A crew of Stanford Drugs researchers have launched SleepFM Scientific, a multimodal sleep basis mannequin that learns from medical polysomnography and predicts long run illness danger from a single night time of sleep. The analysis work is printed in Nature Drugs and the crew has launched the medical code because the open supply sleepfm-clinical repository on GitHub underneath the MIT license.
From in a single day polysomnography to a basic illustration
Polysomnography data mind exercise, eye actions, coronary heart indicators, muscle tone, respiratory effort and oxygen saturation throughout a full night time in a sleep lab. It’s the gold customary check in sleep medication, however most medical workflows use it just for sleep staging and sleep apnea prognosis. The analysis crew deal with these multichannel indicators as a dense physiological time collection and practice a basis mannequin to study a shared illustration throughout all modalities.
SleepFM is educated on about 585,000 hours of sleep recordings from about 65,000 folks, drawn from a number of cohorts. The biggest cohort comes from the Stanford Sleep Drugs Heart, the place about 35,000 adults and kids had in a single day research between 1999 and 2024. That medical cohort is linked to digital well being data, which later permits survival evaluation for a whole lot of illness classes.
Mannequin structure and pretraining goal
On the modeling stage, SleepFM makes use of a convolutional spine to extract native options from every channel, adopted by consideration based mostly aggregation throughout channels and a temporal transformer that operates over brief segments of the night time. The identical core structure already appeared in earlier work on SleepFM for sleep staging and sleep disordered respiratory detection, the place it confirmed that studying joint embeddings throughout mind exercise, electrocardiography and respiratory indicators improves downstream efficiency.
The pretraining goal is depart one out contrastive studying. For every brief time phase, the mannequin builds separate embeddings for every modality group, comparable to mind indicators, coronary heart indicators and respiratory indicators, after which learns to align these modality embeddings in order that any subset predicts the joint illustration of the remaining modalities. This strategy makes the mannequin strong to lacking channels and heterogeneous recording montages, that are widespread in actual world sleep labs.
After pretraining on unlabeled polysomnography, the spine is frozen and small activity particular heads are educated. For normal sleep duties, a light-weight recurrent or linear head maps embeddings to sleep levels or apnea labels. For medical danger prediction, the mannequin aggregates the complete night time right into a single affected person stage embedding, concatenates fundamental demographics comparable to age and intercourse, after which feeds this illustration right into a Cox proportional hazards layer for time to occasion modeling.
Benchmarks on sleep staging and apnea
Earlier than shifting to illness prediction, the analysis crew verified that SleepFM competes with specialist fashions on customary sleep evaluation duties. Prior work already showed that a simple classifier on high of SleepFM embeddings outperforms finish to finish convolutional networks for sleep stage classification and for detection of sleep disordered respiratory, with beneficial properties in macro AUROC and AUPRC on a number of public datasets.
Within the medical examine, the identical pretrained spine is reused for sleep staging and apnea severity classification throughout multi heart cohorts. Outcomes reported within the analysis paper present that SleepFM matches or exceeds present instruments comparable to conventional convolutional fashions and different automated sleep staging programs, which validates that the illustration captures core sleep physiology and never solely statistical artifacts from a single dataset.
Predicting 130 illnesses and mortality from one night time of sleep
The core contribution of this Stanford’s analysis paper is illness prediction. The analysis crew maps prognosis codes within the Stanford digital well being data to phecodes and defines greater than 1,000 candidate illness groupings. For every phecode, they compute time to first prognosis after the sleep examine and match a Cox mannequin on high of SleepFM embeddings.
SleepFM identifies 130 illness outcomes whose dangers are predictable from a single night time of polysomnography with sturdy discrimination. These embody all trigger mortality, dementia, myocardial infarction, coronary heart failure, continual kidney illness, stroke, atrial fibrillation, a number of cancers and a number of psychiatric and metabolic problems. For a lot of of those situations, efficiency metrics comparable to concordance index and space underneath the receiver working curve are in ranges corresponding to established danger scores, though the mannequin makes use of solely sleep recordings plus fundamental demographics.
The reporting additionally notes that for some cancers, being pregnant issues, circulatory situations and psychological well being problems, predictions based mostly on SleepFM attain accuracy ranges round 80 % for multi yr danger home windows. This implies that refined patterns within the coordination between mind, coronary heart and respiratory indicators carry details about latent illness processes that aren’t but clinically seen.
Comparability with easier baselines
To evaluate added worth, the analysis crew in contrast SleepFM based mostly danger fashions with two baselines. The primary makes use of solely demographic options comparable to age, intercourse and physique mass index. The second trains an finish to finish mannequin instantly on polysomnography and outcomes, with out unsupervised pretraining. Throughout most illness classes, the pretrained SleepFM illustration mixed with a easy survival head yields larger concordance and better lengthy horizon AUROC than each baselines.
This analysis clearly reveals that the acquire comes much less from a fancy prediction head and extra from the muse mannequin that has discovered a basic illustration of sleep physiology. In observe, which means medical facilities can reuse a single pretrained spine, study small website particular heads with comparatively modest labeled cohorts and nonetheless strategy cutting-edge efficiency.
Take a look at the Paper and FULL CODES here. Additionally, be at liberty to comply with us on Twitter and don’t overlook to affix our 100k+ ML SubReddit and Subscribe to our Newsletter. Wait! are you on telegram? now you can join us on telegram as well.
Take a look at our newest launch of ai2025.dev, a 2025-focused analytics platform that turns mannequin launches, benchmarks, and ecosystem exercise right into a structured dataset you possibly can filter, evaluate, and export
Asif Razzaq is the CEO of Marktechpost Media Inc.. As a visionary entrepreneur and engineer, Asif is dedicated to harnessing the potential of Synthetic Intelligence for social good. His most up-to-date endeavor is the launch of an Synthetic Intelligence Media Platform, Marktechpost, which stands out for its in-depth protection of machine studying and deep studying information that’s each technically sound and simply comprehensible by a large viewers. The platform boasts of over 2 million month-to-month views, illustrating its reputation amongst audiences.
