Fig. 5: Temporal dynamics of taste, odour and crossmodal neural signals.
From: Tastes and retronasal odours evoke a shared flavour-specific neural code in the human insula
A Significant differences in accuracy between decoders trained and tested on taste identity within the same day as opposed to across days. B, C No significant difference in within-day and across-day accuracies for decoders trained and tested on odour identity or using crossmodal partitioning. N = 24. Large dots signify means; boxes show the interquartile ranges (IQR); whiskers show the minimum and maximum ranges using 1.5 IQR. D Representational drift of flavour representations in the insula across runs. Model of the neural Representational Dissimilarity Matrix (RDM) from betas of all four conditions across runs (left) as predicted by the runwise distance and the flavour distance. Darker colours represent higher values. Significant reduction in representational drift in the insula between Session 1 and Session 2. E Same as (D) but using the neural RDM extracted from the orbitofrontal cortex (OFC) searchlight peak, showing no reduction in drift between days. Mixed models included all available data but were not corrected for multiple comparison between partitions and ROIs. Shaded areas show 95% confidence intervals of the linear model. N = 25 participants for Session 1; N = 24 participants for Session 2. Source data are provided as a Source Data file.
