Multimodal imaging measures predict rearrest.

Area of Science:

• Neuroscience
• Forensic Psychology
• Cognitive Neuroscience

Background:

• Previous research indicated hemodynamic activity in the anterior cingulate cortex (ACC) could predict rearrest.
• Error-processing deficits are linked to behavioral outcomes in forensic populations.

Purpose of the Study:

• To evaluate the predictive power of multimodal neuroimaging (hemodynamic activity and event-related potentials) for rearrest.
• To investigate neural correlates of error-processing in incarcerated individuals.

Main Methods:

• Collected event-related potentials (ERPs) and hemodynamic activity during a Go/NoGo task in 45 incarcerated males.
• Utilized neural measures from the ACC, error-related negativity (ERN/Ne), and error positivity (Pe) for analysis.
• Employed Cox regression, logistic regression, and support vector machine (SVM) models for neuroprediction.

Main Results:

• Measures from the Pe and ACC differentiated between individuals who were and were not subsequently rearrested.
• All calculated models successfully predicted rearrest, with SVM demonstrating the strongest predictive accuracy.
• Multimodal SVM models achieved 83.33% accuracy in predicting rearrest with cross-validation.

Conclusions:

• Increased Pe amplitude and decreased ACC activation suggest abnormal error-processing and are associated with a higher risk of rearrest.
• Multimodal neuroimaging approaches, particularly SVM, show significant promise for predicting recidivism.
• These findings highlight the potential of neural markers for informing risk assessment in forensic settings.