HypnoGPT: A Hypnogram Language Model for Sleep Staging Enhancement and Sleep Disorder Diagnosis
Tianyou Yu,Zhenghui Gu,Fei Wang,Rui Huang,Man Li,Jin-gang Yu,Zhuliang Yu,Jun Zhang,Yan Xu,Haiteng Jiang,Wenjuan Liu,Guifeng Deng,Zhengrun Gao,Yiwen Wu,Jun Liu,Yu Zhang,Shumin Duan,Yuanqing Li,Jun Xiao,Wei Wu
DOI: https://doi.org/10.1101/2024.10.26.24316166
2024-01-01
Abstract:Sleep hypnograms are essential for understanding sleep architecture and diagnosing prevalent sleep disorders. Recognizing the strong correlations between successive sleep stages and the inherent sequential structure of sleep hypnograms, this study introduces HypnoGPT, the first GPT-based hypnogram language model. By modeling the sequential nature of sleep and capturing transition patterns within hypnograms, HypnoGPT extracts salient sleep information to improve both sleep staging and disorder diagnosis. Following self-supervised pretraining on manually annotated whole-night sleep hypnograms, HypnoGPT yielded significant performance gains in sleep staging and disorder diagnosis across 4 publicly available datasets, with successful blinded replications on 3 independent datasets. Notably, experiments on established sleep staging benchmarks validate HypnoGPT as a robust add-on module that consistently enhances the performance of existing sleep staging methods. Moreover, a HypnoGPT-powered transformer model achieved high sensitivity and specificity in classifying abnormal sleep stage sequences and diagnosing Type-1 narcolepsy without requiring the costly multiple-sleep latency test. Additionally, the model identified a range of sleep disorder-associated patterns embedded within sleep stage sequences. HypnoGPT offers a powerful approach to modeling sleep architecture and developing artificial intelligence (AI) models for downstream applications, opening new avenues to advancing precision medicine for sleep disorders. ### Competing Interest Statement W.W. reports equity from Alto Neuroscience. None of the other authors has financial disclosures to report. ### Funding Statement This work was supported in part by STI2030-Major Projects under Grant 2022ZD0211700, the National Natural Science Foundation of China under Grant 62376098, 62276102, and U22A20293, and GuangDong Basic and Applied Basic Research Foundation 2024A1515011983. ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: The study on the private HANG7 dataset was conducted at Zhejiang University with Institutional Review Board approval, and written consent was obtained from all participants or their caregivers. I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes The SHHS and MNC datasets are provided by the National Sleep Research Resource with appropriate deidentification. Permission and access for these datasets can be obtained via the online portal: https://www.sleepdata.org. The SleepEDF, Physio2018, and CAP datasets are available from PhysioNet at https://physionet.org/content/sleep-edfx/1.0.0/, https://physionet.org/content/challenge-2018/1.0.0/, and https://physionet.org/content/capslpdb/1.0.0/, respectively. The MASS dataset is available from http://ceams-carsm.ca/mass/. The ISRUC dataset can be accessed from https://sleeptight.isr.uc.pt/. Access to the HANG7 dataset is governed by data-use agreements, and it is therefore not publicly available.