A Novel Continuous Sleep State Artificial Neural Network Model Based on Multi-Feature Fusion of Polysomnographic Data

Abstract

Purpose

Sleep structure is crucial in sleep research, characterized by its dynamic nature and temporal progression. Traditional 30-second epochs falter in capturing the intricate subtleties of various micro-sleep states. This paper introduces an innovative artificial neural network model to generate continuous sleep depth value (SDV), utilizing a novel multi-feature fusion approach with EEG data, seamlessly integrating temporal consistency.

Methods

The study involved 50 normal and 100 obstructive sleep apnea–hypopnea syndrome (OSAHS) participants. After segmenting the sleep data into 3-second intervals, a diverse array of 38 feature values were meticulously extracted, including power, spectrum entropy, frequency band duration and so on. The ensemble random forest model calculated the timing fitness value for all the features, from which the top 7 time-correlated features were selected to create detailed sleep sample values ranging from 0 to 1. Subsequently, an artificial neural network (ANN) model was trained to delineate sleep continuity details, unravel concealed patterns, and far surpassed the traditional 5-stage categorization (W, N1, N2, N3, and REM).

Results

The SDV changes from wakeful stage (mean 0.7021, standard deviation 0.2702) to stage N3 (mean 0.0396, standard deviation 0.0969). During the arousal epochs, the SDV increases from the range (0.1 to 0.3) to the range around 0.7, and decreases below 0.3. When in the deep sleep (≤0.1), the probability of arousal of normal individuals is less than 10%, while the average arousal probability of OSA patients is close to 30%.

Conclusion

A sleep continuity model is proposed based on multi-feature fusion, which generates SDV ranging from 0 to 1 (representing deep sleep to wakefulness). It can capture the nuances of the traditional five stages and subtle differences in microstates of sleep, considered as a complement or even an alternative to traditional sleep analysis.

Keywords:

• sleep depth value
• sleep continuity
• EEG features
• timing fitness
• ANN model

Acknowledgments

This study was supported the project of Shandong University Scientific Research Development Program (Grant No. J18KB071), Dongying Science and Technology Development Fund (Grant No. DJB2021010), Shandong Province Undergraduate Teaching Reform Research Project (Grant No. Z2021209 and Grant No. Z2023239) and Dongying Key Laboratory of Intelligent Information Processing and Sub-laboratory of Artificial Intelligence and Data Mining. Yunliang Sun and the Respiratory Department of Binzhou Medical University Hospital provided the data and gave a great help for paper writing and revision.

This paper has been facilitated and supported by the Biomedical Research Ethics Committee of Binzhou Medical University Affiliated Hospital: (1) Detailed explanations of the study were provided to participants, including the purpose, procedures, expected duration, potential risks and benefits, and any possible compensation. (2) Participants were ensured a full understanding of the information provided and had the opportunity to ask questions and discuss their concerns. (3) Voluntary participation was ensured, with participants knowing they could choose not to participate in the study or withdraw at any time without adverse consequences. (4) Completion of Patient Written Informed Consent. Prior to the commencement of the study, all participants were required to sign a written informed consent form. This form detailed all aspects of the study, including but not limited to the purpose, procedures, risks, benefits, and the methods of data collection and processing. (5) Participants were informed about how their personal information would be kept confidential and how the data would be securely stored and processed. (6) It was clearly explained to participants how their data would be used, including whether it would be shared with third parties and under what conditions. (7) Participants were informed of their right to receive updated information during the research process if there were any significant changes to the purpose, procedures, or risks of the study. (8) Communication with participants was maintained throughout the research process to ensure they continued to understand the progress and had the opportunity to ask questions or withdraw.

The research protocol of this study, including the methods of sleep data collection, has been reviewed and approved by the Biomedical Research Ethics Committee of the Affiliated Hospital of Binzhou Medical University. Throughout the research process, this study has continuously complied with the regulations and requirements set forth by the Ethics Review Committee, including the ethical principles outlined in the Declaration of Helsinki.

Disclosure

The authors report no conflicts of interest in this work.