US researchers develop AI tool for early stroke detection during sleep
American scientists have created an AI-based model that analyzes sleep ECG data to assess stroke and heart disease risk early, using just one chest electrode.
Key Takeaways
AIAmerican researchers have developed an artificial intelligence-based analysis to assess the future risk of stroke and heart disease by monitoring heart activity during sleep.
The team used data from around 16,000 people who underwent electrocardiogram (ECG) monitoring during sleep to train the software model, then tested its effectiveness on two additional groups totaling about 23,000 patients.
Results showed that data from a single chest-mounted ECG electrode was sufficient to evaluate risk. The AI analyzed changes in heart rhythm during sleep and identified markers linked to the development of future diseases.
The researchers found that with each standard deviation increase in the model's risk score, the likelihood of stroke rose by 19 to 40 percent, the risk of heart failure increased by 69 to 133 percent, and the risk of atrial fibrillation grew by 103 to 172 percent.
They also noted that the early stage of sleep plays a key role in risk assessment, as rapid changes in the nervous system during this phase may make hidden heart rhythm disorders more detectable.
The researchers believe this method could enable early identification of people at higher risk of stroke and other cardiovascular diseases long before symptoms appear.