A Topological Fingerprint Encodes Motor Skill at Rest.

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Computational Neuroscience

Background:

• Brain network architecture at rest may encode individual behavioral skills.
• Identifying mechanisms for encoding motor skills from resting-state brain activity is challenging.
• Previous models often rely on high-dimensional data, lacking interpretability.

Purpose of the Study:

• To investigate if resting-state brain interaction architecture represents individual behavioral skills.
• To identify a minimal set of topological features for encoding motor skills.
• To test if brain network topology at rest can model individual manual dexterity.

Main Methods:

• Utilized resting-state electrophysiology data from 86 Human Connectome Project subjects.
• Employed a machine learning procedure to identify a predictive topological fingerprint.
• Performed vulnerability analysis by simulating hub disconnections to assess critical nodes.

Main Results:

• An optimal fingerprint, comprising four connector hubs in the alpha frequency band, accurately modeled individual manual dexterity.
• Two specific hubs were identified as critical, with their disconnection significantly reducing predictive performance.
• A functional 'refocusing' mechanism was proposed, involving hub pruning of external connections with increased dexterity.

Conclusions:

• Resting-state brain network topology, particularly in the alpha band, encodes stable behavioral traits like motor skills.
• A compact, low-dimensional fingerprint can capture individual motor performance.
• The findings suggest a novel electrophysiological mechanism of functional refocusing for skill encoding.