Fusion analysis of sMRI-DTI in patients with ulcerative colitis by multimodal CCA + joint ICA and associated neurotransmitter profiles.

Abstract

Background: As a chronic inflammatory disease, ulcerative colitis (UC) is characterized by complex etiology and unclear pathological mechanisms. Neuroimaging research plays a pivotal role in improving diagnostic accuracy and guiding treatment decisions, while enhancing our understanding of disease mechanisms. We aimed to explore the possible associations between gray matter volume (GMV) and white matter fractional anisotropy (FA) abnormalities as well as the neurotransmitter profiles in UC patients.

Methods: Thirty-four UC patients and 21 healthy controls (HCs) participated in structural magnetic resonance imaging and diffusion tensor imaging scans. To identify the joint and specific independent components (ICs) across modalities between groups, multimodal canonical correlation analysis with joint independent component analysis was used for fusion analysis. ICs with significant group differences from the same index across two modalities were considered joint group-discriminative ICs. Joint ICs reveal cross-modal neurophysiological mechanisms of GMV-FA. A modal-specific group-discriminative IC is defined as a component that shows significant group difference only in unimodal, revealing specific neurophysiological mechanisms of GMV/FA. The relationship between neuroimaging findings and clinical characteristics was assessed. We also investigated the spatial correlations between the joint and modality-specific ICs and neurotransmitter profiles.

Results: Compared to HCs, patients with UC showed one joint group-discriminating component (GMV_IC4-FA_IC4) mainly in the default mode network, thalamus, corpus callosum, corona radiata, fornix, posterior thalamic radiation, middle cerebellar peduncle, and corticospinal tract as well as one modality-specific group-discriminating component (FA_IC5). The loadings of GMV_IC5 were significantly negatively correlated with platelet levels in UC. Moreover, significant associations between the abnormalities of GMV_IC4 and the dopamine and gamma-aminobutric acid (GABAa) system, between the abnormalities of FA_IC4 and the dopamine, GABAa, acetylcholine, and glutamate system, and between the abnormalities of FA_IC5 and dopamine and serotonin systems were found in this study.

Conclusions: This study suggested the complex interplay between structural alterations and associated neurotransmitter changes in neural systems subserving emotion dysregulation and visceral sensory processing in UC patients. The identified covarying GMV-FA complements previous findings of structural abnormalities. Furthermore, these findings provided novel insights into the neuropathological mechanisms and potential therapeutic targets of UC.