This work introduces an extension of sparse paradigm free mapping (SPFM) for multiecho (ME) fMRI: ME-SPFM. Based on the ME-fMRI signal model and L1-norm regularized estimators, ME-SPFM produces voxel-wise estimates of time-varying changes in the transverse relaxation ($$$R_2^*$$$) and the net magnetization ($$$S_0$$$) without prior information about experimental paradigms. Our evaluations demonstrate that ME-SPFM significantly outperforms its SE counterpart in terms of sensitivity and specificity, nearly matching that of traditional model-based analyses. ME-SPFM’s ability to blindly detect individual events at the single-subject level makes it an ideal candidate to explore the time-varying nature of brain activity in experimentally unconstrained paradigms (naturalistic, resting-state) or clinical applications (detection of inter-ictal epileptic events).
This abstract and the presentation materials are available to members only; a login is required.