Utilizing Mutual Information Analysis to Explore the Relationship Between Gray and White Matter Structural Pathologies in Schizophrenia

Lyall AE, Savadjiev P, Del Re EC, Seitz J, O’Donnell LJ, Westin C-F, Mesholam-Gately RI, Petryshen T, Wojcik JD, Nestor P, et al. Utilizing Mutual Information Analysis to Explore the Relationship Between Gray and White Matter Structural Pathologies in Schizophrenia. Schizophr Bull. 2019;45(2):386–395.

Abstract

Schizophrenia has been characterized as a neurodevelopmental disorder, with structural brain abnormalities reported at all stages. However, at present, it remains unclear whether gray and white matter abnormalities represent related or independent pathologies in schizophrenia. In this study, we present findings from an integrative analysis exploring the morphological relationship between gray and white matter in 45 schizophrenia participants and 49 healthy controls. We utilized mutual information (MI), a measure of how much information two variables share, to assess the morphological dependence between gray and white matter in three segments of the corpus callsoum, and the gray matter regions these segments connect: (1) the genu and the left and right rostral middle frontal gyrus (rMFG), (2) the isthmus and the left and right superior temporal gyrus (STG), (3) the splenium and the left and right lateral occipital gyrus (LOG). We report significantly reduced MI between white matter tract dispersion of the right hemispheric callosal connections to the STG and both cortical thickness and area in the right STG in schizophrenia patients, despite a lack of group differences in cortical thickness, surface area, or dispersion. We believe that this reduction in morphological dependence between gray and white matter may reflect a possible decoupling of the developmental processes that shape morphological features of white and gray matter early in life. The present study also demonstrates the importance of studying the relationship between gray and white matter measures, as opposed to restricting analyses to gray and white matter measures independently.
Last updated on 02/26/2023