Newsletter
Artificial intelligence (AI) may be able to predict whether a person is likely to suffer sudden cardiac death, according to a new study published in MedicalXpress on Monday. The data used to train the AI model was far-reaching, and included more than one million hospital diagnoses and 10 million medication prescriptions. It came from the records of 25,000 people who had died from sudden heart attacks. Then, the AI model crossed the data profile against health information from 70,000 people in the general public. The study found that AI models could detect who had over 90% risk of dying suddenly.
“Predicting [sudden cardiac death] is difficult,” said Xavier Jouven, M.D., Ph.D., the lead author of the study, in the report. “The usual approaches fail to identify high-risk people, particularly at an individual level.” This is where AI comes into play. AI can analyze the full portfolio of a person’s medical information, seeing what the human brain misses. That’s how AI can detect the probability of a sudden heart attack at a higher level than human doctors. While past research has tried to link past cardiovascular history to future heart attacks, this new AI model is considered “a mix of neurological, psychiatric, metabolic and cardiovascular data.”
There are notable limitations to this study. The researchers note how medical histories often include proxies, which are averages or substitutes, instead of raw data. The differences between data proxies and the raw data, at scale, could affect the success of the AI training model. In turn, this could decrease AI’s effectiveness predicting heart attacks.
AI has been used in health in the past, and will be used in the future
Artificial intelligence is no stranger to the health space. Specifically, AI can analyze large data sets thoroughly much more quickly and effectively than human researchers. It is also useful when tied into the vast array of consumer health devices out in the wild. For example, researchers at the Mayo Clinic used ECG recordings from Apple Watches to identify weak hearts in a study published last year. Rumors also point to Apple itself working on an AI-equipped health coach for future release, but these are unconfirmed.
Jouven notes that medical professionals do have recourse to prevent heart attacks in high-risk patients. The hard part is identifying which patients are high-risk. “The use of AI is necessary to detect in a given subject a succession of medical information registered over the years that will form a trajectory associated with an increased risk of sudden cardiac death,” Jouven said. The researchers hope that with AI, heart attacks can be prevented with personalized health reports in the future.