AI Predicts Heart Failure Risk from Sleep Patterns
A new study reveals that heart activity during sleep can forecast cardiovascular risks up to a decade in advance.
AI can predict heart failure and stroke risks using ECG data from sleep.
The study involved nearly 16,000 patients and showed significant predictive capabilities.
Incorporating sleep ECG data enhances traditional cardiovascular risk models.
Recent research indicates that the way a person's heart functions during sleep can provide crucial insights into their future cardiovascular health. A study published in the journal SLEEP demonstrates that by analyzing heart electrical activity captured through a single electrocardiogram (ECG) lead while sleeping, artificial intelligence (AI) can predict the likelihood of heart failure, irregular heartbeat, stroke, or even death within the next ten years.
The research, spearheaded by scientists at Emory University, utilized data from nearly 16,000 patients who participated in a sleep study at Massachusetts General Hospital. The AI system was trained on this extensive dataset, which included ECG recordings taken during sleep. After training, the AI was tested on additional datasets from Emory University Hospital and Beth Israel Deaconess Medical Center, encompassing nearly 23,000 more patients. The results indicated that the AI could categorize participants based on their long-term cardiovascular risk levels.
The findings revealed alarming statistics: for each standard deviation increase in the AI's risk score, there was a 103 to 172 percent increase in the risk of atrial fibrillation, a 19 to 40 percent increase in stroke risk, and a 69 to 133 percent increase in heart failure risk. Additionally, the AI's predictions showed a 65 to 82 percent higher risk of mortality from any cause. Notably, only one group demonstrated a significantly increased risk of future heart attacks.
The implications of this research are profound, particularly for public health and patient care. Traditional cardiovascular risk assessments often overlook the critical data available during sleep. By integrating ECG data from sleep studies, healthcare providers could identify individuals at heightened risk for cardiovascular diseases long before symptoms manifest. This proactive approach could lead to earlier interventions and better health outcomes for patients.
Looking ahead, researchers emphasize the importance of further validating these findings and integrating AI-driven analyses into routine diagnostic procedures. As the technology evolves, it holds the potential to revolutionize how cardiovascular risks are assessed, particularly for patients with sleep disorders. Future studies may refine these predictive models, ultimately enhancing the accuracy of cardiovascular risk assessments and improving patient care strategies.
