Figure 6.

Speech-driven STRFs in auditory regions defined by functional connectivity with non-auditory regions. A row (A-E) is dedicated to each auditory region (right insets: region mask displayed on inflated standard topology surfaces and labeled with the corresponding seed region). Regions defined by the main effect of seed connectivity (A-C) are labeled as ‘AC’ and regions defined by the interaction of seed connectivity with speech intelligibility (D-E) are labeled as ‘AC-Int.’ Overall responses (STRF) are shown in the first column, while effects of intelligibility (STRF_between) and responses within intelligible (STRF_intel) and unintelligible (STRF_unintel) trials are shown in the second through fourth columns, respectively. For each region (A-E), the color scale is based on the overall STRF min/max and applied to each of the decomposed STRFs at right to allow a visualization of the relative contribution of each component STRF to the overall STRF. The overall STRFs are annotated with second-level alignment strength (Z) and Bayes Factors (geometric mean across surface nodes) indicating quality of LOO predictions within intelligible (BF_I = BF_intel) and unintelligible (BF_UI = BF_unintel) trials. The decomposed STRFs are annotated with the max amplitude (Amp; STRF_between) or second-level alignment strength (Z; STRF_intel and STRF_unintel) and the proportional relative importance (Imp).