Model for SALSA MC002 NBS Results Interpretation

The model is designed to analyze image outputs from SALSA MC002 SMA NEWBORN SCREEN ASSAY (MRC Holland).
The machine used is Applied Biosystems 7500 Real-Time PCR System.
The software used is HID Real-Time PCR Analysis Software v1.2. with default export parameters (background grid, axis and legend included).

What is included?

This repository includes:

1. .h5 file (saved model with both structure and weights)
2. static folder with .html page
3. .py flask backend
4. examples of input images
5. .ipynb file used for model training

What should I install in order to run this model?

The prerequisites are following:

1. Python 3.X
2. Tensorflow and it's dependencies
3. Flask library

How can I run it locally?

pip3 install tensorflow python3-flask  
git clone https://github.com/trickovicmatija/SMA.git  
cd SMA
export FLASK_APP=predict_app.py
flask run --host=127.0.0.1  

In order to run the model, you should:

1. Install the prerequisites
2. Clone the repository on local machine via "git clone"
3. Run the flask script
4. Manually navigate to "/static/index.html" in address bar
5. Upload the image and predict the outcome

This is created as a part of Master's thesis at the Centre for Human Molecular Genetics, University of Belgrade - Faculty of Biology.