• clustering
• MRI
• neuroanatomical
• PCL-R
• psychopathy
• structural

Background: Psychopathy is a personality disorder characterized by core emotional deficits, such as guilt reduction and low empathy, impulsivity and an increased risk for antisocial behaviour. Past research has successfully adopted clustering algorithms to distinguish between different psychopathic traits profiles among offenders. Such a person-centered approach has been suggested as a useful framework to identify specific neurobiological correlates of psychopathy.
Methods: We computed seven morphology measures of 31 cortical ROIs per hemisphere, in a sample of 89 Caucasian male offenders, that were subdivided into four clusters using a k-means clustering algorithm based on PCL-R facets scores.
Results: No correlation between morphology measures and PCL-R scores reached statistical significance after Bonferroni correction, nor did the comparisons between groups of participants based on the standard cut-offs of PCL-R total score. In contrast, comparisons between clusters produced significant results. Prototypical psychopaths showed a significantly increased left retrosplenial cortex volume, compared with each of the other clusters, but also increased surface area and folding index in the same area and increased folding index in the pars orbitalis of the right inferior frontal gyrus, when compared to antisocial offenders.
Conclusions: Our findings support the hypothesis of a neurovelopmental origin of psychopathy and highlight the utility of a person-centered approach to identify its neurobiological correlates.