Speakers - PAWC2025

Abstract

Objectives
This ongoing study investigates how dynamic and directed resting‐state connectivity underlie inhibitory control deficits in cocaine use disorder (CUD) and how rTMS modulates these pathways. Specifically, it tests whether altered dynamic connectivity patterns predict lower drift rates and more impulsive thresholds from Hierarchical Drift Diffusion Modeling (HDDM) on the Go/No Go task, and whether rTMS-driven changes in top-down connectivity improve inhibitory performance.

Methodology
Using the cross-sectional SUDMEX-CONN dataset, I will first apply a multilayer network approach to derive graph-theoretical metrics (e.g., flexibility, temporal variability) that capture time-varying connectivity reconfigurations. Meanwhile, I will leverage regression Dynamic Causal Modeling (rDCM) to provide directed connectivity estimates between fronto-parietal and fronto-striatal regions. These metrics will be correlated with HDDM-derived parameters to elucidate how inhibitory control differs between CUD patients and healthy controls. In the longitudinal extension using the SUDMEX-TMS dataset, I will track CUD participants undergoing active or sham rTMS across five time points, assessing whether changes in dynamic/directed connectivity predicted improvements in inhibitory control. 

Expected Results
I anticipate that preliminary cross-sectional analyses will reveal reduced flexibility in frontoparietal hubs and weaker top-down connectivity (PFC→striatum) among CUD participants. These disruptions are expected to correlate with lower drift rates and more impulsive decision thresholds in HDDM. In the longitudinal phase with rTMS, I predict that active stimulation (vs. sham) will lead to enhanced frontostriatal directed connectivity and increased time in a “control-dominant” dynamic state, paralleling gains in inhibitory performance. 

Conclusion
If supported, these results highlight the critical role of flexible, top-down pathways in regulating inhibitory responses. They reinforce notions of flexible hubs and salience-network switching, wherein rapid network reconfiguration is vital for suppressing prepotent actions. Practically, this work may guide personalized neuromodulation strategies demonstrating that rTMS-induced changes in dynamic and directed connectivity can be harnessed to improve evidence accumulation and decision caution.