Asymmetric coding of reward prediction errors in human insula and dorsomedial prefrontal cortex

Abstract

The signed value and unsigned salience of reward prediction errors (RPEs) are critical to understanding reinforcement learning (RL) and cognitive control. Dorsomedial prefrontal cortex (dMPFC) and insula (INS) are key regions for integrating reward and surprise information, but conflicting evidence for both signed and unsigned activity has led to competing proposals for the nature of RPE representations in these brain areas. Recently, the distributional RL theory (dRL) has been used to explain RPE coding diversity in the rodent midbrain by proposing that dopaminergic neurons have differential sensitivity to positive and negative RPEs. Here, we use intracranially recorded high frequency activity (HFA) to show that thisasymmetric scalingstrategy captures RPE coding diversity in human dMPFC and INS. We found neural populations responding to valence-specific positive and negative RPEs, as well as unsigned RPE salience, which are spatially interleaved within each region. Furthermore, directional connectivity estimates suggest a leading role of INS in communicating positive and unsigned RPEs to dMPFC. These findings support asymmetric scaling across distinct but intermingled neural populations as a core principle in RPE coding, expand the scope of dRL, and reconcile longstanding theoretical debates on the role of dMPFC and INS in RL and cognitive control.

An intracranial dissection of human escape circuits

“…We computed the directed FC between all pairs of electrodes across different brain lobes. To obtain the directed FC, we used the time-lagged cross-correlation analysis, as described in [45], with a lag range of −300 to 300 ms. Subsequently, we examined the directed FC within the contexts of fast-attacking and slow-attacking scenarios, both at the average level across all electrode pairs and at the individual electrode pair level.…”

“…Functional connectivity (FC) between two ROIs was assessed by computing Pearson correlations between all inter-lobe electrode pairs. To estimate the directed FC, we employed the time-lagged cross-correlation analysis, as described in the study conducted by [45]. The lag range of cross-correlation is −300 to 300 ms, with a step size of 10 ms.…”

Changes in rest-activity rhythms in adolescents as they age: associations with brain changes and behavior in the ABCD study

“…Changes in sleep timing were associated with change in gray matter predominantly in subcortical volumes. Previous studies in adolescents and adults 43,44 had reported a relationship between greater eveningness and larger GMV and cortical thickness in the medial prefrontal cortex, insula and precuneus, which are regions involved in reward evaluation and processing 45 and in self-referential thinking 46 . A UK biobank study in older adults, reported that in addition to an association of eveningness with GMV in cortical regions there was also an association with subcortical regions including greater GMV in nucleus accumbens, caudate, putamen, thalamus and pallidum 47 .…”

“…Previous studies reported a relationship between greater eveningness and larger GMV and cortical thickness in the medial prefrontal cortex, insula and precuneus, which are regions involved in reward evaluation and processing (Ciaramelli et al, 2021;Hoy et al, 2023) and in self-referential thinking (Cavanna & Trimble, 2006), in adolescents and adults (Rosenberg, Jacobs, Maximov, Reske, & Shah, 2018;Vulser et al, 2023). In a UK biobank study, in addition to cortical regions, GMV associations with greater eveningness were found for subcortical regions including greater GMV in bilateral nucleus accumbens, caudate, putamen, thalamus and pallidum in older adults (Norbury, 2020).…”