Magnus Ruud Kjaer

Diagnosing sleep disordered breathing requires manual annotation of events from sleep studies, such as nocturnal polysomnography, a process that is time-intensive, costly, and prone to inter-rater variability. Automatic approaches exist but lack generalizability due to signal variability across centers. We develop an automatic apneic breathing event detector to localize and classify obstructive apneas, central apneas, hypopneas, and isolated respiratory events without arousals or desaturations. The model is trained on 5456 polysomnographies and tested on 1099 polysomnographies from six cohorts uses an end-to-end deep learning architecture. The model’s predictions show a strong correlation with expert annotations for apnea-hypopnea index (r² = 0.84) and achieve an F1 score of 0.78 across apnea event types, with specific F1 scores of 0.71, 0.51, and 0.65 for obstructive apnea, central apnea, and hypopnea events, respectively. In two independent, multi-scored datasets, The model performs comparably or better than individual expert raters. The model’s probabilistic output, termed “apnotyping,” provides insights into sleep disordered breathing etiology, with event probabilities correlating more strongly with key sleep apnea traits—such as loop gain and pharyngeal muscle compensation—than traditional apnea indexes. This probabilistic approach may enhance diagnostic accuracy and support personalized treatment strategies, leading to improved patient outcomes.

Sleep is a fundamental biological process with broad implications for physical and mental health, yet its complex relationship with disease remains poorly understood. Polysomnography (PSG)—the gold standard for sleep analysis—captures rich physiological signals but is underutilized due to challenges in standardization, generalizability and multimodal integration. To address these challenges, we developed SleepFM, a multimodal sleep foundation model trained with a new contrastive learning approach that accommodates multiple PSG configurations. Trained on a curated dataset of over 585,000 hours of PSG recordings from approximately 65,000 participants across several cohorts, SleepFM produces latent sleep representations that capture the physiological and temporal structure of sleep and enable accurate prediction of future disease risk. From one night of sleep, SleepFM accurately predicts 130 conditions with a C-Index of at least 0.75 (Bonferroni-corrected P < 0.01), including all-cause mortality (C-Index, 0.84), dementia (0.85), myocardial infarction (0.81), heart failure (0.80), chronic kidney disease (0.79), stroke (0.78) and atrial fibrillation (0.78). Moreover, the model demonstrates strong transfer learning performance on a dataset from the Sleep Heart Health Study—a dataset that was excluded from pretraining—and performs competitively with specialized sleep-staging models such as U-Sleep and YASA on common sleep analysis tasks, achieving mean F ₁ scores of 0.70–0.78 for sleep staging and accuracies of 0.69 and 0.87 for classifying sleep apnea severity and presence. This work shows that foundation models can learn the language of sleep from multimodal sleep recordings, enabling scalable, label-efficient analysis and disease prediction.

Polysomnography (PSG), the gold standard test for sleep analysis, generates vast amounts of multimodal clinical data, presenting an opportunity to leverage self-supervised representation learning (SSRL) for pre-training foundation models to enhance sleep analysis. However, progress in sleep foundation models is hindered by two key limitations: (1) the lack of a shared dataset and benchmark with diverse tasks for training and evaluation, and (2) the absence of a systematic evaluation of SSRL approaches across sleep-related tasks. To address these gaps, we introduce Stanford Sleep Bench, a large-scale PSG dataset comprising 17,467 recordings totaling over 163,000 hours from a major sleep clinic, including 13 clinical disease prediction tasks alongside canonical sleep-related tasks such as sleep staging, apnea diagnosis, and age estimation. We systematically evaluate SSRL pre-training methods on Stanford Sleep Bench, assessing downstream performance across four tasks: sleep staging, apnea diagnosis, age estimation, and disease and mortality prediction. Our results show that multiple pretraining methods achieve comparable performance for sleep staging, apnea diagnosis, and age estimation. However, for mortality and disease prediction, contrastive learning significantly outperforms other approaches while also converging faster during pretraining. To facilitate reproducibility and advance sleep research, we will release Stanford Sleep Bench along with pretrained model weights, training pipelines, and evaluation code.