= 0.5] = 1\n",
    "                \n",
    "                img = np.concatenate((_ct[..., np.newaxis], _pt[..., np.newaxis]), -1)\n",
    "                \n",
    "                if i % 1 == 0 :\n",
    "                    #print(\"rotation, flip deformation\")\n",
    "                    #rotation, flip augmentation\n",
    "                    random_state = np.random.RandomState()\n",
    "                    f = get_augment_operations(random_state)\n",
    "                    img = f(img)\n",
    "                    _mask = f(_mask)\n",
    "            \n",
    "                data.append(img) \n",
    "                #binarization again for deformated masks.\n",
    "\n",
    "                labels.append(_mask)\n",
    "                \n",
    "            data = np.asarray(data)\n",
    "            labels = np.asarray(labels)\n",
    "            yield data, labels\n",
    "            i = i + 1\n",
    "            j = j + 1\n",
    "\n",
    "        else:\n",
    "            file_chunk = file_list[i*batch_size:(i+1)*batch_size] \n",
    "            data = []\n",
    "            labels = []\n",
    "            for file in file_chunk:\n",
    "                _ct=np.asarray(nib.load(all_ct_images[file]).dataobj)\n",
    "                _pt=np.asarray(nib.load(all_pt_images[file]).dataobj)\n",
    "                _mask=np.asarray(nib.load(all_mask_images[file]).dataobj)\n",
    "                \n",
    "                _ct = preprocess(_ct)\n",
    "                _pt = preprocess(_pt)\n",
    "                \n",
    "                img = np.concatenate((_ct[..., np.newaxis], _pt[..., np.newaxis]), -1)\n",
    "                \n",
    "                data.append(img) \n",
    "                \n",
    "                _mask = make144(_mask)\n",
    "                labels.append(_mask)\n",
    "                \n",
    "            data = np.asarray(data)\n",
    "            labels = np.asarray(labels)\n",
    "            yield data, labels\n",
    "            i = i + 1\n",
    "            \n",
    "batch_size = BATCH_SIZE\n",
    "augment_rate = AUG_RATE\n",
    "#tf_data_generator(train_data_list)\n",
    "\n",
    "\n",
    "train_dataset = tf.data.Dataset.from_generator(tf_data_generator,args= [train_data_list, batch_size, augment_rate],\n",
    "                                         output_types = (tf.float32, tf.float32),\n",
    "                                         output_shapes = ((None,144,144,144,2),(None,144,144,144)))        \n",
    "\n",
    "valid_dataset = tf.data.Dataset.from_generator(tf_data_generator,args= [val_data_list, batch_size],\n",
    "                                         output_types = (tf.float32, tf.float32),\n",
    "                                         output_shapes = ((None,144,144,144,2),(None,144,144,144))) \n",
    "\n",
    "#test the data generator\n",
    "num = 0\n",
    "for data, labels in train_dataset:\n",
    "    print(data.shape, labels.shape)\n",
    "    print(labels.shape)\n",
    "    num = num + 1\n",
    "    if num > 7: break\n",
    "print(\"testing valid pipeline\")  \n",
    "for data, labels in valid_dataset:\n",
    "    print(data.shape, labels.shape)\n",
    "\n",
    "    num = num + 1\n",
    "    if num > 7: break"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Densely 3D Unet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            [(None, 144, 144, 14 0                                            \n",
      "__________________________________________________________________________________________________\n",
      "conv3d (Conv3D)                 (None, 144, 144, 144 2640        input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization (BatchNorma (None, 144, 144, 144 192         conv3d[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "activation (Activation)         (None, 144, 144, 144 0           batch_normalization[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_1 (Conv3D)               (None, 144, 144, 144 20752       activation[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "dropout (Dropout)               (None, 144, 144, 144 0           conv3d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "concatenate (Concatenate)       (None, 144, 144, 144 0           dropout[0][0]                    \n",
      "                                                                 conv3d[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_1 (BatchNor (None, 144, 144, 144 256         concatenate[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "activation_1 (Activation)       (None, 144, 144, 144 0           batch_normalization_1[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_2 (Conv3D)               (None, 144, 144, 144 27664       activation_1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "dropout_1 (Dropout)             (None, 144, 144, 144 0           conv3d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_1 (Concatenate)     (None, 144, 144, 144 0           dropout_1[0][0]                  \n",
      "                                                                 concatenate[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_2 (BatchNor (None, 144, 144, 144 320         concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_2 (Activation)       (None, 144, 144, 144 0           batch_normalization_2[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_3 (Conv3D)               (None, 144, 144, 144 34576       activation_2[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "dropout_2 (Dropout)             (None, 144, 144, 144 0           conv3d_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_2 (Concatenate)     (None, 144, 144, 144 0           dropout_2[0][0]                  \n",
      "                                                                 concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_3 (BatchNor (None, 144, 144, 144 384         concatenate_2[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_3 (Activation)       (None, 144, 144, 144 0           batch_normalization_3[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_4 (Conv3D)               (None, 144, 144, 144 9312        activation_3[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "dropout_3 (Dropout)             (None, 144, 144, 144 0           conv3d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_5 (Conv3D)               (None, 72, 72, 144,  9312        dropout_3[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_4 (BatchNor (None, 72, 72, 144,  384         conv3d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_4 (Activation)       (None, 72, 72, 144,  0           batch_normalization_4[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_6 (Conv3D)               (None, 72, 72, 144,  41488       activation_4[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "dropout_4 (Dropout)             (None, 72, 72, 144,  0           conv3d_6[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_3 (Concatenate)     (None, 72, 72, 144,  0           dropout_4[0][0]                  \n",
      "                                                                 conv3d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_5 (BatchNor (None, 72, 72, 144,  448         concatenate_3[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_5 (Activation)       (None, 72, 72, 144,  0           batch_normalization_5[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_7 (Conv3D)               (None, 72, 72, 144,  48400       activation_5[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "dropout_5 (Dropout)             (None, 72, 72, 144,  0           conv3d_7[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_4 (Concatenate)     (None, 72, 72, 144,  0           dropout_5[0][0]                  \n",
      "                                                                 concatenate_3[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_6 (BatchNor (None, 72, 72, 144,  512         concatenate_4[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_6 (Activation)       (None, 72, 72, 144,  0           batch_normalization_6[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_8 (Conv3D)               (None, 72, 72, 144,  55312       activation_6[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "dropout_6 (Dropout)             (None, 72, 72, 144,  0           conv3d_8[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_5 (Concatenate)     (None, 72, 72, 144,  0           dropout_6[0][0]                  \n",
      "                                                                 concatenate_4[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_7 (BatchNor (None, 72, 72, 144,  576         concatenate_5[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_7 (Activation)       (None, 72, 72, 144,  0           batch_normalization_7[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_9 (Conv3D)               (None, 72, 72, 144,  62224       activation_7[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "dropout_7 (Dropout)             (None, 72, 72, 144,  0           conv3d_9[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_6 (Concatenate)     (None, 72, 72, 144,  0           dropout_7[0][0]                  \n",
      "                                                                 concatenate_5[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_8 (BatchNor (None, 72, 72, 144,  640         concatenate_6[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_8 (Activation)       (None, 72, 72, 144,  0           batch_normalization_8[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_10 (Conv3D)              (None, 72, 72, 144,  25760       activation_8[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "dropout_8 (Dropout)             (None, 72, 72, 144,  0           conv3d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_11 (Conv3D)              (None, 36, 36, 144,  25760       dropout_8[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_9 (BatchNor (None, 36, 36, 144,  640         conv3d_11[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_9 (Activation)       (None, 36, 36, 144,  0           batch_normalization_9[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_12 (Conv3D)              (None, 36, 36, 144,  69136       activation_9[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "dropout_9 (Dropout)             (None, 36, 36, 144,  0           conv3d_12[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_7 (Concatenate)     (None, 36, 36, 144,  0           dropout_9[0][0]                  \n",
      "                                                                 conv3d_11[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_10 (BatchNo (None, 36, 36, 144,  704         concatenate_7[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_10 (Activation)      (None, 36, 36, 144,  0           batch_normalization_10[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_13 (Conv3D)              (None, 36, 36, 144,  76048       activation_10[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_10 (Dropout)            (None, 36, 36, 144,  0           conv3d_13[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_8 (Concatenate)     (None, 36, 36, 144,  0           dropout_10[0][0]                 \n",
      "                                                                 concatenate_7[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_11 (BatchNo (None, 36, 36, 144,  768         concatenate_8[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_11 (Activation)      (None, 36, 36, 144,  0           batch_normalization_11[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_14 (Conv3D)              (None, 36, 36, 144,  82960       activation_11[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_11 (Dropout)            (None, 36, 36, 144,  0           conv3d_14[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_9 (Concatenate)     (None, 36, 36, 144,  0           dropout_11[0][0]                 \n",
      "                                                                 concatenate_8[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_12 (BatchNo (None, 36, 36, 144,  832         concatenate_9[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_12 (Activation)      (None, 36, 36, 144,  0           batch_normalization_12[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_15 (Conv3D)              (None, 36, 36, 144,  89872       activation_12[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_12 (Dropout)            (None, 36, 36, 144,  0           conv3d_15[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_10 (Concatenate)    (None, 36, 36, 144,  0           dropout_12[0][0]                 \n",
      "                                                                 concatenate_9[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_13 (BatchNo (None, 36, 36, 144,  896         concatenate_10[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_13 (Activation)      (None, 36, 36, 144,  0           batch_normalization_13[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_16 (Conv3D)              (None, 36, 36, 144,  96784       activation_13[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_13 (Dropout)            (None, 36, 36, 144,  0           conv3d_16[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_11 (Concatenate)    (None, 36, 36, 144,  0           dropout_13[0][0]                 \n",
      "                                                                 concatenate_10[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_14 (BatchNo (None, 36, 36, 144,  960         concatenate_11[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_14 (Activation)      (None, 36, 36, 144,  0           batch_normalization_14[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_17 (Conv3D)              (None, 36, 36, 144,  57840       activation_14[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_14 (Dropout)            (None, 36, 36, 144,  0           conv3d_17[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_18 (Conv3D)              (None, 18, 18, 144,  57840       dropout_14[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_15 (BatchNo (None, 18, 18, 144,  960         conv3d_18[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_15 (Activation)      (None, 18, 18, 144,  0           batch_normalization_15[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_19 (Conv3D)              (None, 18, 18, 144,  103696      activation_15[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_15 (Dropout)            (None, 18, 18, 144,  0           conv3d_19[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_12 (Concatenate)    (None, 18, 18, 144,  0           dropout_15[0][0]                 \n",
      "                                                                 conv3d_18[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_16 (BatchNo (None, 18, 18, 144,  1024        concatenate_12[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_16 (Activation)      (None, 18, 18, 144,  0           batch_normalization_16[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_20 (Conv3D)              (None, 18, 18, 144,  110608      activation_16[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_16 (Dropout)            (None, 18, 18, 144,  0           conv3d_20[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_13 (Concatenate)    (None, 18, 18, 144,  0           dropout_16[0][0]                 \n",
      "                                                                 concatenate_12[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_17 (BatchNo (None, 18, 18, 144,  1088        concatenate_13[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_17 (Activation)      (None, 18, 18, 144,  0           batch_normalization_17[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_21 (Conv3D)              (None, 18, 18, 144,  117520      activation_17[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_17 (Dropout)            (None, 18, 18, 144,  0           conv3d_21[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_14 (Concatenate)    (None, 18, 18, 144,  0           dropout_17[0][0]                 \n",
      "                                                                 concatenate_13[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_18 (BatchNo (None, 18, 18, 144,  1152        concatenate_14[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_18 (Activation)      (None, 18, 18, 144,  0           batch_normalization_18[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_22 (Conv3D)              (None, 18, 18, 144,  124432      activation_18[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_18 (Dropout)            (None, 18, 18, 144,  0           conv3d_22[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_15 (Concatenate)    (None, 18, 18, 144,  0           dropout_18[0][0]                 \n",
      "                                                                 concatenate_14[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_19 (BatchNo (None, 18, 18, 144,  1216        concatenate_15[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_19 (Activation)      (None, 18, 18, 144,  0           batch_normalization_19[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_23 (Conv3D)              (None, 18, 18, 144,  131344      activation_19[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_19 (Dropout)            (None, 18, 18, 144,  0           conv3d_23[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_16 (Concatenate)    (None, 18, 18, 144,  0           dropout_19[0][0]                 \n",
      "                                                                 concatenate_15[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_20 (BatchNo (None, 18, 18, 144,  1280        concatenate_16[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_20 (Activation)      (None, 18, 18, 144,  0           batch_normalization_20[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_24 (Conv3D)              (None, 18, 18, 144,  138256      activation_20[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_20 (Dropout)            (None, 18, 18, 144,  0           conv3d_24[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_17 (Concatenate)    (None, 18, 18, 144,  0           dropout_20[0][0]                 \n",
      "                                                                 concatenate_16[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_21 (BatchNo (None, 18, 18, 144,  1344        concatenate_17[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_21 (Activation)      (None, 18, 18, 144,  0           batch_normalization_21[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_25 (Conv3D)              (None, 18, 18, 144,  113232      activation_21[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_21 (Dropout)            (None, 18, 18, 144,  0           conv3d_25[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_26 (Conv3D)              (None, 9, 9, 144, 33 113232      dropout_21[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_22 (BatchNo (None, 9, 9, 144, 33 1344        conv3d_26[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_22 (Activation)      (None, 9, 9, 144, 33 0           batch_normalization_22[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_27 (Conv3D)              (None, 9, 9, 144, 16 145168      activation_22[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_22 (Dropout)            (None, 9, 9, 144, 16 0           conv3d_27[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_18 (Concatenate)    (None, 9, 9, 144, 35 0           dropout_22[0][0]                 \n",
      "                                                                 conv3d_26[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_23 (BatchNo (None, 9, 9, 144, 35 1408        concatenate_18[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_23 (Activation)      (None, 9, 9, 144, 35 0           batch_normalization_23[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_28 (Conv3D)              (None, 9, 9, 144, 16 152080      activation_23[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_23 (Dropout)            (None, 9, 9, 144, 16 0           conv3d_28[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_19 (Concatenate)    (None, 9, 9, 144, 36 0           dropout_23[0][0]                 \n",
      "                                                                 concatenate_18[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_24 (BatchNo (None, 9, 9, 144, 36 1472        concatenate_19[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_24 (Activation)      (None, 9, 9, 144, 36 0           batch_normalization_24[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_29 (Conv3D)              (None, 9, 9, 144, 16 158992      activation_24[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_24 (Dropout)            (None, 9, 9, 144, 16 0           conv3d_29[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_20 (Concatenate)    (None, 9, 9, 144, 38 0           dropout_24[0][0]                 \n",
      "                                                                 concatenate_19[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_25 (BatchNo (None, 9, 9, 144, 38 1536        concatenate_20[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_25 (Activation)      (None, 9, 9, 144, 38 0           batch_normalization_25[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_30 (Conv3D)              (None, 9, 9, 144, 16 165904      activation_25[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_25 (Dropout)            (None, 9, 9, 144, 16 0           conv3d_30[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_21 (Concatenate)    (None, 9, 9, 144, 40 0           dropout_25[0][0]                 \n",
      "                                                                 concatenate_20[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_26 (BatchNo (None, 9, 9, 144, 40 1600        concatenate_21[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_26 (Activation)      (None, 9, 9, 144, 40 0           batch_normalization_26[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_31 (Conv3D)              (None, 9, 9, 144, 16 172816      activation_26[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_26 (Dropout)            (None, 9, 9, 144, 16 0           conv3d_31[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_22 (Concatenate)    (None, 9, 9, 144, 41 0           dropout_26[0][0]                 \n",
      "                                                                 concatenate_21[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_27 (BatchNo (None, 9, 9, 144, 41 1664        concatenate_22[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_27 (Activation)      (None, 9, 9, 144, 41 0           batch_normalization_27[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_32 (Conv3D)              (None, 9, 9, 144, 16 179728      activation_27[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_27 (Dropout)            (None, 9, 9, 144, 16 0           conv3d_32[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_23 (Concatenate)    (None, 9, 9, 144, 43 0           dropout_27[0][0]                 \n",
      "                                                                 concatenate_22[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_28 (BatchNo (None, 9, 9, 144, 43 1728        concatenate_23[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_28 (Activation)      (None, 9, 9, 144, 43 0           batch_normalization_28[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_33 (Conv3D)              (None, 9, 9, 144, 16 186640      activation_28[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_28 (Dropout)            (None, 9, 9, 144, 16 0           conv3d_33[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_24 (Concatenate)    (None, 9, 9, 144, 44 0           dropout_28[0][0]                 \n",
      "                                                                 concatenate_23[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_transpose (Conv3DTranspo (None, 18, 18, 144,  1354864     concatenate_24[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_25 (Concatenate)    (None, 18, 18, 144,  0           conv3d_transpose[0][0]           \n",
      "                                                                 concatenate_17[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_29 (BatchNo (None, 18, 18, 144,  1792        concatenate_25[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_29 (Activation)      (None, 18, 18, 144,  0           batch_normalization_29[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_34 (Conv3D)              (None, 18, 18, 144,  193552      activation_29[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_29 (Dropout)            (None, 18, 18, 144,  0           conv3d_34[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_26 (Concatenate)    (None, 18, 18, 144,  0           dropout_29[0][0]                 \n",
      "                                                                 concatenate_25[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_30 (BatchNo (None, 18, 18, 144,  1856        concatenate_26[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_30 (Activation)      (None, 18, 18, 144,  0           batch_normalization_30[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_35 (Conv3D)              (None, 18, 18, 144,  200464      activation_30[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_30 (Dropout)            (None, 18, 18, 144,  0           conv3d_35[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_27 (Concatenate)    (None, 18, 18, 144,  0           dropout_30[0][0]                 \n",
      "                                                                 concatenate_26[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_31 (BatchNo (None, 18, 18, 144,  1920        concatenate_27[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_31 (Activation)      (None, 18, 18, 144,  0           batch_normalization_31[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_36 (Conv3D)              (None, 18, 18, 144,  207376      activation_31[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_31 (Dropout)            (None, 18, 18, 144,  0           conv3d_36[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_28 (Concatenate)    (None, 18, 18, 144,  0           dropout_31[0][0]                 \n",
      "                                                                 concatenate_27[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_32 (BatchNo (None, 18, 18, 144,  1984        concatenate_28[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_32 (Activation)      (None, 18, 18, 144,  0           batch_normalization_32[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_37 (Conv3D)              (None, 18, 18, 144,  214288      activation_32[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_32 (Dropout)            (None, 18, 18, 144,  0           conv3d_37[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_29 (Concatenate)    (None, 18, 18, 144,  0           dropout_32[0][0]                 \n",
      "                                                                 concatenate_28[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_33 (BatchNo (None, 18, 18, 144,  2048        concatenate_29[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_33 (Activation)      (None, 18, 18, 144,  0           batch_normalization_33[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_38 (Conv3D)              (None, 18, 18, 144,  221200      activation_33[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_33 (Dropout)            (None, 18, 18, 144,  0           conv3d_38[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_30 (Concatenate)    (None, 18, 18, 144,  0           dropout_33[0][0]                 \n",
      "                                                                 concatenate_29[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_34 (BatchNo (None, 18, 18, 144,  2112        concatenate_30[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_34 (Activation)      (None, 18, 18, 144,  0           batch_normalization_34[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_39 (Conv3D)              (None, 18, 18, 144,  228112      activation_34[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_34 (Dropout)            (None, 18, 18, 144,  0           conv3d_39[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_31 (Concatenate)    (None, 18, 18, 144,  0           dropout_34[0][0]                 \n",
      "                                                                 concatenate_30[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_transpose_1 (Conv3DTrans (None, 36, 36, 144,  1410144     concatenate_31[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_32 (Concatenate)    (None, 36, 36, 144,  0           conv3d_transpose_1[0][0]         \n",
      "                                                                 concatenate_11[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_35 (BatchNo (None, 36, 36, 144,  1344        concatenate_32[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_35 (Activation)      (None, 36, 36, 144,  0           batch_normalization_35[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_40 (Conv3D)              (None, 36, 36, 144,  145168      activation_35[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_35 (Dropout)            (None, 36, 36, 144,  0           conv3d_40[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_33 (Concatenate)    (None, 36, 36, 144,  0           dropout_35[0][0]                 \n",
      "                                                                 concatenate_32[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_36 (BatchNo (None, 36, 36, 144,  1408        concatenate_33[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_36 (Activation)      (None, 36, 36, 144,  0           batch_normalization_36[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_41 (Conv3D)              (None, 36, 36, 144,  152080      activation_36[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_36 (Dropout)            (None, 36, 36, 144,  0           conv3d_41[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_34 (Concatenate)    (None, 36, 36, 144,  0           dropout_36[0][0]                 \n",
      "                                                                 concatenate_33[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_37 (BatchNo (None, 36, 36, 144,  1472        concatenate_34[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_37 (Activation)      (None, 36, 36, 144,  0           batch_normalization_37[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_42 (Conv3D)              (None, 36, 36, 144,  158992      activation_37[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_37 (Dropout)            (None, 36, 36, 144,  0           conv3d_42[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_35 (Concatenate)    (None, 36, 36, 144,  0           dropout_37[0][0]                 \n",
      "                                                                 concatenate_34[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_38 (BatchNo (None, 36, 36, 144,  1536        concatenate_35[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_38 (Activation)      (None, 36, 36, 144,  0           batch_normalization_38[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_43 (Conv3D)              (None, 36, 36, 144,  165904      activation_38[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_38 (Dropout)            (None, 36, 36, 144,  0           conv3d_43[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_36 (Concatenate)    (None, 36, 36, 144,  0           dropout_38[0][0]                 \n",
      "                                                                 concatenate_35[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_39 (BatchNo (None, 36, 36, 144,  1600        concatenate_36[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_39 (Activation)      (None, 36, 36, 144,  0           batch_normalization_39[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_44 (Conv3D)              (None, 36, 36, 144,  172816      activation_39[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_39 (Dropout)            (None, 36, 36, 144,  0           conv3d_44[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_37 (Concatenate)    (None, 36, 36, 144,  0           dropout_39[0][0]                 \n",
      "                                                                 concatenate_36[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_transpose_2 (Conv3DTrans (None, 72, 72, 144,  898640      concatenate_37[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_38 (Concatenate)    (None, 72, 72, 144,  0           conv3d_transpose_2[0][0]         \n",
      "                                                                 concatenate_6[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_40 (BatchNo (None, 72, 72, 144,  960         concatenate_38[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_40 (Activation)      (None, 72, 72, 144,  0           batch_normalization_40[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_45 (Conv3D)              (None, 72, 72, 144,  103696      activation_40[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_40 (Dropout)            (None, 72, 72, 144,  0           conv3d_45[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_39 (Concatenate)    (None, 72, 72, 144,  0           dropout_40[0][0]                 \n",
      "                                                                 concatenate_38[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_41 (BatchNo (None, 72, 72, 144,  1024        concatenate_39[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_41 (Activation)      (None, 72, 72, 144,  0           batch_normalization_41[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_46 (Conv3D)              (None, 72, 72, 144,  110608      activation_41[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_41 (Dropout)            (None, 72, 72, 144,  0           conv3d_46[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_40 (Concatenate)    (None, 72, 72, 144,  0           dropout_41[0][0]                 \n",
      "                                                                 concatenate_39[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_42 (BatchNo (None, 72, 72, 144,  1088        concatenate_40[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_42 (Activation)      (None, 72, 72, 144,  0           batch_normalization_42[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_47 (Conv3D)              (None, 72, 72, 144,  117520      activation_42[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_42 (Dropout)            (None, 72, 72, 144,  0           conv3d_47[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_41 (Concatenate)    (None, 72, 72, 144,  0           dropout_42[0][0]                 \n",
      "                                                                 concatenate_40[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_43 (BatchNo (None, 72, 72, 144,  1152        concatenate_41[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_43 (Activation)      (None, 72, 72, 144,  0           batch_normalization_43[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_48 (Conv3D)              (None, 72, 72, 144,  124432      activation_43[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_43 (Dropout)            (None, 72, 72, 144,  0           conv3d_48[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_42 (Concatenate)    (None, 72, 72, 144,  0           dropout_43[0][0]                 \n",
      "                                                                 concatenate_41[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_transpose_3 (Conv3DTrans (None, 144, 144, 144 525376      concatenate_42[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_43 (Concatenate)    (None, 144, 144, 144 0           conv3d_transpose_3[0][0]         \n",
      "                                                                 concatenate_2[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_44 (BatchNo (None, 144, 144, 144 640         concatenate_43[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_44 (Activation)      (None, 144, 144, 144 0           batch_normalization_44[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_49 (Conv3D)              (None, 144, 144, 144 69136       activation_44[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_44 (Dropout)            (None, 144, 144, 144 0           conv3d_49[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_44 (Concatenate)    (None, 144, 144, 144 0           dropout_44[0][0]                 \n",
      "                                                                 concatenate_43[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_45 (BatchNo (None, 144, 144, 144 704         concatenate_44[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_45 (Activation)      (None, 144, 144, 144 0           batch_normalization_45[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_50 (Conv3D)              (None, 144, 144, 144 76048       activation_45[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_45 (Dropout)            (None, 144, 144, 144 0           conv3d_50[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_45 (Concatenate)    (None, 144, 144, 144 0           dropout_45[0][0]                 \n",
      "                                                                 concatenate_44[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_46 (BatchNo (None, 144, 144, 144 768         concatenate_45[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_46 (Activation)      (None, 144, 144, 144 0           batch_normalization_46[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_51 (Conv3D)              (None, 144, 144, 144 82960       activation_46[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_46 (Dropout)            (None, 144, 144, 144 0           conv3d_51[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_46 (Concatenate)    (None, 144, 144, 144 0           dropout_46[0][0]                 \n",
      "                                                                 concatenate_45[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_52 (Conv3D)              (None, 144, 144, 144 2508        concatenate_46[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "activation_47 (Activation)      (None, 144, 144, 144 0           conv3d_52[0][0]                  \n",
      "==================================================================================================\n",
      "Total params: 9,995,948\n",
      "Trainable params: 9,969,580\n",
      "Non-trainable params: 26,368\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import tensorflow.keras as keras\n",
    "\n",
    "from tensorflow.keras import Model, Input\n",
    "from tensorflow.keras.layers import Dense, Flatten, Conv3D, MaxPooling3D, concatenate, Conv3DTranspose, add, multiply\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "\n",
    "\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.layers import Input, MaxPooling3D\n",
    "from tensorflow.keras.layers import Conv3D, Conv3DTranspose\n",
    "from tensorflow.keras.layers import BatchNormalization\n",
    "from tensorflow.keras.layers import Activation, Dropout\n",
    "from tensorflow.keras.layers import concatenate\n",
    "from tensorflow.keras import backend as K\n",
    "from tensorflow.keras.regularizers import l2\n",
    "from tensorflow.keras.utils import plot_model\n",
    "from tensorflow.keras.layers import Dropout\n",
    "from tensorflow.keras.regularizers import l2\n",
    "\n",
    "\n",
    "def denseBlock(t, nb_layers):\n",
    "    for _ in range(nb_layers):\n",
    "        tmp = t\n",
    "        t = BatchNormalization(axis=-1,\n",
    "                                gamma_regularizer=l2(0.0001),\n",
    "                                beta_regularizer=l2(0.0001))(t)\n",
    "\n",
    "        t = Activation('relu')(t)\n",
    "        t = Conv3D(16, kernel_size=(3, 3, 3), padding='same', kernel_initializer='he_uniform', data_format='channels_last')(t)\n",
    "        t = Dropout(0.2)(t)\n",
    "        t = concatenate([t, tmp])\n",
    "    return t\n",
    "\n",
    "def transitionDown(t, nb_features):\n",
    "    t = BatchNormalization(axis=-1,\n",
    "                            gamma_regularizer=l2(0.0001),\n",
    "                            beta_regularizer=l2(0.0001))(t)\n",
    "    t = Activation('relu')(t)\n",
    "    t = Conv3D(nb_features, kernel_size=(1, 1, 1), padding='same', kernel_initializer='he_uniform', data_format='channels_last')(t)\n",
    "    t = Dropout(0.2)(t)\n",
    "    #t = MaxPooling3D(pool_size=(2, 2, 2), strides=2, padding='same', data_format='channels_last')(t)\n",
    "    t = Conv3D(nb_features, kernel_size=(1, 1, 1),strides=(2,2,1), padding='same', data_format='channels_last')(t)\n",
    "    \n",
    "    return t\n",
    "\n",
    "def Tiramisu3D(layer_per_block, n_pool=4, growth_rate=16):\n",
    "    input_layer = Input(shape=INPUT_SHAPE)\n",
    "    t = Conv3D(48, kernel_size=(3, 3, 3), strides=(1, 1, 1), padding='same')(input_layer)\n",
    "\n",
    "    #dense block\n",
    "    nb_features = 48\n",
    "    skip_connections = []\n",
    "    for i in range(n_pool):\n",
    "        t = denseBlock(t, layer_per_block[i])\n",
    "        skip_connections.append(t)\n",
    "        nb_features += growth_rate * layer_per_block[i]\n",
    "        t = transitionDown(t, nb_features)\n",
    "\n",
    "    t = denseBlock(t, layer_per_block[n_pool]) # bottle neck\n",
    "\n",
    "    skip_connections = skip_connections[::-1] #subvert the array\n",
    "\n",
    "    for i in range(n_pool):\n",
    "        keep_nb_features = growth_rate * layer_per_block[n_pool + i]\n",
    "        t = Conv3DTranspose(keep_nb_features, strides=(2,2,1), kernel_size=(3, 3, 3), padding='same', data_format='channels_last')(t) # transition Up\n",
    "        t = concatenate([t, skip_connections[i]])\n",
    "\n",
    "        t = denseBlock(t, layer_per_block[n_pool+i+1])\n",
    "\n",
    "    t = Conv3D(12, kernel_size=(1, 1, 1), padding='same', kernel_initializer='he_uniform', data_format='channels_last')(t)\n",
    "    output_layer = Activation('softmax')(t)\n",
    "    return Model(inputs=input_layer, outputs=output_layer)\n",
    "\n",
    "layer_per_block =  [3, 4, 5, 6, 7, 6, 5, 4, 3]\n",
    "#layer_per_block = [4, 5, 7, 10, 12, 15, 12, 10, 7, 5, 4]\n",
    "tiramisu = Tiramisu3D(layer_per_block)\n",
    "\n",
    "tiramisu.summary()\n",
    "#plot_model(model, to_file='model.png')\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Attention gate input"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import tensorflow.keras as keras\n",
    "\n",
    "from tensorflow.keras import Model, Input\n",
    "from tensorflow.keras.layers import Dense, Flatten, Conv3D, MaxPooling3D, concatenate, Conv3DTranspose, add, multiply\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "\n",
    "\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.layers import Input, MaxPooling3D\n",
    "from tensorflow.keras.layers import Conv3D, Conv3DTranspose\n",
    "from tensorflow.keras.layers import BatchNormalization\n",
    "from tensorflow.keras.layers import Activation, Dropout\n",
    "from tensorflow.keras.layers import concatenate\n",
    "from tensorflow.keras import backend as K\n",
    "from tensorflow.keras.regularizers import l2\n",
    "from tensorflow.keras.utils import plot_model\n",
    "from tensorflow.keras.layers import Dropout\n",
    "from tensorflow.keras.regularizers import l2\n",
    "\n",
    "PROJECT_NAME = 'hector_attention_gate'\n",
    "project_name = PROJECT_NAME\n",
    "\n",
    "def attention_block_3d(x, g, inter_channel, data_format='channels_last'):\n",
    "    # theta_x(?,g_height,g_width,inter_channel)\n",
    "\n",
    "    theta_x = Conv3D(inter_channel, [1, 1, 1], strides=[1, 1, 1], data_format=data_format)(x)\n",
    "\n",
    "    # phi_g(?,g_height,g_width,inter_channel)\n",
    "\n",
    "    phi_g = Conv3D(inter_channel, [1, 1, 1], strides=[1, 1, 1], data_format=data_format)(g)\n",
    "\n",
    "    # f(?,g_height,g_width,inter_channel)\n",
    "\n",
    "    f = Activation('relu')(add([theta_x, phi_g]))\n",
    "\n",
    "    # psi_f(?,g_height,g_width,1)\n",
    "\n",
    "    psi_f = Conv3D(1, [1, 1, 1], strides=[1, 1, 1], data_format=data_format)(f)\n",
    "\n",
    "    rate = Activation('sigmoid')(psi_f)\n",
    "\n",
    "    # rate(?,x_height,x_width)\n",
    "\n",
    "    # att_x(?,x_height,x_width,x_channel)\n",
    "\n",
    "    att_x = multiply([x, rate])\n",
    "\n",
    "    return att_x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_3\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_4 (InputLayer)            [(None, 144, 144, 14 0                                            \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_91 (Conv3D)              (None, 144, 144, 144 1760        input_4[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_83 (BatchNo (None, 144, 144, 144 128         conv3d_91[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_66 (Activation)      (None, 144, 144, 144 0           batch_normalization_83[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_91 (Concatenate)    (None, 144, 144, 144 0           input_4[0][0]                    \n",
      "                                                                 activation_66[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_92 (Conv3D)              (None, 144, 144, 144 29408       concatenate_91[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_84 (BatchNo (None, 144, 144, 144 576         conv3d_92[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_67 (Activation)      (None, 144, 144, 144 0           batch_normalization_84[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "dropout_65 (Dropout)            (None, 144, 144, 144 0           activation_67[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_92 (Concatenate)    (None, 144, 144, 144 0           input_4[0][0]                    \n",
      "                                                                 dropout_65[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling3d_8 (MaxPooling3D)  (None, 72, 72, 72, 3 0           concatenate_92[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_93 (Conv3D)              (None, 72, 72, 72, 6 58816       max_pooling3d_8[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_85 (BatchNo (None, 72, 72, 72, 6 256         conv3d_93[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_68 (Activation)      (None, 72, 72, 72, 6 0           batch_normalization_85[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_93 (Concatenate)    (None, 72, 72, 72, 9 0           max_pooling3d_8[0][0]            \n",
      "                                                                 activation_68[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_94 (Conv3D)              (None, 72, 72, 72, 6 169408      concatenate_93[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_86 (BatchNo (None, 72, 72, 72, 6 288         conv3d_94[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_69 (Activation)      (None, 72, 72, 72, 6 0           batch_normalization_86[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "dropout_66 (Dropout)            (None, 72, 72, 72, 6 0           activation_69[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_94 (Concatenate)    (None, 72, 72, 72, 9 0           max_pooling3d_8[0][0]            \n",
      "                                                                 dropout_66[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling3d_9 (MaxPooling3D)  (None, 36, 36, 36, 9 0           concatenate_94[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_95 (Conv3D)              (None, 36, 36, 36, 1 338816      max_pooling3d_9[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_87 (BatchNo (None, 36, 36, 36, 1 512         conv3d_95[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_70 (Activation)      (None, 36, 36, 36, 1 0           batch_normalization_87[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_95 (Concatenate)    (None, 36, 36, 36, 2 0           max_pooling3d_9[0][0]            \n",
      "                                                                 activation_70[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_96 (Conv3D)              (None, 36, 36, 36, 1 781184      concatenate_95[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_88 (BatchNo (None, 36, 36, 36, 1 144         conv3d_96[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_71 (Activation)      (None, 36, 36, 36, 1 0           batch_normalization_88[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "dropout_67 (Dropout)            (None, 36, 36, 36, 1 0           activation_71[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_96 (Concatenate)    (None, 36, 36, 36, 2 0           max_pooling3d_9[0][0]            \n",
      "                                                                 dropout_67[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling3d_10 (MaxPooling3D) (None, 18, 18, 18, 2 0           concatenate_96[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_97 (Conv3D)              (None, 18, 18, 18, 2 1562368     max_pooling3d_10[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_89 (BatchNo (None, 18, 18, 18, 2 1024        conv3d_97[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_72 (Activation)      (None, 18, 18, 18, 2 0           batch_normalization_89[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_97 (Concatenate)    (None, 18, 18, 18, 4 0           max_pooling3d_10[0][0]           \n",
      "                                                                 activation_72[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_98 (Conv3D)              (None, 18, 18, 18, 2 3331840     concatenate_97[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_90 (BatchNo (None, 18, 18, 18, 2 72          conv3d_98[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_73 (Activation)      (None, 18, 18, 18, 2 0           batch_normalization_90[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "dropout_68 (Dropout)            (None, 18, 18, 18, 2 0           activation_73[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_98 (Concatenate)    (None, 18, 18, 18, 4 0           max_pooling3d_10[0][0]           \n",
      "                                                                 dropout_68[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling3d_11 (MaxPooling3D) (None, 9, 9, 9, 482) 0           concatenate_98[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_99 (Conv3D)              (None, 9, 9, 9, 512) 6663680     max_pooling3d_11[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_91 (BatchNo (None, 9, 9, 9, 512) 2048        conv3d_99[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_74 (Activation)      (None, 9, 9, 9, 512) 0           batch_normalization_91[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_99 (Concatenate)    (None, 9, 9, 9, 994) 0           max_pooling3d_11[0][0]           \n",
      "                                                                 activation_74[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_100 (Conv3D)             (None, 9, 9, 9, 512) 13741568    concatenate_99[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_92 (BatchNo (None, 9, 9, 9, 512) 36          conv3d_100[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "activation_75 (Activation)      (None, 9, 9, 9, 512) 0           batch_normalization_92[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "dropout_69 (Dropout)            (None, 9, 9, 9, 512) 0           activation_75[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_100 (Concatenate)   (None, 9, 9, 9, 994) 0           max_pooling3d_11[0][0]           \n",
      "                                                                 dropout_69[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_transpose_12 (Conv3DTran (None, 18, 18, 18, 2 2035968     concatenate_100[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_101 (Concatenate)   (None, 18, 18, 18, 7 0           conv3d_transpose_12[0][0]        \n",
      "                                                                 concatenate_98[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_101 (Conv3D)             (None, 18, 18, 18, 2 5101312     concatenate_101[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_93 (BatchNo (None, 18, 18, 18, 2 72          conv3d_101[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "activation_76 (Activation)      (None, 18, 18, 18, 2 0           batch_normalization_93[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_102 (Concatenate)   (None, 18, 18, 18, 9 0           concatenate_101[0][0]            \n",
      "                                                                 activation_76[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_102 (Conv3D)             (None, 18, 18, 18, 2 6870784     concatenate_102[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_94 (BatchNo (None, 18, 18, 18, 2 72          conv3d_102[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "activation_77 (Activation)      (None, 18, 18, 18, 2 0           batch_normalization_94[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "dropout_70 (Dropout)            (None, 18, 18, 18, 2 0           activation_77[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_103 (Concatenate)   (None, 18, 18, 18, 9 0           concatenate_101[0][0]            \n",
      "                                                                 dropout_70[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_transpose_13 (Conv3DTran (None, 36, 36, 36, 1 1017984     concatenate_103[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_104 (Concatenate)   (None, 36, 36, 36, 3 0           conv3d_transpose_13[0][0]        \n",
      "                                                                 concatenate_96[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_103 (Conv3D)             (None, 36, 36, 36, 1 1223552     concatenate_104[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_95 (BatchNo (None, 36, 36, 36, 1 144         conv3d_103[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "activation_78 (Activation)      (None, 36, 36, 36, 1 0           batch_normalization_95[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_105 (Concatenate)   (None, 36, 36, 36, 4 0           concatenate_104[0][0]            \n",
      "                                                                 activation_78[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_104 (Conv3D)             (None, 36, 36, 36, 1 1665920     concatenate_105[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_96 (BatchNo (None, 36, 36, 36, 1 144         conv3d_104[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "activation_79 (Activation)      (None, 36, 36, 36, 1 0           batch_normalization_96[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "dropout_71 (Dropout)            (None, 36, 36, 36, 1 0           activation_79[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_106 (Concatenate)   (None, 36, 36, 36, 4 0           concatenate_104[0][0]            \n",
      "                                                                 dropout_71[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_transpose_14 (Conv3DTran (None, 72, 72, 72, 6 246848      concatenate_106[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_107 (Concatenate)   (None, 72, 72, 72, 1 0           conv3d_transpose_14[0][0]        \n",
      "                                                                 concatenate_94[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_105 (Conv3D)             (None, 72, 72, 72, 6 280000      concatenate_107[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_97 (BatchNo (None, 72, 72, 72, 6 288         conv3d_105[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "activation_80 (Activation)      (None, 72, 72, 72, 6 0           batch_normalization_97[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_108 (Concatenate)   (None, 72, 72, 72, 2 0           concatenate_107[0][0]            \n",
      "                                                                 activation_80[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_106 (Conv3D)             (None, 72, 72, 72, 6 390592      concatenate_108[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_98 (BatchNo (None, 72, 72, 72, 6 288         conv3d_106[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "activation_81 (Activation)      (None, 72, 72, 72, 6 0           batch_normalization_98[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "dropout_72 (Dropout)            (None, 72, 72, 72, 6 0           activation_81[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_109 (Concatenate)   (None, 72, 72, 72, 2 0           concatenate_107[0][0]            \n",
      "                                                                 dropout_72[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_transpose_15 (Conv3DTran (None, 144, 144, 144 57888       concatenate_109[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_110 (Concatenate)   (None, 144, 144, 144 0           conv3d_transpose_15[0][0]        \n",
      "                                                                 concatenate_92[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_107 (Conv3D)             (None, 144, 144, 144 57056       concatenate_110[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_99 (BatchNo (None, 144, 144, 144 576         conv3d_107[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "activation_82 (Activation)      (None, 144, 144, 144 0           batch_normalization_99[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_111 (Concatenate)   (None, 144, 144, 144 0           concatenate_110[0][0]            \n",
      "                                                                 activation_82[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_108 (Conv3D)             (None, 144, 144, 144 84704       concatenate_111[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_100 (BatchN (None, 144, 144, 144 576         conv3d_108[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "activation_83 (Activation)      (None, 144, 144, 144 0           batch_normalization_100[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "dropout_73 (Dropout)            (None, 144, 144, 144 0           activation_83[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_112 (Concatenate)   (None, 144, 144, 144 0           concatenate_110[0][0]            \n",
      "                                                                 dropout_73[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv3d_109 (Conv3D)             (None, 144, 144, 144 99          concatenate_112[0][0]            \n",
      "==================================================================================================\n",
      "Total params: 45,718,799\n",
      "Trainable params: 45,715,177\n",
      "Non-trainable params: 3,622\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "\n",
    "def denseBlock(t, nb_features, drop_rate=0.2):\n",
    "    tmp = t\n",
    "    t = Conv3D(nb_features, kernel_size=(3, 3, 3), padding='same', kernel_initializer='he_uniform', data_format='channels_last')(t)\n",
    "    t = BatchNormalization(axis=-1,\n",
    "                            gamma_regularizer=l2(0.0001),\n",
    "                            beta_regularizer=l2(0.0001))(t)\n",
    "\n",
    "    t = Activation('relu')(t)\n",
    "    t = concatenate([tmp, t], axis=-1)\n",
    "    t = Conv3D(nb_features, kernel_size=(3, 3, 3), padding='same', kernel_initializer='he_uniform', data_format='channels_last')(t)\n",
    "    t = BatchNormalization(axis=1,\n",
    "                            gamma_regularizer=l2(0.0001),\n",
    "                            beta_regularizer=l2(0.0001))(t)\n",
    "    t = Activation('relu')(t)\n",
    "    t = Dropout(drop_rate)(t)\n",
    "    t = concatenate([tmp, t], axis=-1)\n",
    "    return t\n",
    "\n",
    "def TransposeDenseblock(t, skip, nb_features, drop_rate=0.2):\n",
    "\n",
    "    tmp = concatenate([Conv3DTranspose(nb_features, (2, 2, 2), strides=(2, 2, 2), padding='same')(t), skip], axis=-1)\n",
    "    \n",
    "    t = Conv3D(nb_features, kernel_size=(3, 3, 3), padding='same', kernel_initializer='he_uniform', data_format='channels_last')(tmp)\n",
    "    t = BatchNormalization(axis=1,\n",
    "                            gamma_regularizer=l2(0.0001),\n",
    "                            beta_regularizer=l2(0.0001))(t)\n",
    "\n",
    "    t = Activation('relu')(t)\n",
    "    t = concatenate([tmp, t], axis=-1)\n",
    "    t = Conv3D(nb_features, kernel_size=(3, 3, 3), padding='same', kernel_initializer='he_uniform', data_format='channels_last')(t)\n",
    "    t = BatchNormalization(axis=1,\n",
    "                            gamma_regularizer=l2(0.0001),\n",
    "                            beta_regularizer=l2(0.0001))(t)\n",
    "    t = Activation('relu')(t)\n",
    "    t = Dropout(drop_rate)(t)\n",
    "    t = concatenate([tmp, t], axis=-1)\n",
    "    return t\n",
    "\n",
    "\n",
    "def dense_BN_unet(att_input= True, inter_channel = 1):\n",
    "    feature_maps = [32, 64, 128, 256, 512]\n",
    "    drop_rate = 0.2\n",
    "    \n",
    "    inputs = Input(INPUT_SHAPE)\n",
    "    #if att_input:\n",
    "    #    att = attention_block_3d(inputs[...,0,np.newaxis], inputs[...,1, np.newaxis], inter_channel)\n",
    "    #    conv11 = Conv3D(32, (3, 3, 3), activation='relu', padding='same')(att)\n",
    "    #else:\n",
    "    #    conv11 = Conv3D(32, (3, 3, 3), activation='relu', padding='same')(inputs)\n",
    "    \n",
    "    skips = []\n",
    "    x = inputs\n",
    "    for nb_features in feature_maps[:-1]:\n",
    "        x = denseBlock(x, nb_features, drop_rate=drop_rate)\n",
    "        skips.append(x)\n",
    "        x = MaxPooling3D(pool_size=(2, 2, 2))(x)\n",
    "    \n",
    "    x = denseBlock(x, feature_maps[-1], drop_rate=drop_rate)\n",
    "    \n",
    "    feature_maps.reverse()\n",
    "    skips.reverse()\n",
    "\n",
    "    for i in range(len(skips)):   \n",
    "        x = TransposeDenseblock(x, skips[i], feature_maps[1 + i], drop_rate=0.2) \n",
    "\n",
    "    conv10 = Conv3D(1, (1, 1, 1), activation='sigmoid')(x)\n",
    "\n",
    "    model = Model(inputs=[inputs], outputs=[conv10])\n",
    "\n",
    "    model.summary()\n",
    "\n",
    "    return model\n",
    "\n",
    "model = dense_BN_unet(att_input= False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.callbacks import ModelCheckpoint, CSVLogger\n",
    "project_name = PROJECT_NAME\n",
    "\n",
    "def dice_coef(y_true, y_pred):\n",
    "    y_true_f = K.flatten(y_true)\n",
    "    y_pred_f = K.flatten(y_pred)\n",
    "    intersection = K.sum(y_true_f * y_pred_f)\n",
    "    return (2. * intersection +K.epsilon()) / (K.sum(y_true_f) + K.sum(y_pred_f) + K.epsilon())\n",
    "\n",
    "\n",
    "def dice_coef_loss(y_true, y_pred):\n",
    "    return 1-dice_coef(y_true, y_pred)\n",
    "\n",
    "def binary_focal_loss_fixed(y_true, y_pred, gamma=2., alpha=.25):\n",
    "    \"\"\"\n",
    "    :param y_true: A tensor of the same shape as `y_pred`\n",
    "    :param y_pred:  A tensor resulting from a sigmoid\n",
    "    :return: Output tensor.\n",
    "    \"\"\"\n",
    "    y_true = tf.cast(y_true, tf.float32)\n",
    "    # Define epsilon so that the back-propagation will not result in NaN for 0 divisor case\n",
    "    epsilon = K.epsilon()\n",
    "    # Add the epsilon to prediction value\n",
    "    # y_pred = y_pred + epsilon\n",
    "    # Clip the prediciton value\n",
    "    y_pred = K.clip(y_pred, epsilon, 1.0 - epsilon)\n",
    "    # Calculate p_t\n",
    "    p_t = tf.where(K.equal(y_true, 1), y_pred, 1 - y_pred)\n",
    "    # Calculate alpha_t\n",
    "    alpha_factor = K.ones_like(y_true) * alpha\n",
    "    alpha_t = tf.where(K.equal(y_true, 1), alpha_factor, 1 - alpha_factor)\n",
    "    # Calculate cross entropy\n",
    "    cross_entropy = -K.log(p_t)\n",
    "    weight = alpha_t * K.pow((1 - p_t), gamma)\n",
    "    # Calculate focal loss\n",
    "    loss = weight * cross_entropy\n",
    "    # Sum the losses in mini_batch\n",
    "    loss = K.mean(loss)\n",
    "    return loss\n",
    "\n",
    "def hyper_loss(y_true, y_pred):\n",
    "    return  K.mean(dice_coef_loss(y_true, y_pred) + (0.5*binary_focal_loss_fixed(y_true, y_pred)))\n",
    "\n",
    "def bce_dice_loss(y_true, y_pred):\n",
    "    return 0.5 * tf.keras.losses.binary_crossentropy(y_true, y_pred) - dice_coef(y_true, y_pred)\n",
    "\n",
    "\n",
    "def train(att_input= True, inter_channel = 1, loss_function= dice_coef_loss, retrain = False, L2 = False):\n",
    "    batch_size = BATCH_SIZE\n",
    "    project_name = PROJECT_NAME\n",
    "    print(project_name)\n",
    "    print('-'*30)\n",
    "    print('Creating and compiling model...')\n",
    "    print('-'*30)\n",
    "    \n",
    "    #create model\n",
    "    model = dense_BN_unet(att_input= att_input, inter_channel = inter_channel)\n",
    "    weight_dir = 'weights'\n",
    "    if not os.path.exists(weight_dir):\n",
    "        os.mkdir(weight_dir)\n",
    "        \n",
    "    if retrain != False:\n",
    "        # It can be used to reconstruct the model identically.\n",
    "        model = tf.keras.models.load_model(os.path.join(weight_dir,retrain))\n",
    "        \n",
    "    model_checkpoint = ModelCheckpoint(os.path.join(weight_dir, project_name + '.h5'), \n",
    "                                       monitor='val_loss', save_best_only=True)\n",
    "\n",
    "    log_dir = 'logs'\n",
    "    if not os.path.exists(log_dir):\n",
    "        os.mkdir(log_dir)\n",
    "    csv_logger = CSVLogger(os.path.join(log_dir,  project_name + '.txt'), separator=',', append=False)\n",
    "\n",
    "    print('-'*30)\n",
    "    print('Fitting model...')\n",
    "    print('-'*30)\n",
    "\n",
    "    steps_per_epoch = np.int(np.ceil((len(train_data_list)*(1+ AUG_RATE) )/batch_size))\n",
    "    validation_steps = np.int(np.ceil(len(val_data_list)/batch_size))\n",
    "\n",
    "    print(\"steps_per_epoch = \", steps_per_epoch)\n",
    "    print(\"validation_steps = \", validation_steps)\n",
    "    \n",
    "    from tensorflow.keras import metrics\n",
    "    \n",
    "\n",
    "    model.compile(optimizer=Adam(lr=1e-5), \n",
    "                  loss=loss_function, metrics=[metrics.binary_crossentropy, dice_coef])\n",
    "\n",
    "\n",
    "    #model.fit(imgs_train, imgs_mask_train, batch_size=1, epochs=50, verbose=1, shuffle=True, validation_split=0.10, callbacks=[model_checkpoint, csv_logger])\n",
    "    model.fit(train_dataset, validation_data = valid_dataset, steps_per_epoch = steps_per_epoch,\n",
    "         validation_steps = validation_steps, epochs = EPOCHS, callbacks=[model_checkpoint, csv_logger,PlotLossesKerasTF()],\n",
    "             verbose = 1)\n",
    "\n",
    "    print('-'*30)\n",
    "    print('Training finished')\n",
    "    print('-'*30)\n",
    "    \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "binary_crossentropy\n",
      "\ttraining         \t (min:    0.008, max:    0.625, cur:    0.009)\n",
      "\tvalidation       \t (min:    0.008, max:    0.399, cur:    0.010)\n",
      "dice_coef\n",
      "\ttraining         \t (min:    0.021, max:    0.793, cur:    0.784)\n",
      "\tvalidation       \t (min:    0.027, max:    0.750, cur:    0.750)\n",
      "Loss\n",
      "\ttraining         \t (min:    0.386, max:    1.160, cur:    0.395)\n",
      "\tvalidation       \t (min:    0.431, max:    1.154, cur:    0.431)\n",
      "67/67 [==============================] - 424s 6s/step - loss: 0.3953 - binary_crossentropy: 0.0092 - dice_coef: 0.7835 - val_loss: 0.4307 - val_binary_crossentropy: 0.0098 - val_dice_coef: 0.7504\n",
      "Epoch 26/100\n",
      " 3/67 [>.............................] - ETA: 3:46 - loss: 0.4286 - binary_crossentropy: 0.0056 - dice_coef: 0.7495"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-42-e9d101664d71>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0mos\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0menviron\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"CUDA_VISIBLE_DEVICES\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'1'\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0mPROJECT_NAME\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m'dense_BN_augmentation'\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m \u001b[0mtrain\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0matt_input\u001b[0m\u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m<ipython-input-41-095413d66d98>\u001b[0m in \u001b[0;36mtrain\u001b[0;34m(att_input, inter_channel, loss_function, retrain, L2)\u001b[0m\n\u001b[1;32m     92\u001b[0m     model.fit(train_dataset, validation_data = valid_dataset, steps_per_epoch = steps_per_epoch,\n\u001b[1;32m     93\u001b[0m          \u001b[0mvalidation_steps\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mvalidation_steps\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mepochs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mEPOCHS\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcallbacks\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mmodel_checkpoint\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcsv_logger\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mPlotLossesKerasTF\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 94\u001b[0;31m              verbose = 1)\n\u001b[0m\u001b[1;32m     95\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     96\u001b[0m     \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'-'\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0;36m30\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/tf2/lib/python3.7/site-packages/tensorflow/python/keras/engine/training.py\u001b[0m in \u001b[0;36m_method_wrapper\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m     64\u001b[0m   \u001b[0;32mdef\u001b[0m \u001b[0m_method_wrapper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     65\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_in_multi_worker_mode\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m  \u001b[0;31m# pylint: disable=protected-access\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 66\u001b[0;31m       \u001b[0;32mreturn\u001b[0m \u001b[0mmethod\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     67\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     68\u001b[0m     \u001b[0;31m# Running inside `run_distribute_coordinator` already.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/tf2/lib/python3.7/site-packages/tensorflow/python/keras/engine/training.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_batch_size, validation_freq, max_queue_size, workers, use_multiprocessing)\u001b[0m\n\u001b[1;32m    846\u001b[0m                 batch_size=batch_size):\n\u001b[1;32m    847\u001b[0m               \u001b[0mcallbacks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_train_batch_begin\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 848\u001b[0;31m               \u001b[0mtmp_logs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtrain_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0miterator\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    849\u001b[0m               \u001b[0;31m# Catch OutOfRangeError for Datasets of unknown size.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    850\u001b[0m               \u001b[0;31m# This blocks until the batch has finished executing.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/tf2/lib/python3.7/site-packages/tensorflow/python/eager/def_function.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwds)\u001b[0m\n\u001b[1;32m    578\u001b[0m         \u001b[0mxla_context\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mExit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    579\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 580\u001b[0;31m       \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    581\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    582\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mtracing_count\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_get_tracing_count\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/tf2/lib/python3.7/site-packages/tensorflow/python/eager/def_function.py\u001b[0m in \u001b[0;36m_call\u001b[0;34m(self, *args, **kwds)\u001b[0m\n\u001b[1;32m    609\u001b[0m       \u001b[0;31m# In this case we have created variables on the first call, so we run the\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    610\u001b[0m       \u001b[0;31m# defunned version which is guaranteed to never create variables.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 611\u001b[0;31m       \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_stateless_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# pylint: disable=not-callable\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    612\u001b[0m     \u001b[0;32melif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_stateful_fn\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    613\u001b[0m       \u001b[0;31m# Release the lock early so that multiple threads can perform the call\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/tf2/lib/python3.7/site-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   2418\u001b[0m     \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_lock\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2419\u001b[0m       \u001b[0mgraph_function\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwargs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_maybe_define_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2420\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0mgraph_function\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_filtered_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# pylint: disable=protected-access\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2421\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2422\u001b[0m   \u001b[0;34m@\u001b[0m\u001b[0mproperty\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/tf2/lib/python3.7/site-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36m_filtered_call\u001b[0;34m(self, args, kwargs)\u001b[0m\n\u001b[1;32m   1663\u001b[0m          if isinstance(t, (ops.Tensor,\n\u001b[1;32m   1664\u001b[0m                            resource_variable_ops.BaseResourceVariable))),\n\u001b[0;32m-> 1665\u001b[0;31m         self.captured_inputs)\n\u001b[0m\u001b[1;32m   1666\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1667\u001b[0m   \u001b[0;32mdef\u001b[0m \u001b[0m_call_flat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcaptured_inputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcancellation_manager\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/tf2/lib/python3.7/site-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36m_call_flat\u001b[0;34m(self, args, captured_inputs, cancellation_manager)\u001b[0m\n\u001b[1;32m   1744\u001b[0m       \u001b[0;31m# No tape is watching; skip to running the function.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1745\u001b[0m       return self._build_call_outputs(self._inference_function.call(\n\u001b[0;32m-> 1746\u001b[0;31m           ctx, args, cancellation_manager=cancellation_manager))\n\u001b[0m\u001b[1;32m   1747\u001b[0m     forward_backward = self._select_forward_and_backward_functions(\n\u001b[1;32m   1748\u001b[0m         \u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/tf2/lib/python3.7/site-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36mcall\u001b[0;34m(self, ctx, args, cancellation_manager)\u001b[0m\n\u001b[1;32m    596\u001b[0m               \u001b[0minputs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    597\u001b[0m               \u001b[0mattrs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mattrs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 598\u001b[0;31m               ctx=ctx)\n\u001b[0m\u001b[1;32m    599\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    600\u001b[0m           outputs = execute.execute_with_cancellation(\n",
      "\u001b[0;32m~/anaconda3/envs/tf2/lib/python3.7/site-packages/tensorflow/python/eager/execute.py\u001b[0m in \u001b[0;36mquick_execute\u001b[0;34m(op_name, num_outputs, inputs, attrs, ctx, name)\u001b[0m\n\u001b[1;32m     58\u001b[0m     \u001b[0mctx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mensure_initialized\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     59\u001b[0m     tensors = pywrap_tfe.TFE_Py_Execute(ctx._handle, device_name, op_name,\n\u001b[0;32m---> 60\u001b[0;31m                                         inputs, attrs, num_outputs)\n\u001b[0m\u001b[1;32m     61\u001b[0m   \u001b[0;32mexcept\u001b[0m \u001b[0mcore\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_NotOkStatusException\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     62\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mname\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "EPOCHS = 100\n",
    "os.environ[\"CUDA_VISIBLE_DEVICES\"]='1'\n",
    "PROJECT_NAME = 'dense_BN_augmentation'\n",
    "train(att_input= False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}