Causal Evidence for Learning-Dependent Frontal Lobe Contributions to Cognitive Control

Area of Science:

• Cognitive neuroscience
• Computational psychiatry
• Neuroimaging and brain stimulation

Background:

• The lateral prefrontal cortex (LPFC) is crucial for top-down control of perception, prioritizing relevant sensory information for goal-directed behavior.
• Excessive cognitive control can be detrimental, necessitating regulatory mechanisms that adapt to environmental demands.
• The flexible control model (FCM) proposes that LPFC control is regulated by a reinforcement-learning system predicting cognitive demand based on environmental statistics.

Purpose of the Study:

• To causally test the hypothesis that LPFC-based cognitive control is engaged based on anticipated cognitive demand, as predicted by the FCM.
• To investigate the role of a self-adjusting reinforcement-learning mechanism in regulating control engagement in dynamic environments.
• To determine the specific learning signals that regulate top-down biases in stimulus processing.

Main Methods:

• Combined computational modeling (FCM) with transcranial magnetic stimulation (TMS) for a causal investigation.
• Participants completed a nonstationary Stroop task with dynamically changing conflict probabilities.
• TMS was applied over the LPFC or a control site on each trial before stimulus onset.

Main Results:

• In control conditions, participants exhibited performance fluctuations consistent with demand predictions derived from trial history, aligning with FCM predictions.
• Transiently perturbing the LPFC with TMS abolished these adaptive performance fluctuations.
• TMS over the LPFC did not impair basic cognitive or motor functions.

Conclusions:

• The study provides causal evidence for a learning-based account of cognitive control, supporting the FCM's role in adaptive regulation.
• Findings delineate the nature of learning signals that govern the engagement of prefrontal cortex control functions.
• This research demonstrates the critical role of the LPFC in flexibly adjusting cognitive control based on predicted environmental demands.