Researchers at Stanford University have developed an AI-backed system that uses sleep data to predict the future risk of disease onset. These predictions are based off a single night’s sleep data. Their findings suggest that the physiological signals collected while you sleep can be valuable pointers to the diseases that you could suffer from years later.
SleepFM, as the system is called, leverages data collected on breathing patterns, heart function, brain activity, leg movements as well as the other physical signals that happen while you sleep. In all, the system was trained using data on 65,000 people. The information was collected during polysomnography analyses conducted from 1999 to 2024 at the Sleep Medicine Center at Stanford. The electronic health records of the people who underwent the sleep tests were also collected. The medical records spanned more than two decades for most patients.
This data was then used during the AI system’s training and it emerged that the system could predict with high accuracy the likelihood that someone could suffer from any of the 130 conditions that the researched selected from a broad list.
Those 130 conditions were narrowed down to exclude those where SleepFM had less than 80% accuracy rates in its predictions. The highest predictions covered circulatory diseases, pregnancy complications, cancers and mental health issues. Parkinson’s disease stood out with an accuracy rate of 89%, same as prostate cancer risk prediction, while dementia came close to it with an 85% prediction accuracy rate.
The researchers noted that for many of the conditions that had lower prediction rates in the SleepFM system, hospitals are already leveraging AI to predict those states. These included systems that help in predicting whether a given patient will respond well to a certain cancer treatment approach.
While assessing the SleepFM system, the Stanford team discovered that it performed best if it used different sets of data to predict any disease. For example, one would expect that the system would rely on circulatory data alone while predicting the risk of diseases that affect the circulatory system. However, doing that lowered accuracy rates and the predictions only improved when the system also used seemingly unrelated signals like brain function and breathing activity.
The researchers are now working to make further improvements to the ability of the AI system to predict disease risk, and they also hope to find ways to incorporate additional data taken from wearable devices so that the range of the diseases that SleepFM can predict expands beyond the 130 diseases.
As companies like Microsoft Corp. (NASDAQ: MSFT) continue their aggressive buildout of AI data centers while also developing better Large Language Models (LLMs), there is no telling what additional innovative uses of AI could emerge in the years to come.
About TrillionDollarClub
TrillionDollarClub (“TDC”) is a specialized communications platform with a focus on the biggest and brightest companies covered by IBN. It is one of 75+ brands within the Dynamic Brand Portfolio @ IBN that delivers: (1) access to a vast network of wire solutions via InvestorWire to efficiently and effectively reach a myriad of target markets, demographics and diverse industries; (2) article and editorial syndication to 5,000+ outlets; (3) enhanced press release enhancement to ensure maximum impact; (4) social media distribution via IBN to millions of social media followers; and (5) a full array of tailored corporate communications solutions. With broad reach and a seasoned team of contributing journalists and writers, TDC is uniquely positioned to best serve private and public companies that want to reach a wide audience of investors, influencers, consumers, journalists, and the general public. By cutting through the overload of information in today’s market, TDC brings its clients unparalleled recognition and brand awareness.
TDC is where breaking news, insightful content and actionable information converge.
To receive SMS alerts from TrillionDollarClub, text “Trillion” to 888-902-4192 (U.S. Mobile Phones Only)
For more information, please visit https://www.TrillionDollarClub.net
Please see full terms of use and disclaimers on the TrillionDollarClub website applicable to all content provided by TDC, wherever published or re-published: https://www.TrillionDollarClub.net/Disclaimer
TrillionDollarClub
Austin, TX
www.TrillionDollarClub.net
512.354.7000 Office
[email protected]
TrillionDollarClub is powered by IBN
