first commit

.gitignore

@@ -0,0 +1,31 @@
+# Windows default autosave extension
+*.asv
+
+# OSX / *nix default autosave extension
+*.m~
+
+# Compiled MEX binaries (all platforms)
+*.mex*
+
+# Packaged app and toolbox files
+*.mlappinstall
+*.mltbx
+
+# Generated helpsearch folders
+helpsearch*/
+
+# Simulink code generation folders
+slprj/
+sccprj/
+
+# Matlab code generation folders
+codegen/
+
+# Simulink autosave extension
+*.autosave
+
+# Simulink cache files
+*.slxc
+
+# Octave session info
+octave-workspace

README.md

@@ -0,0 +1 @@
+# NeurAugVR

erd/main.m

@@ -0,0 +1,98 @@
+%% Add paths
+clear, clc
+addpath('C:/Users/guta_/Documents/MATLAB/eeglab2021.0')
+varsbefore = who; eeglab; varsnew = []; varsnew = setdiff(who, varsbefore); clear(varsnew{:})
+
+%% Load
+load('./data/EEG.mat','EEG');
+
+%% Preprocess (prep. 1)
+bandlims = [1, 40];
+resamp = 250;
+EEG = pop_select(EEG, 'nochannel',{'ECG','EKG'});
+EEG = pop_resample(EEG, resamp);
+EEG = pop_eegfiltnew(EEG, 'locutoff',bandlims(1), 'plotfreqz',0);
+EEG = pop_eegfiltnew(EEG, 'hicutoff',bandlims(2), 'plotfreqz',0);
+EEG = pop_reref(EEG, []);
+
+%% Epoch
+dire = {'L'};
+epochlims = [-5.5, 5.5];
+EEGdi = pop_epoch(EEG, dire, epochlims, 'epochinfo', 'yes');
+
+%% Select channel
+chan = 'C4';
+chn = find(ismember({EEG.chanlocs(:).labels}, chan));
+tubmat = double(EEGdi.data(chn, :, :));  % I use the word 'tub' for a 1x1xN array, and 'tubmat' for a 1xMxN array
+
+%% Time-Frequency decomposition of each trial
+baseline = nan; % Indiferent actually
+scale = 'abs';
+cycles = 0;
+[~, ~, ~, time_vec, freq_vec, ~, ~, amp_vol] = newtimef_trueamp(tubmat, size(tubmat,2), [EEGdi.times(1) EEGdi.times(end)], EEGdi.srate, cycles,'plotitc','off', 'baseline', baseline, 'scale', scale, 'plotersp', 'off');
+time_vec = time_vec/1000;
+amp_vol = double(amp_vol);
+
+%% Calculate power
+pow_vol = amp_vol.^2;
+
+%% Average trials
+pow_mat = mean(pow_vol,3);
+
+%% Calculate ERD%
+baselims = [-4, 0];
+erd_mat = (pow_mat./mean(pow_mat(:,time_vec > baselims(1) & time_vec < baselims(2)),2) - 1)*100;
+
+%% Average alpha's frequency band
+freqlims = [8, 12];
+erd_vec = mean(erd_mat(freq_vec > freqlims(1) & freq_vec < freqlims(2),:),1);
+
+%% Average task's time interval
+tasklims = [0.5, 4];
+erd_val = mean(erd_vec(time_vec > tasklims(1) & time_vec < tasklims(2)));
+
+%% Plot ERSP%
+xlims = [time_vec(1), time_vec(end)];
+ylims = [0, 30];
+clims = [-100,100];
+figure
+imagesc('XData',time_vec,'YData',freq_vec,'CData',erd_mat,clims)
+hold on
+colormap jet
+ch = colorbar;
+set(get(ch,'title'),'string','ERSP%');
+% caxis(clims)
+plot(xlims,freqlims(1)*[1,1],'Color',[0,0,0])
+hold on
+plot(xlims,freqlims(2)*[1,1],'Color',[0,0,0])
+hold on
+plot([0,0],ylims,'Color',[0,0,0],'LineStyle','--')
+hold on
+ylim(ylims)
+xlim(xlims)
+xlabel('Time [s]')
+ylabel('Frequency [Hz]')
+
+%% Plot time-Power
+pow_vec = mean(pow_mat(freq_vec > freqlims(1) & freq_vec < freqlims(2),:),1);
+pow_std_vec = std(pow_mat(freq_vec > freqlims(1) & freq_vec < freqlims(2),:),0,1);
+pow_val = mean(pow_vec(time_vec > tasklims(1) & time_vec < tasklims(2)));
+top_margin_vec = pow_vec + pow_std_vec;
+bot_margin_vec = pow_vec - pow_std_vec;
+
+colors = lines(2);
+figure
+p(1) = plot(time_vec, pow_vec, 'Color', colors(1,:));  % use colors(2,:) for orange color
+hold on
+h = fill([time_vec, time_vec(end:-1:1)], [bot_margin_vec, top_margin_vec(end:-1:1)], colors(1,:),'LineStyle','none'); % we make the shape to fill starting w/ bottom edge from left to right, and then top edge from right to left
+set(h,'facealpha',.3)
+hold on
+text(time_vec(end),pow_val,sprintf(' %d µV^2', round(pow_val)),'FontSize',10, 'Color', colors(1,:))
+xlim(xlims)
+xlabel('Time [s]')
+ylabel('Power [µV^2]')
+legend([p(1)],'Left - C4')
+
+
+
+