%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% DESCRIPTION
% This is supposed to be the main script used to run total activation and
% clustering of total activation outputs into innovation-driven
% co-activation patterns
%
% REFERENCES
% For theoretical references about those techniques and applications to 
% functional brain imaging, see the following:
%
% - A Signal Processing Approach to Generalized 1-D Total Variation
% (Karahanoglu et al., IEEE Transactions on Signal Processing, vol. 59, no.
% 11, November 2011) for details on the linear operator used in total
% activation (deconvolution coupled to derivation) and its discrete
% implementation
%
% - Total activation: fMRI deconvolution through spatio-temporal
% regularization (Karahanoglu et al., Neuroimage, vol. 73, pp. 121-134,
% 2013) for an overview of total activation in its first version (still
% used as such for temporal regularization; see Appendix A. for an outline
% of the algorithm)
%
% - Transient brain activity disentangles fMRI resting-state dynamics in
% terms of spatially and temporally overlapping networks (Karahanoglu et
% al., Nature communications, DOI: 10.1038/ncomms8751, 2015) for an
% application of total activation to resting-state brain data, and the
% introduction of the thresholding and iCAP steps
%
% - Regularized spatiotemporal deconvolution of fMRI data using gray-matter
% constrained total variation (Farouj et al., ISBI abstract, 2016) for an
% overview of the presently used version of spatial regularization
%
%
% ACKNOWLEDGMENTS
% The present codes are adaptations from the initial version developed
% by Dr Isik Karahanoglu, former PhD student at MIP:Lab, and Dr Younes
% Farouj, former visiting PhD student and now post-doc at MIP:Lab
%
% UPDATES ON UTILITIES
% V1.0, December 9th 2016, Thomas Bolton: simplified total activation 
% scripts running for one subject (not yet functional)
% V1.1, December 23rd 2016, Thomas Bolton: finished modifying total
% activation routines for one subject
% V1.2, December 28th 2016, Thomas Bolton: everything written down as
% functions for total activation, and tested on .img/.hdr functional and
% segmentation files
% V2.0, May 2018, Daniela Zoelle: including of updates of all lab members:
%     - changed structure of results saving
%     - separated TA, thesholding, clustering and regression in different
%     functions
%     - included checking whether TA has already been done
%
%       Jun 2018, Younes Farouj:
%     - Check writing permission
%     - Check post-interpolation data size 

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

clc;clear all;close all;
% Adds the paths required for total activation
AddPaths();


dwtmode('sym') % Default mode


% EO: Set whether to use GPU implementation (= 1) or original Matlab.
param.use_cuda = 0;

% setting up all parameters to run TA
Inputs_TA_DATA_OpenfMRI      % OpenfMRI test
Inputs_TA

% run TA
Run_TA(param);