College of Mathematics and Informatics, South China Agricultural University, Guangzhou, 510642, China.
College of Information Engineering, Shanghai Maritime University, Shanghai, 201306, China.
J Mol Neurosci. 2022 Oct;72(10):2094-2105. doi: 10.1007/s12031-022-02061-3. Epub 2022 Aug 25.
Depression is characterized by poor emotion regulation that makes it difficult to escape the effects of emotional pain, but the neuromodulation behind these symptoms is still unclear. This study investigated the neural mechanism of emotional state-related responses during music stimuli in participants with major depressive disorder (MDD) compared to never-depressed (ND) controls. A novel two-level feature selection method, integrating recursive feature elimination based on support vector machine (SVM-RFE) and random forest algorithm (RF), was proposed to screen emotional recognition brain regions (ERBRs). On this basis, the differences of functional connectivity (FC) were systematically analyzed by two-sample t-test. The results demonstrate that ND participants show eight pairs of FCs with a significant difference between positive emotional music stimuli (pEMS) versus negative emotional music stimuli (nEMS) in 15 ERBRs of MDD, but the participants with MDD show one pair of significant difference in FC. The decreased number reflects the fuzzy response to positive and negative emotions in MDD, which appears to arise from obstacle to emotional cognition and regulation. Furthermore, there was no significant difference in FC between MDDs and NDs under pEMS, but a significant difference was detected between the two groups under nEMS (p < 0.01), revealing a 'bias' against the negative state in MDD. The current study may help to better comprehend the abnormal evolution from normal to depression and inform the utilization of pEMS in formal treatment for depression.
抑郁症的特征是情绪调节能力差，难以摆脱情绪痛苦的影响，但这些症状背后的神经调节仍不清楚。本研究调查了与从未抑郁（ND）对照组相比，重度抑郁症（MDD）患者在音乐刺激期间情绪状态相关反应的神经机制。提出了一种新的两级特征选择方法，将基于支持向量机（SVM）的递归特征消除（SVM-RFE）和随机森林算法（RF）相结合，用于筛选情绪识别脑区（ERBRs）。在此基础上，通过两样本 t 检验系统地分析了功能连接（FC）的差异。结果表明，ND 参与者在 MDD 的 15 个 ERBRs 中，pEMS 与 nEMS 之间存在 8 对 FC 差异具有统计学意义，而 MDD 参与者的 FC 存在 1 对差异具有统计学意义。减少的数量反映了 MDD 对正性和负性情绪的模糊反应，这似乎源于情绪认知和调节的障碍。此外，在 pEMS 下，MDD 和 ND 之间的 FC 没有显著差异，但在 nEMS 下两组之间存在显著差异（p<0.01），这揭示了 MDD 对负性状态的“偏向”。本研究有助于更好地理解从正常到抑郁的异常演变，并为 pEMS 在抑郁症正式治疗中的应用提供信息。
