Researchers at Stanford University have developed a new AI model which can predict more than 100 health conditions including cancer, mental health, cardiovascular issues and death using sleep study data. The AI model has been trained on nearly 600,000 hours of sleep data collected from 65,000 participants which had been split into five-second increments to combine multiple body signals, such as brain activity, heart activity, muscle activity, pulse and breathing and understand how they relate to each other. After being trained to identify and link different body signals, SleepFM was taught how to understand standard sleep analysis tasks such as different stages of sleep and diagnosing the severity of sleep apnea, a serious sleep disorder where breathing repeatedly stops and starts often due to airway blockage or the brain failing to signal muscles. Once the model was seen successfully identifying sleep-related issues, it was paired with the sleep recordings and health data of 35,000 patients collected over 25 years for it to learn how to identify chronic diseases through nighttime body signals.
Thank you for reading this post, don't forget to subscribe!Scientists Forecast Over One Hundred Conditions Through One Night Sleep Using New AI Model
