Use qsiprep to inform MCR estimations

[marcelzwiers]

In the MCR_GPUWorker > gpuMCRMWI call add:

        % extradata : Optional additional data
        %   .freqBKG: 3D/4D initial estimation of total field [Hz] (highly recommended)
        %   .pini   : 3D initial estimation of B1 offset [rad]  (highly recommended)
        %   .ff     : 3D/4D fibre fraction map, [x,y,z,nF] (for GRE-DIMWI only)
        %   .theta  : 3D/4D angle between B0 and fibre orientation, [x,y,z, nF] (for GRE-DIMWI only)
        %   .IWF    : 3D volume fractino IC/(IC+EC), [x,y,z] (for GRE-DIMWI only)

And for the MCRWorker > mwi_3cx_2R1R2s_dimwi call add:

% DIMWI
% =====
% 	.theta      : angle between fibre and B0 directions, range [0, pi/2];
%   .ff         : fibre fraction, [row, col, slice, (fibre)]
%   .fo         : fibre direction maps, [row, col, slice, fibre vector, (fibre)]
% 	.icvf       : intra-axonal/cellular water volume fraction, range [0,1]