Identifying individuals with antisocial personality disorder using resting-state FMRI

Area of Science:

• Neuroscience
• Psychiatry
• Machine Learning

Background:

• Antisocial personality disorder (ASPD) is linked to criminal behavior, necessitating objective diagnostic tools.
• Understanding brain functional connectivity in ASPD is crucial for explaining behavioral abnormalities.

Purpose of the Study:

• To investigate functional connectivity changes in ASPD patients using machine learning and resting-state fMRI.
• To develop a data-driven classifier for objective ASPD diagnosis.

Main Methods:

• Utilized resting-state functional magnetic resonance imaging (fMRI) data from 32 ASPD subjects and 35 controls.
• Employed an exploratory machine learning classifier to analyze brain functional connectivity patterns.
• Performed voxel-based morphometry to assess gray and white matter volumes.

Main Results:

• The machine learning classifier achieved 86.57% accuracy, distinguishing ASPD individuals from controls.
• Identified significant uncoupling between the default mode network and attention network in ASPD.
• Found abnormal gray matter in the parietal lobule and white matter in the precuneus in ASPD patients.

Conclusions:

• Resting-state fMRI with machine learning can effectively identify functional connectivity alterations in ASPD.
• The precuneus, superior parietal gyrus, and cerebellum show high discriminative power for ASPD classification.
• This study provides insights into the neural underpinnings of ASPD and a potential tool for objective diagnosis.