Global and network-level topographic mapping of functional connectivity networks in schizophrenia and bipolar disorder using extended-duration resting-state fMRI

Background: Functional connectivity MRI studies have identified widespread dysconnectivity in schizophrenia and bipolar disorder. However, most approaches rely on group-defined atlases that assume fixed network boundaries, potentially obscuring effects due to inter-individual variability in network organization. Here, we examined topographic abnormalities using individualized functional mapping. Methods: Resting-state fMRI data (1 h acquisition) were obtained from 56 healthy controls (HC), 45 bipolar disorder (BP), and 31 schizophrenia (SZ) participants (ages 18–33). Individualized functional networks were derived using template matching. Topographic Abnormality Index (TAI) quantified network-specific spatial deviations relative to normative boundaries, while Vertexwise Functional Deviation Index (VFDI) reflected global deviation across all cortical vertices. Group differences were assessed using ANOVA and ANCOVA (covarying age, sex, and motion), with false discovery rate correction. Clinical correlations were examined within BP and SZ. Results: Significant group effects were observed across multiple networks, including temporo-insular (TIN), cingulo-opercular (CON), sensorimotor, dorsal attention (DAN), language (LAN), and default mode (DMN) networks (q < 0.05). BP showed prominent abnormalities in sensorimotor and perceptual networks, whereas SZ exhibited greater involvement of higher-order associative networks. Global metrics demonstrated robust group discrimination: TAI average (F = 9.01, p = 2 × 10⁻⁴) and VFDI (F = 11.06, p = 3.7 × 10⁻⁵), with both BP and SZ elevated relative to HC. Mania severity correlated with sensorimotor (body) network TAI (q = 0.002). Global metrics were not related to symptom severity. Conclusions: Schizophrenia and bipolar disorder are characterized by widespread but distinct disruptions in functional network topography. Global measures, particularly VFDI, provide sensitive indices of cross-network abnormality and may offer utility for biomarker development beyond traditional network-specific approaches.