Fig. 3: Crossmodal decoding accuracy is driven by the extent of tuning index overlap in all three ROIs.
From: Tastes and retronasal odours evoke a shared flavour-specific neural code in the human insula
A Contrasting the statistical parametric maps (T-maps) of sweet against savoury in both modalities results in a spatial map of tuning indices to each condition. Tuning index overlap can be derived by correlating tuning index maps of each modality. B–D Tuning index overlap in a given ROI is strongly correlated with crossmodal decoder accuracy in the leave-one-subject-out ROI (uncorrected Pearson correlations for multiple comparisons, r = .507, p = .010) (B), the granular insula ROI (r = .557, p = .004) (C) and the dysgranular and agranular insula ROI (r = .535, p = .006) (D). Shaded areas indicate 95% simultaneous prediction bounds. N = 25 participants for panels (B–D). Source data are provided as a Source Data file.
