Task-specific functional brain geometry from model maps.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Medical Imaging

Background:

• Functional magnetic resonance imaging (fMRI) measures brain activity via blood oxygen level dependent (BOLD) signals.
• Understanding the brain's functional geometry and resource allocation is crucial for neuroscience.
• Existing methods struggle to quantitatively assess subtle differences in brain interactivity.

Purpose of the Study:

• To propose and validate a novel method, 'model maps,' for representing the intrinsic functional geometry of the brain.
• To establish a quantitative tool for exploring global and local patterns of brain resource allocation.
• To investigate differences in brain interactivity between baseline and reward-related tasks in different subject groups.

Main Methods:

• Developed 'model maps' to encode the coherence of fMRI BOLD signals over time using Markov chains.
• Mapped spatial brain positions to a new metric space where distances reflect signal co-dependencies.
• Applied the method to 29 fMRI time sequences across 4 conditions for two subject groups.

Main Results:

• Model maps successfully represent the functional, not anatomical, geometry of the brain.
• Demonstrated quantitative differences in brain interactivity between baseline and reward tasks.
• Identified significantly lower differentiation in brain interactivity in drug abusers compared to controls.

Conclusions:

• Model maps offer a powerful quantitative tool for analyzing brain functional geometry and resource allocation.
• The findings reveal distinct patterns of brain interactivity in drug abusers, highlighting a previously unquantifiable deficit.
• This approach advances the study of brain function and its alterations in neurological and substance use disorders.