Understanding Cognitive Impairment in Early Psychosis through Functional Brain Dysconnectivity: A Whole-Brain Voxel-wise Analysis Approach

Abstract

Early psychosis (EP) is the early stage of onset of psychosis symptoms, characterized by a loss of touch with reality. Cognitive impairment is prominent and even precedes symptom onset in individuals with psychosis, which is commonly driven by alterations in functional brain connectivity. Previous work has focused on brain regions defined a priori and relied on the assumption of normal distributions, which can hinder result generalizability. This project aims to build upon previous findings in a data-driven way and share code following open science principles. We utilize the Human Connectome Project-Early Psychosis dataset of 183 participants, consisting of resting-state functional magnetic resonance imaging (rs-fMRI) scans and phenotypic data from a variety of cognitive domains, such as attention, memory, and processing speed. Functional connectivity is calculated in a voxel-wise manner to derive voxel-whole brain connectivity patterns. Multivariate Distance Matrix Regression (MDMR), a non-parametric technique, is then applied to assess relationships between functional connectivity differences and cognitive scores. Statistics for hypothesis testing follow an asymptotic null distribution, and theoretical p-values can be calculated such that the results are more robust. Further, scripts detailing the data analysis will be shared on GitHub, linked to an Open Science Framework project to aid replicability. We expect to identify some well-known brain areas implicating specific impairments, which will be externally validated in a follow-up study. This may provide a more holistic view on brain-cognition relationships for early psychosis and guide future applications in related fields of research.