Poster D111, Monday, March 27, 8:00 – 10:00 am, Pacific Concourse
Strategic adaptation to non-reward prediction error qualities and contextual volatility in fMRI
Daniel S. Kluger1,2, Ricarda I. Schubotz1,2,3; 1University of Muenster, Germany, 2Otto-Creutzfeldt-Center for Cognitive and Behavioral Neuroscience, University of Muenster, Germany, 3University Hospital Cologne, Germany
Prediction errors are deemed necessary for the updating of internal models of the environment, prompting us to stop or shield current action plans and helping us to adapt to environmental features. The aim of the present study was twofold: First, we sought to determine the neural underpinnings of qualitatively different abstract prediction errors in a serial pattern detection task. Second, we investigated predictive processing as a function of statistical context (volatility). We hypothesised that the prospective extent of model-based expectations might be adapted to the stability of respective contexts in such a way that unstable environments call for more frequent comparisons of model-based predictions with sensory input, resulting in incremental predictions. Distinct frontoparietal circuits were found for sequential terminations (inferior frontal gyrus, anterior operculum/insula, and precuneus) and extensions (superior frontal sulcus, posterior cingulate cortex, and angular gyrus), respectively. These findings provide a novel approach of distinguishing non-reward prediction error signals with regard to behavioural consequences they entail. Furthermore, changes in environmental stability resulted in activation of the temporoparietal junction and inferior frontal gyrus for the highly volatile context at potential points of violation (checkpoints). Notably, this effect was not due to fluctuations in stimulus-bound surprise. Data point towards a context-dependent adaptation of predictive strategies. Conceivably, enhanced BOLD responses at sequential checkpoints reflect stepwise processing rather than a full-length prediction. This strategic adjustment presumably relies on the iterant evaluation of model information retrieved from working memory, as suggested by strengthened functional connectivity of the parahippocampal area during epochs of high volatility.
Topic Area: PERCEPTION & ACTION: Other