Stanford researchers propose SleepFM: a new multimodal basic model for sleep analysis

Sleep medicine is a fundamental field that involves monitoring and evaluating physiological signals to diagnose sleep disorders and understand sleep patterns. Techniques such as polysomnography (PSG) record brain, heart, and respiratory activities during sleep, providing a detailed overview of a person's sleep health. These signals are essential for categorizing sleep stages and identifying sleep disorders. PSG typically includes electroencephalograms (EEG), electrooculograms (EOG), electromyograms (EMG), electrocardiograms (ECG), and respiratory channels. Each modality offers a unique perspective: brain activity signals (BAS) measure brain function, ECG monitors heart rhythms, and respiratory sensors quantify breathing patterns, together providing a comprehensive assessment of sleep health.

Accurately analyzing sleep data is crucial due to the complexity of sleep disorders. Manual analysis, which involves visual inspection by trained technicians, is time-consuming, labor-intensive, and error-prone. This traditional method faces significant challenges, especially with the increasing volume of sleep data. Therefore, there is a pressing need for automated techniques that can efficiently and accurately analyze sleep data across multiple physiological signals. The goal is to develop robust models that can handle the complexity of sleep data and provide reliable diagnoses.

Current methods for sleep data analysis are mainly based on supervised deep learning models. These models have shown promise in automating sleep staging and classifying sleep disorders such as sleep-disordered breathing (SDB). However, most existing methods rely on limited task-labeled data and do not take advantage of the full range of physiological signals available from the PSG. For example, DL models such as CNN and RNN have been proposed for sleep scoring tasks, but they often need to catch up in generalization and robustness. Furthermore, while contrastive learning (CL) has been successful in other domains, its application in integrating BAS, ECG, and respiratory signals for sleep analysis remains unexplored.

Researchers from Stanford University and the Technical University of Denmark presented SleepFM, an innovative multimodal core model for sleep analysis. This model leverages a large data set of multimodal sleep recordings from more than 14,000 participants, with a total of more than 100,000 hours of sleep data collected between 1999 and 2020 at the Stanford Sleep Clinic. SleepFM uses a contrastive learning approach to integrate brain activity, ECG and respiratory signals. This integration allows the model to capture comprehensive physiological representations, significantly improving the accuracy of sleep analysis.

SleepFM employs three 1D convolutional neural networks (CNN) to generate embeddings of each modality (BAS, ECG, and respiratory signals). The architecture of these models is based on a 1D CNN developed to classify ECG measurements. Each CNN is designed to handle the specific characteristics of its respective modality: 10 channels for BAS, 2 for ECG, and 7 for respiratory channels. A novel leave-one-out contrastive learning technique is introduced, which significantly outperforms standard pairwise contrastive learning in capturing the synergy between different physiological signals.

In sleep stage classification, SleepFM achieved a macro AUROC of 0.88 and a macro AUPRC of 0.72, compared to 0.72 and 0.48 for end-to-end CNNs. SleepFM outperformed CNNs with an AUROC of 0.85 and AUPRC of 0.77 for detecting sleep-disordered breathing, compared to 0.69 and 0.61 for CNNs. Additionally, SleepFM additions demonstrated an average accuracy of 48% in retrieving corresponding recording clips from other modalities from 90,000 candidates. These results underline the model's ability to integrate various physiological signals and improve the accuracy and efficiency of sleep analysis.

The success of the model is primarily attributed to its ability to learn rich, multimodal representations of physiological data, which are crucial for accurate sleep analysis. SleepFM also excelled in demographic attribute classification, showing high accuracy in predicting age and sex from 30-second physiological data clips. The model achieved AUROCs of 0.982, 0.852, 0.784, and 0.915 for the 0-18, 18-35, 35-50, and 50+ age groups, respectively. For gender classification, the AUROC was 0.850, significantly outperforming the reference models.

In conclusion, SleepFM represents significant progress in sleep medicine by providing an automated, accurate, and efficient method for analyzing multimodal sleep data. SleepFM offers a holistic approach to understanding sleep patterns and diagnosing disorders by integrating brain activity, ECG and respiratory signals. The model's superior performance on various tasks, including sleep stage classification, sleep-disordered breathing detection, and demographic prediction, highlights its potential to transform clinical practices in sleep medicine. The success of SleepFM demonstrates the value of holistic, multimodal sleep modeling to capture the richness of sleep records, ultimately contributing to a better understanding and improvement of sleep health.

Review the Paper and GitHub. All credit for this research goes to the researchers of this project. Also, don't forget to follow us on twitter.com/Marktechpost”>twitter. Join our Telegram channel, Discord channeland LinkedIn Grabove.

If you like our work, you will love our Newsletter..

Don't forget to join our 43k+ ML SubReddit | Also, check out our ai Event Platform