Brain entropy as a biomarker of major depression in adolescents and young adults: insights from multimodal resting-state functional magentic resonance imaging

作者：Ruoxi Lu, Jie Li, Yiran Li, Xinglin Zeng, Yan Guo, Danian Li, Ying Cui, Xinyu Liang, Hanyue Zhang, Yi-Xiang Wang, Baohua Cheng, Yujie Liu, Ze Wang, Senhui Qiu · 发表于：Psychoradiology · 年份：2026 · DOI：10.1093/psyrad/kkag009 · 被引用次数：2 · 研究领域：Functional Brain Connectivity Studies、Mental Health Research Topics、EEG and Brain-Computer Interfaces

Background: Major depressive disorder (MDD) in adolescents and young adults is increasingly prevalent, yet accurate diagnosis remains challenging due to the limitations of conventional neuroimaging metrics. Traditional resting-state functional magnetic resonance imaging (rs-fMRI) measures such as amplitude of low-frequency fluctuations (ALFF), regional homogeneity (ReHo), and functional connectivity density (FCD) primarily capture static aspects of brain activity and may overlook critical neural dynamics. Brain entropy (BEN), which quantifies temporal irregularity in rs-fMRI signals, may offer a complementary approach to better characterize neural alterations in MDD. Methods: We analyzed multimodal rs-fMRI data from 204 individuals aged 12-24 years (119 with MDD and 85 healthy controls). BEN was computed alongside ALFF, ReHo, and FCD to extract region-wise features across the brain. A support vector machine with recursive feature elimination (SVM-RFE) was used to classify MDD and healthy controls based on various feature combinations. Classification performance was evaluated using repeated cross-validation and permutation testing. Additionally, partial Spearman correlations were performed between selected brain features and clinical measures including depression severity, childhood trauma, sleep quality, and cognitive control. Results: < 0.001). The most frequently selected brain regions contributing to MDD classification included the putamen, paracentral lobule, cuneus, midd...