# Functional network comparative area and topography analysis (FUNCATA) in non-affective psychosis: a replication study

## TL;DR

This study examines brain network differences in early psychosis using individualized fMRI analysis, identifying potential biomarkers and distinct subtypes of psychosis.

## Contribution

The study introduces FUNCATA, a novel method for analyzing individual variability in functional brain networks in non-affective psychosis.

## Key findings

- NAP participants showed significantly altered network sizes in dorsal attention, default mode, and sensorimotor-body networks.
- Altered DAN and DMN organization may serve as early biomarkers for psychosis detection and intervention.
- Three distinct psychosis biotypes were identified based on network profiles and symptom patterns.

## Abstract

Resting-state fMRI studies consistently demonstrate widespread dysconnectivity in schizophrenia (SZ), yet conventional analytic methods often fail to account for individual variability in functional brain organization. This study utilized individualized assessments of network size and topography to examine functional alterations in early psychosis. MRI data were drawn from the Human Connectome Project – Early Psychosis study (ages 18–34), including 86 individuals with non-affective psychosis (NAP) and 57 healthy controls (HC). Ten large-scale functional networks were delineated using template-matching procedures. Group differences in network size were evaluated using ANOVA, while network topography was examined with vertex-wise chi-square analyses and the Topographic Abnormality Index (TAI). Compared to controls, NAP participants showed significantly larger dorsal attention (DAN) and default mode (DMN) networks, along with a smaller sensorimotor-body (SBN) network (effect sizes d = 0.39–0.48). NAP also exhibited greater topographic abnormalities in the DAN, DMN, and cingulo-opercular (CON) networks. DMN size was inversely related to mania symptoms, antipsychotic treatment duration, and working memory performance, while smaller SBN size was also linked to reduced working memory. A k-means clustering revealed three psychosis biotypes. Biotype 1 had enlarged DAN and language network size, with higher antipsychotic exposure. Biotype 2 showed near-normal network profiles but elevated mood symptoms. Biotype 3 exhibited enlarged DMN/DAN and reduced frontoparietal network size, with prominent negative symptoms. Consistent with prior schizophrenia studies, DAN enlargement was present in early psychosis, suggesting stability across illness stages. Altered DAN and DMN organization may serve as early biomarkers to guide detection and intervention strategies.