Quantitative deconvolution of neuronal-related BOLD events with Multi-Echo Sparse Free Paradigm Mapping

This work proposes a novel formulation of the Sparse Free Paradigm Mapping (SPFM) algorithm for Multi-Echo (ME) fMRI that produces estimates of ΔR2* with interpretable units (s-1) without prior timing information of task events. Here we show how ΔR2* time-series estimated with ME-SPFM have physiologically plausible values, in agreement with previously published estimates of ΔR2* for neuronal events. We also show, how having easily interpretable units permits detection of artefactual events that can be easily removed from subsequent analysis and interpretation. Our results suggest that ME-SPFM may provide a pseudo-quantitative method to study the time-varying nature of brain activity in experimentally unconstrained paradigms (naturalistic, resting-state) or clinical applications (detection of inter-ictal epileptic events).