Answers are grounded in this paper and cite their sources.
Learning a motor skill makes movements more accurate and faster, but whether the same circuits support both gains remains unclear. Thirteen adults learned a novel foot-drawing skill over seven sessions. Nine had usable resting-state fMRI and matched behavioural data for the pre-to-post brain analysis. We used PCMCI to estimate directed, time-lagged BOLD dependencies between 112 cortical and subcortical regions and related their change to behavioural gains. Inference used small-sample resampling with family-wise error control. The primary direct change-score analysis yielded no family-wise-error-corrected accuracy or speed association. In the complementary longitudinal interaction analysis, one network-level Post \(× \) accuracy association from VisCent to LimbicA survived maxT correction across 361 blocks ( \(p=0.020\) ), and one Post \(× \) speed edge from left retrosplenial cortex to left putamen reached the maxT threshold across 12,432 edges ( \(p=0.047\) ; Monte-Carlo 95% interval 0.043–0.051). Broader accuracy- and speed-related maps are exploratory. We report the complete \(2× 2\) interaction results, direct accuracy-versus-speed contrasts, nested cross-validation of the baseline analysis, and model-matched sensitivity. These preliminary findings provide replication targets for larger studies of learning-related changes in directed resting-state BOLD coupling.
No takes yet. Share an insight, caveat, or question.
Rezaei et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: