Comprehensive study of sex-based anatomical variations of human brain and development of sex-specific brain templates

Abstract

The sex-based human brain structural variations alongside the necessity and development process for sex-specific brain templates were investigated in this study. Comparing magnetic resonance images of 500 female and 500 male subjects, no significant sex-based difference was observed for average cortical thickness, however, all the volumetric values, including the total brain volume (TBV) and major 19 brain regions, were found to be significantly different between females and males. Moreover, analyzing the fractional volume of the regions showed that these sex variations were not proportional to TBV for all regions. These findings underscore the importance of distinguishing the sex-based differences in human brain studies. While brain templates have been developed for general population and cohorts with the same characteristics such as race or age, there is a lack of sex-specific brain templates. To fill this gap and find a representative reference brain image for each sex, nonlinear templates were developed for female, male, and mixed population subjects. Next, a separate set of 109 female and 109 male brain images were used to evaluate the sex-specificity of the brain templates. It was observed that the female and male test subjects were registered to their sex-specific templates with the lowest amount of deformation/warping confirming better representativeness of the sex-specific templates for their target population. The findings of this study including the templates and the reported variations can be used in research involving sex dimorphic brain disorders, diseases, and/or injuries such as traumatic brain injury that is affected by the sex-based brain anatomical differences. Statement of significanceHuman brain exhibits sex-based variation both in size and volumetric composition of different regions. Despite these differences, there is a paucity of sex-specific brain templates. Addressing this gap marks the significance of our study as briefly explained here. We have shown that differences in male and female brain go beyond simple scaling and the observation of regional differences that are not proportional to the sex-based total brain volume variations has motivated us to develop sex-specific templates. The representativeness and difference of these sex-specific templates were assessed by measuring the amount of required warping in nonlinear registration of test subjects to them. It was shown that registration of female and male subjects to their corresponding sex-specific template involved lower level of warping compared to their registration to their opposite sex or mixed population brain template.