Abstract：Major depressive disorder (MDD) is a mental disorder caused by a variety of factors, such as congenital genetic abnormalities, acquired environmental mutations. MDD can cause serious damage to patients’ daily life and social economy. Therefore, seeking more effective and objective physiological indicators and diagnostic methods to assist diagnosis has great significance for the early diagnosis and prevention of MDD. Based on the emotional face-word stroop task, this paper used traditional machine learning, ensemble learning and deep learning methods to study 31 MDD patients and 31 healthy controls by collecting behavioral data and ERP data of subjects. The evaluation indicators for different classification methods include accuracy (ACC), F1-score (F1), recall (Recall), specificity (Specificity), positive predictive value (PPV), and negative predictive value (NPV). The data set was randomly divided into training set, verification set and test set in a ratio of 7:2:1. The process was repeated ten times, and the final classification result was the mean ± standard deviation of ten times. The results showed that the accuracy of convolutional neural network (CNN) method, which can automatically learn and extract features from the data, achieved 89.76% ± 19.18% in the identification of MDD based on behavioral data. Based on ERP data, it was found that CNN obtained the optimal result under all six indicators, and the accuracy of MDD recognition was 90.71% ± 14.17%. This paper proposed a multi-modal deep learning neural network based on behavioral data and ERP data, which is referred as behavior-ERP parallel temporal convolution neural network (BEPTCNN). It achieved excellent results in all indicators of MDD identification task, and the recognition accuracy could reach 95.48% ± 7.31%. These results showed that both behavioral data and ERP data could be used as effective physiological indicators for the auxiliary diagnosis of MDD. In addition, the BEPTCNN model proposed in this paper could be used as an effective method for the recognition of MDD.
侯峰, 张明, 蔺向彬, 张威, 马榕. 基于行为与事件相关电位的机器学习重度抑郁识别研究[J]. 中国生物医学工程学报, 2023, 42(5): 542-553.
Hou Feng, Zhang Ming, Lin Xiangbin, Zhang Wei, Ma Rong. Recognition of Major Depression Using Machine Learning Methods Based on Behavioral and Event-Related Potentials. Chinese Journal of Biomedical Engineering, 2023, 42(5): 542-553.
[1] Friedrich MJ. Depression is the leading cause of disability around the world[J]. JAMA, 2017, 317(15):1517-1517. [2] Moussavi S, Chatterji S, Verdes E, et al. Depression, chronic diseases, and decrements in health: results from the World Health Surveys[J]. Lancet,2007, 370(9590):851-858. [3] Tran T, Milanovic M, Holshausen K, et al. What is normal cognition in depression? Prevalence and functional correlates of normative versus idiographic cognitive impairment[J]. Neuro-psychology, 2021, 35(1):33-41. [4] Cai L, Bao Y, X Fu, et al. Causal links between major depressive disorder and insomnia: A Mendelian randomisation study[J]. Gene, 2020, 768:145271. [5] Gananca L, Galfalvy HC, Cisneros-Trujillo S, et al. Relation-ships between inflammatory markers and suicide risk status in major depression[J]. Journal of Psychiatric Research, 2020, 134: 192-199. [6] Tak S, Lee S, Park CA, et al. Altered Effective Connectivity within the fronto-limbic circuitry in response to negative emotional task in female patients with major depressive disorder[J]. Brain Connect, 2021,11(4):264-277. [7] Vos T, Flaxman AD, Naghavi M, et al. Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010[J]. Lancet. 2012,380(9859):2163-2196. [8] CJ Funkhouser, Kaiser A, Alqueza K L, et al. Depression risk factors and affect dynamics: an experience sampling study[J]. Journal of Psychiatric Research, 2021, 135:68-75. [9] Saunders EFH, Mukherjee D, Waschbusch DA, et al. Predictors of diagnostic delay: assessment of psychiatric disorders in the clinic[J]. Depression and Anxiety, 2020,38:562-570. [10] Hasanzadeh F, Mohebbi M, Rostami R. Single channel EEG classification: a case study on prediction of major depressive disorder treatment outcome[J]. IEEE Access,2021,9: 3417-3427. [11] Li X, Li J, Hu B, et al. Attentional bias in MDD: ERP components analysis and classification using a dot-probe task[J]. Computer Methods and Programs in Biomedicine, 2018, 164:169-179. [12] 陆林,沈渔邨. 精神病学:抑郁障碍. 第二版[M]. 北京:人民卫生出版社, 1988:391-392. [13] Kraepelin E. Manic-depressive insanity and paranoia (classic reprint)[J]. Journal of Nervous & Mental Disease, 1921, 53(4):350-350. [14] Zhu Y, Shang Y, Shao Z, et al. Automated depression diagnosis based on deep networks to encode facial appearance and dynamics[J]. IEEE Transactions on Affective Computing, 2018, 9(4): 578-584. [15] 刘岩,李幼军,陈萌. 基于固有模态分解和深度学习的抑郁症脑电信号分类分析[J]. 中国医学物理学杂志,2017,34(9):963-967. [16] Funahashi S, Andreau JM. Prefrontal cortex and neural mechanisms of executive function - ScienceDirect[J]. Journal of Physiology Paris, 2013, 107(6):471-482. [17] Wang, L, LaBar KS. Prefrontal mechanisms for executive control over emotional distraction are altered in major de-pression[J]. Psychiatry Res, 2008, 163(2): 143-155. [18] Fenske MJ, Eastwood JD. Modulation of focused attention by faces expressing emotion: evidence from flanker tasks[J]. Emotion, 2003, 3(4): 327-343. [19] Ruz M, Tudela P. Emotional conflict in interpersonal interactions. Neuroimage.2011;54(2):1685-1691. [20] 刘吉梅.人机互动过程中情绪冲突的研究[D].重庆:西南大学, 2013. [21] Cservenka A , Stroup ML , Etkin A , et al. The effects of age, sex, and hormones on emotional conflict-related brain response during adolescence[J]. Brain &Cognition, 2015, 99(10):135-150. [22] Demšar J. Statistical comparisons of classifiers over multiple data sets[J]. The Journal of Machine Learning Research, 2006, 7:1-30. [23] Alhanai T, Ghassemi M, Glass J. Detecting depression with audio/text sequence modeling of interviews[C]//2018 Proceedings of the 19th Annual Conference of the International Speech Communication Association. Hyderabad: INTERSPEECH, 2018:1716-1720. [24] Li, Xiaowei, Tong Cao, Shuting Sun, et al. Classification study on eye movement data: Towards a new approach in depression detection[C]// 2016 IEEE Congress on Evolutionary Computation (CEC). Vancouver: IEEE , 2016: 1227-1232. [25] Qureshi S A, Saha S, Hasanuzzaman M, et al. Multi-task representation learning for multimodal estimation of depression level[J]. Intelligent Systems, IEEE, 2019, 34(5):45-52. [26] 章文佩,沈群伦,宋锦涛,等. 基于事件相关电位 (ERPs) 和机器学习的考试焦虑诊断[J]. 心理学报, 2019, 51(10): 1116-1127. [27] Zhao Z, Wang C, Y uan Q, et al. Dynamic changes of brain networks during feedback-related processing of reinforcement learning in schizophrenia[J]. Brain Research, 2020, 1746: 146979. [28] O’Halloran L, Rueda-Delgado LM, Jollans L, et al. Inhibitory-control event-related potentials correlate with individual differences in alcohol use[J]. Addiction Biology, 2020,25(2): e12729. [29] Mirowski PW, Lecun Y, Madhavan D, et al. Comparing SVM and convolutional networks for epileptic seizure predic-tion from intracranial EEG[C]//2008 IEEE Workshop on Machine Learning for Signal Processing. Cancun: IEEE , 2008: 244-249. [30] Gilanie G, Bajwa UI, Waraich MM, et al. Coronavirus (COVID-19) detection from chest radiology images using con-volutional neural networks[J]. Biomedical Signal Processing and Control, 2021,66: 102490. [31] Bai SJ, Kolter JZ , Koltun V, et al. An empirical evaluation of generic convolutional and recurrent networks for sequence modeling[J/OL]. https://doi.org/10.48550/arXiv.1803.01271, 2018-03-04/2021-11-18. [32] Amin SU, Alsulaiman M, Muhammad G, et al. Deep learning for EEG motor imagery classification based on multi-layer CNNs feature fusion[J]. Future Generation Computer Systems, 2019, 101:542-554. [33] 曾柏泉. 基于卷积神经网络的情感脑电识别研究[D]. 广州:华南理工大学, 2019. [34] Xie J, Wang Q. Benchmarking machine learning algorithms on blood glucose prediction for type i diabetes in comparison with classical time-series models[J]. IEEE Transactions on Biomedical Engineering, 2020, 67(11): 3101-3124. [35] Hewage P, Behera A, Trovati M, et al. Temporal convolutional neural(TCN) network for an effective weather forecasting using time-series data from the local weather station[J]. Soft Computing, 2020, 24(21):16453-16482.
