Early psychosis structural abnormalities in the midbrain correlate with positive and negative symptoms




0000-0002-0604-0113 (Zhou, Zicong)

Journal Title

Journal ISSN

Volume Title



Early psychosis (EP) is a critical period in developing psychotic disorders. During the EP period, timely intervention strategies can effectively mitigate symptoms of psychotic disorders. A primary target in current therapeutics for psychotic disorders is the dopaminergic system; however, current therapeutics are often inadequate in treating the advanced stages of psychotic disorders. Currently, there are hypotheses that the advent of psychotic disorders are alterations in the brain’s structure and functional connectivity leading to aberrant network organization. Given the prominence of the midbrain in synthesizing dopamine and being a hub in the circuitry of dopaminergic function in this study, we will quantify the structural properties of the midbrain using a publicly available dataset of EP subjects. The Human Connectome Project (HCP) is a global effort to determine human brain connectivity objectively. Specifically, there is a subset where the focus is on subjects within five years of their first psychotic episode, which is the focus of this study. In this project, we employ various computational tools (including FSL, ANTs, packages of Python and RStudio, etc.) to acquire relevant measures of brain structure in EP. Specifically, we performed gross anatomical analysis of brain volumetrics, regional microstructural analysis, and correlation analysis between brain and behavior indices. First, Deformation-based Morphometry (DBM) is a test performed on T1 MRI scans to determine brain volumes. DBM measures shape movements to align individuals with a registration template using the Jacobian determinant (JD), whereas a standard VBM focuses on voxel intensity differences between individuals and the template. In our results, DBM revealed mild changes around the midbrain between EP and healthy controls. Second, we analyzed Diffusion Tensor Imaging (DTI) data by obtaining indices such as Fractional Anisotropy (FA), Axial Diffusivity (AD), and Mean Diffusivity (MD). A midbrain mask was created based on Freesurfer atlas ROI labels, allowing a seed-based analysis centered around the midbrain. Group differences were estimated using the Welch two-group t-tests on ROI means of JD, FA, AD, MD, and midbrain volumes. Interestingly, group differences in JD and midbrain volumes were insignificant, but differences were more pronounced for FA, AD, and MD. Third, we employed Tract-Based Spatial Statistics (TBSS) to determine microstructural changes in white matter tracts. TBSS successfully captured structural variabilities within the midbrain, aligning with our study’s expectations, in addition to moderate changes in other main white matter tracts, such as corticospinal tract and cingulum, suggesting an initiation of altered brain connectivity emanating from alterations in the midbrain or a putative reorganization of dopaminergic circuitry. Fourth, correlations between these quantities in the EP group and behavioral scores (i.e., PANSS and CAINS tests) were explored. It is found that midbrain volume noticeably correlates with the Cognitive score of PANSS and JD, strongly correlates with AD and MD; FA correlates with the Negative score of PANSS, and MD correlates with the Positive score of PANSS. Overall, these findings contribute to understanding midbrain involvement in early psychosis and underscore the interest for further investigation in this research path.