Artificial intelligence can detect diseases 10 years in advance: Heart rhythm in the light phase of sleep gives a 10-year risk of stroke and crisis

A new clinical study that will revolutionize the early diagnosis of cardiovascular diseases in the world of cardiology, neurology, artificial intelligence and sleep medicine has been announced to the public. The research, conducted jointly by Emory University, Georgia Institute of Technology and Harvard Medical School-affiliated hospitals, documented that micro-changes in the electrical conduction of the heart in the early stages of sleep predict vital risks.

The study published in the journal SLEEP; It was carried out as a result of processing the polysomnography (sleep test) and ECG data of more than 39 thousand patients with deep learning algorithms.

Research findings; proved that the rapid autonomic changes experienced by the nervous system during the transition to sleep reveal hidden abnormalities in the electrical conduction and beat dynamics of the heart. ECG data, especially recorded during the lightest phase of sleep, makes cardiovascular risks visible in individuals who do not yet show any clinical symptoms.

The developed deep learning algorithm classified the 10-year cardiovascular risk profile of patients with high accuracy, even after taking into account traditional risk factors (age, gender, diabetes, hypertension and smoking).

For each unit increase in the risk score produced by the model; It was found to increase the risk of atrial fibrillation (irregular heart rhythm) by 103% to 172%, the risk of stroke by 19% to 40%, the risk of heart failure by 69% to 133%, and the risk of death from all causes by 65% ​​to 82%.

One of the authors of the study, Prof. Dr. Gari Clifford stated that heart data obtained during sleep tests have been ignored until now, and that scanning these data with artificial intelligence has improved existing risk prediction models by up to 9%.

Acting Director of the NIH National Heart, Lung and Blood Institute, Dr. David Goff announced that the method will be directly integrated into routine sleep disorder screening tests in the near future.

Experts report that by identifying patients in the risk group years before symptoms appear, personalized preventive treatments will be initiated and preventable heart attacks and strokes will be prevented.