Biological correlates of spontaneous brain activity alterations in diabetic retinopathy: an amplitude of low-frequency fluctuation-based multimodal study.
Authors
Affiliations (8)
Affiliations (8)
- The Affiliated Eye Hospital, Jiangxi Medical College, Nanchang University.
- Jiangxi Province Key Laboratory of Ophthalmology and Vision Sciences.
- Jiangxi Clinical Research Center for Ophthalmic Disease.
- Jiangxi Provincial Key Laboratory of Vitreoretinal Diseases for Health.
- School of Ophthalmology and Optometry.
- Die Hu, Qi-Ran Zhang, and Yuan-Zhi He contributed equally to the writing of this article.
- The Second Clinical Medical College and.
- Queen Mary School, Jiangxi Medical College, Nanchang University, Nanchang, Jiangxi Province, China.
Abstract
This study aimed to investigate diabetic retinopathy-related alterations in spontaneous brain activity using ALFF and to explore whether these alterations show indirect spatial associations with normative molecular, cellular, and neurochemical brain architectures. Resting-state functional MRI data were acquired from 46 patients with diabetic retinopathy and 44 healthy controls. The amplitude of low-frequency fluctuation (ALFF) was used to assess regional spontaneous brain activity. ALFF alteration maps were further integrated with cortical gene expression data to perform imaging transcriptomic analysis, functional enrichment, cell-type-specific expression analysis, neurotransmitter map association analysis, and machine learning classification. Compared with healthy controls, patients with diabetic retinopathy showed increased ALFF in the cerebellum VIII region, left inferior temporal gyrus, right hippocampus, and left pallidum, and decreased ALFF in the bilateral middle occipital gyri and left calcarine cortex. Imaging transcriptomic analysis linked ALFF alterations to gene expression patterns enriched in neural development, synaptic structure, cytoskeletal organization, cell adhesion, and immune-inflammatory pathways. Cell-type analysis implicated astrocytes, microglia, oligodendrocyte precursor cells, and oligodendrocytes. Neurotransmitter mapping suggested associations with multiple neurochemical systems. Machine learning models based on ALFF features demonstrated moderate-to-good classification performance, with logistic regression achieving the highest mean area under the receiver operating characteristic curve of 0.854 ± 0.028. Patients with diabetic retinopathy showed altered spontaneous brain activity in visual and extra-visual regions. Multimodal analyses provided exploratory biological annotation and preliminary discriminative information for these ALFF alterations, but the transcriptomic and neurotransmitter findings should be interpreted as indirect atlas-based associations requiring further validation.