Related Experiment Video
Updated: Sep 23, 2026
Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data
Published on: June 26, 2013
Altered Kullback-Leibler similarity gradients in cocaine use disorder: Connectome constraints and multiscale
Rui Wang1, Dan Luo2, Jiajun Xu2
1Department of Radiology, Institution of Radiology and Medical Imaging, West China Hospital of Sichuan University, Chengdu, 610041, Sichuan, China.
Background:
Cortical morphological alterations are evident in cocaine use disorder (CUD), yet their organization within macroscale cortical hierarchies, relationship to structural connectome architecture, and associations with multiscale biological reference maps remain incompletely understood.
Methods:
Using structural MRI from 71 CUD patients and 60 healthy controls, we constructed individual-level morphological networks and calculated gradients. A connectome constraint model was used to assess how structural connectome architecture relates to altered gradient patterns, while a network diffusion model identified putative epicenters. Multiscale reference maps, transcriptomic data, and mitochondrial atlases were used for exploratory spatial contextualization of the neurochemical, genetic, and molecular features associated with these alterations.
Results:
Group comparisons demonstrated decreased gradient scores in lower-order sensorimotor systems and increased scores in higher-order associative networks. The spatial patterning of these gradient alterations was significantly constrained by the structural connectome. We identified the bilateral precentral and postcentral gyri as putative epicenters associated with the spatial distribution of the observed gradient alteration patterns. Multiscale annotations identified spatial associations between these alterations and cortical microstructure, neurotransmitter receptor profiles, and mitochondrial features. Transcriptomic contextualization showed that gradient alterations in CUD were related to genes involved in synaptic structure and organelle inner membranes, while psychiatric enrichment indicated associations with alcohol abuse disorder and schizophrenia-related gene sets.
Conclusion:
These findings characterize the structural connectome constraints and multiscale biological associations of macroscale gradient alterations in CUD, providing a hypothesis-generating framework for contextualizing their network-level organization.
