SleepFM: Multi-modal Representation Learning for Sleep Across Brain Activity, ECG and Respiratory Signals

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Thapa, Rahul, He, Bryan, Kjaer, Magnus Ruud, Moore, Hyatt, Ganjoo, Gauri, Mignot, Emmanuel, Zou, James
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916262905380864
author Thapa, Rahul
He, Bryan
Kjaer, Magnus Ruud
Moore, Hyatt
Ganjoo, Gauri
Mignot, Emmanuel
Zou, James
author_facet Thapa, Rahul
He, Bryan
Kjaer, Magnus Ruud
Moore, Hyatt
Ganjoo, Gauri
Mignot, Emmanuel
Zou, James
contents Sleep is a complex physiological process evaluated through various modalities recording electrical brain, cardiac, and respiratory activities. We curate a large polysomnography dataset from over 14,000 participants comprising over 100,000 hours of multi-modal sleep recordings. Leveraging this extensive dataset, we developed SleepFM, the first multi-modal foundation model for sleep analysis. We show that a novel leave-one-out approach for contrastive learning significantly improves downstream task performance compared to representations from standard pairwise contrastive learning. A logistic regression model trained on SleepFM's learned embeddings outperforms an end-to-end trained convolutional neural network (CNN) on sleep stage classification (macro AUROC 0.88 vs 0.72 and macro AUPRC 0.72 vs 0.48) and sleep disordered breathing detection (AUROC 0.85 vs 0.69 and AUPRC 0.77 vs 0.61). Notably, the learned embeddings achieve 48% top-1 average accuracy in retrieving the corresponding recording clips of other modalities from 90,000 candidates. This work demonstrates the value of holistic multi-modal sleep modeling to fully capture the richness of sleep recordings. SleepFM is open source and available at https://github.com/rthapa84/sleepfm-codebase.
format Preprint
id arxiv_https___arxiv_org_abs_2405_17766
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle SleepFM: Multi-modal Representation Learning for Sleep Across Brain Activity, ECG and Respiratory Signals
Thapa, Rahul
He, Bryan
Kjaer, Magnus Ruud
Moore, Hyatt
Ganjoo, Gauri
Mignot, Emmanuel
Zou, James
Machine Learning
Artificial Intelligence
Signal Processing
Sleep is a complex physiological process evaluated through various modalities recording electrical brain, cardiac, and respiratory activities. We curate a large polysomnography dataset from over 14,000 participants comprising over 100,000 hours of multi-modal sleep recordings. Leveraging this extensive dataset, we developed SleepFM, the first multi-modal foundation model for sleep analysis. We show that a novel leave-one-out approach for contrastive learning significantly improves downstream task performance compared to representations from standard pairwise contrastive learning. A logistic regression model trained on SleepFM's learned embeddings outperforms an end-to-end trained convolutional neural network (CNN) on sleep stage classification (macro AUROC 0.88 vs 0.72 and macro AUPRC 0.72 vs 0.48) and sleep disordered breathing detection (AUROC 0.85 vs 0.69 and AUPRC 0.77 vs 0.61). Notably, the learned embeddings achieve 48% top-1 average accuracy in retrieving the corresponding recording clips of other modalities from 90,000 candidates. This work demonstrates the value of holistic multi-modal sleep modeling to fully capture the richness of sleep recordings. SleepFM is open source and available at https://github.com/rthapa84/sleepfm-codebase.
title SleepFM: Multi-modal Representation Learning for Sleep Across Brain Activity, ECG and Respiratory Signals
topic Machine Learning
Artificial Intelligence
Signal Processing
url https://arxiv.org/abs/2405.17766