Assessing causal contributions of parietal cortex to learned cognitive flexibility
Christina Bejjani1, Peter Whitehead1, Anthony Sali2, Yu-Chin Chiu3, Tobias Egner1; 1Duke University, 2Wake Forest University, 3Purdue University
Adaptive behavior is characterized by our ability to create, maintain, and update rules by which we categorize and respond to stimuli across changing contexts (cognitive flexibility). Previously, we deployed an fMRI version of a standard task-switching paradigm in which the frequency of switch trials varies over time, ranging from 25% to 50% to 75%. Consistent with previous literature, we found a behavioral interaction between trial type (switch vs. repeat) and switch likelihood (25%, 50%, 75%) such that switch costs were reduced when switch trials were more frequent (“learned switch-readiness”). We then fit a reinforcement learning model to the behavioral data, tracking how participants learn the likelihood of switch trials over time. Using the trial-wise prediction error estimates in model-based fMRI analyses, we found that activity within the left inferior parietal cortex (LIPC) prominently tracked trial-by-trial updating of switch likelihood (unsigned prediction error), implicating this region as a potential key player in learning switch-readiness. In the present experiment, we followed up on this correlational finding with a causality test: we probed whether the normal function of LIPC is necessary for adapting switch readiness to different contexts by disrupting its activity via fMRI-guided TMS. Comparing behavioral performance under LIPC TMS with control site (vertex) TMS, we found little preliminary evidence (n = 18; preregistered target n = 26) that the LIPC is causally involved in learned cognitive flexibility. This suggests that a wider network of interacting brain regions beyond the LIPC likely mediates learned changes in switch-readiness.
Topic Area: EXECUTIVE PROCESSES: Goal maintenance & switching