Morphometrical brain markers of sex difference

Abstract

Many major neuropsychiatric pathologies, some of which appear in adolescence, show differentiated prevalence, onset, and symptomatology across the biological sexes. Therefore, mapping differences in brain structure between males and females during this critical developmental period may provide information about the neural mechanisms underlying the dimorphism of these pathologies. Utilizing a large dataset collected through the Adolescent Brain Cognitive Development study, we investigated the differences of adolescent (9-10 years old) male and female brains (n = 8325) by using a linear Support-Vector Machine Classifier to predict sex based on morphometry and image intensity values of structural brain imaging data. The classifier correctly classified the sex of 86% individuals with the insula, the precentral and postcentral gyri, and the pericallosal sulcus as the most discernable features. The role of these significant dimorphic features in psychopathology was explored by testing them as mediators between sex and clinical symptomology. The results demonstrate the existence of morphometrical brain markers of sex difference. Significance StatementMany psychiatric pathologies express differently across the sexes. Therefore, an understanding of the differences in brain structure between males and females during the critical developmental period of adolescence may provide the insights about the dimorphism of clinical symptomology and the general functions of the dimorphic brain structures. Using machine learning, we successfully classified males and females with a high accuracy based on morphometry and image intensity data extracted from structural MRI scans. The features which significantly contributed to classification were examined to determine brain regions which are dimorphic during adolescence. The relevance of these brain regions to the expression of psychopathology symptoms was also explored.