Identifying Depressive Disorder With Sleep Electroencephalogram Data: A Study Based on Deep Learning

TAO Ran , DING Sheng-nan, CHEN Jie, ZHU Xue-min, NI Zhao-jun, HU Ling-ming, ZHANG Yang, XU Yan, SUN Hong-qiang

Abstract

To explore the effectiveness of using deep learning network combined Vision Transformer (ViT) and Transformer to identify patients with depressive disorder on the basis of their sleep electroencephalogram (EEG) signals.   Methods  The sleep EEG signals of 28 patients with depressive disorder and 37 normal controls were preprocessed. Then, the signals were converted into image format and the feature information on frequency domain and spatial domain was retained. After that, the images were transmitted to the ViT-Transformer coding network for deep learning of the EEG signal characteristics of the rapid eye movement (REM) sleep and non-rapid eye movement (NREM) sleep in patients with depressive disorder and those in normal controls, respectively, and to identify patients with depressive disorder.   Results  Based on the ViT-Transformer network, after examining different EEG frequencies, we found that the combination of delta, theta, and beta waves produced better results in identifying depressive disorder. Among the different EEG frequencies, EEG signal features of delta-theta-beta combination waves in REM sleep achieved 92.8% accuracy and 93.8% precision for identifying depression, with the recall rate of patients with depression being 84.7%, and the F0.5 value being 0.917±0.074. When using the delta-theta-beta combination EEG signal features in NREM sleep to identify depressive disorder, the accuracy was 91.7%, the precision was 90.8%, the recall rate was 85.2%, and the F0.5 value was 0.914±0.062. In addition, through visualization of the sleep EEG of different sleep stages for the whole night, it was found that classification errors usually occurred during transition to a different sleep stage.   Conclusion  Using the deep learning ViT-Transformer network, we found that the EEG signal features in REM sleep based on delta-theta-beta combination waves showed better effect in identifying depressive disorder.

 

Keywords: Depressive disorder, Sleep electroencephalogram, Deep learning, Rapid eye movement sleep, Non-rapid eye movement sleep


Full Text:

PDF


References


Organization WHO. Depression and other common mental disorders: global health estimates: World Health Organization， 2017. (2017-02)[2022-12-13].https://apps.who.int/iris/handle/10665/254610.

RIEMANN D， KRONE L B， WULFF K， et al. Sleep, insomnia, and depression. Neuropsychopharmacology，2020，45(1): 74–89. doi: 10.1038/s41386-019-0411-y.

NOLLET M， HICKS H， MCCARTHY A P， et al. REM sleep's unique associations with corticosterone regulation, apoptotic pathways, and behavior in chronic stress in mice. Proc Natl Acad Sci U S A，2019， 116(7): 2733–2742. doi: 10.1073/pnas. 1816456116.

COAN A， ALLEN J. Frontal EEG asymmetry as a moderator and mediator of emotion. Biol Psychol，2004，67(1/2): 7–49. doi: 10.1016/j. biopsycho.2004.03.002.

ZHAO Y， ZHANG G， ZHANG Y， et al. Multi-view cross-subject seizure detection with information bottleneck attribution. J Neural Eng，2022， 19(4): 046011. doi: 10.1088/1741-2552/ac7d0d.

TSIAMALOU A， DARDIOTIS E， PATERAKIS K， et al. EEG in neurorehabilitation: a bibliometric analysis and content review. Neurol Int，2022，14(4): 1046–1061. doi: 10.3390/neurolint14040084.

AY B， YILDIRIM O， TALO M， et al. Automated depression detection using deep representation and sequence learning with EEG signals. J Med Syst，2019，43(7): 205. doi: 10.1007/s10916-019-1345-y.

ULKE C， TENKE C E， KAYSER J， et al. Resting EEG Measures of Brain Arousal in a Multisite Study of Major Depression. Clin EEG Neurosci， 2019，50(1): 3–12. doi: 10.1177/1550059418795578.

ACHARYA U R， OH S L， HAGIWARA Y， et al. Automated EEG-based screening of depression using deep convolutional neural network. Comput Methods Programs Biomed，2018，161: 103–113. doi: 10.1016/j. cmpb.2018.04.012.

THAKRE T P， KULKARNI H， ADAMS K S， et al. Polysomnographic identification of anxiety and depression using deep learning. J Psychiatr Res，2022，150: 54–63. doi: 10.1016/j. jpsychires.2022.03.027.

BERRY R B， BROOKS R， GAMALDO C， et al. AASM scoring manual updates for 2017 (Version 2.4). J Clin Sleep Med，2017，13(5): 665–666. doi: 10.5664/jcsm.6576.

SANTANGELI O， PORKKA T， VIRKKALA J， et al. Sleep and slow-wave activity in depressed adolescent boys: a preliminary study. Sleep Med，2017，38: 24–30. doi: 10.1016/j. sleep.2017.06.029.

BLASKOVICH B， REICHARDT R， GOMBOS F， et al. Cortical hyperarousal in NREM sleep normalizes from pre- to post- REM periods in individuals with frequent nightmares. Sleep，2020，43(1): zsz201. doi: 10.1093/sleep/zsz201.

NOFZINGER E A， PRICE J C， MELTZER C， et al. Towards a neurobiology of dysfunctional arousal in depression: the relationship between beta EEG power and regional cerebral glucose metabolism during NREM sleep. Psychiatry Res，2000，98(2): 71–91. doi: 10.1016/s0925-4927(00)00045-7.

ALHAJ H， WISNIEWSKI G， MCALLISTER R H. The use of the EEG in measuring therapeutic drug action: focus on depression and antidepressants. J Psychopharmacol，2011，25(9): 1175–1191. doi: 10. 1177/0269881110388323.

LEUCHTER A F， COOK I A， HUNTER A， et al. Use of clinical neurophysiology for the selection of medication in the treatment of major depressive disorder: the state of the evidence. Clin EEG Neurosci，2009， 40(2): 78–83. doi: 10.1177/155005940904000207.

SUN S， LI X， ZHU J， et al. Graph theory analysis of functional connectivity in major depression disorder with high-density resting state EEG data. IEEE Trans Neural Syst Rehabil Eng，2019，27(3): 429–439. doi: 10.1109/TNSRE.2019.2894423.

Van STRAATEN E C， DENHAAN J， DEWAAL H， et al. Disturbed phase relations in white matter hyperintensity based vascular dementia: an EEG directed connectivity study. Clin Neurophysiol，2015，126(3): 497–504. doi: 10.1016/j.clinph.2014.05.018.

CHIANG H S， CHEN M Y， LIAO L S. Cognitive depression detection cyber-medical system based on EEG analysis and deep learning approaches. IEEE J Biomed Health Inform，2023，27(2): 608–616. doi: 10. 1109/JBHI.2022.3200522.

ADAMCZYK M， GAZEA M， WOLLWEBER B， et al. Cordance derived from REM sleep EEG as a biomarker for treatment response in depression--a naturalistic study after antidepressant medication. J Psychiatr Res，2015，63: 97–104. doi: 10.1016/j. jpsychires.2015.02.007.

MIKOTEIT T， BRAND S， PERREN S， et al. Visually detected non-rapid eye movement stage 2 sleep spindle density at age five years predicted prosocial behavior positively and hyperactivity scores negatively at age nine years. Sleep Med，2018，48: 101–106. doi: 10.1016/j. sleep.2018.03. 028.

BYEON H. Application of machine learning technique to distinguish Parkinson's disease dementia and Alzheimer's dementia: predictive power of parkinson's disease-related non-motor symptoms and neuropsychological profile. J Pers Med，2020，10(2): 31. doi: 10.3390/jpm10020031.

DOSOVITSKIY A， BEYER L， KOLESNIKOV A， et al. An image is worth 16x16 words: transformers for image recognition at scale. arXiv:2010.11929. https://arxiv.org/abs/2010.11929v2. doi: 10.48550/arXiv.2010.11929.


Refbacks

  • There are currently no refbacks.