depth-weighted version for wMEM

[Edouard2laire]

This update implements several updates to wMEM and improves the support of fNIRS data.

The addition includes for wMEM:

• Support of depth-weighting
• Support for stable parcellation
• New alpha initialization using MNE ( alpha method = 6)
• New noise covariance calculation (NoiseCov method = 6)
• Baseline shuffling for resting state analysis

Bugfixes:

• wMEM: improve the support of multimodal data fusion.
• fNIRS: remove alpha normalization

Other:

• New graphical interface for MEM
• New default options for cMEM and wMEM (see https://neuroimage.usc.edu/brainstorm/Tutorials/TutBEst#A.28Advanced.29_Change_of_default_parameters_.28v3.0.0.29 )
• Unification of the code between EEG/MEG and NIRS

Tutorials to update:

• https://neuroimage.usc.edu/brainstorm/Tutorials/TutBEst
• https://neuroimage.usc.edu/brainstorm/Tutorials/Epilepsy#Source_analysis:_MEM_inverse_solution