Do you want to know when you’re going to die?
Discover our latest podcast
Researchers at Geisinger, a health care provider in Pennsylvania, USA, have been able to predict which patients are likely to die within the year. To do this, they trained artificial intelligence (AI) to detect signs of potential heart problems in the future, such as heart attacks or atrial fibrillation.
The machine examined the results of 1.77 million electrocardiograms (ECGs), which are the recordings of heart activity, of almost 400,000 patients. Two versions of the AI have been developed. One analysed the raw data, the other received the age and sex of the participants. The conclusions of the study, reported by New Scientist, were presented at the American Heart Association congress in Dallas (USA) in November 2019.
Things we can’t see
The AI predictions were then analysed, using what is called ‘AUC.’ This metric measures the performance of a model that distinguishes between two groups of people. In this case, the deceased on the one hand, and the survivors on the other. A score of 0.5 indicates no difference between the groups. A score of 1 is perfect. However, the technology systematically scored higher than 0.85.Doctors, on the other hand, scored between 0.65 and 0.8, according to the authors of the study.
According to them, the AI model works better than existing methods when it comes to detecting potential deaths. It even identified cardiac problems in patients already studied by cardiologists. Three doctors each examined ‘normal looking’ ECGs and were not able to detect risk profiles, as the machine did. Brandon Fornwalt, the study's lead researcher, said to New Scientist:
This finding suggests that the model sees things that humans probably can't see, or at least that we don't know and think are normal. Artificial intelligence can teach us things that we may have misinterpreted for decades.
Still unclear how it works
However, scientists still do not know what patterns are detected by the AI, and thus have difficulty explaining how it works. It is this lack of knowledge that worries health professionals, who are reluctant to make decisions based on a misunderstood algorithm.
This study is not the first attempt to predict death. Last year, Google researchers created a predictive model using electronic health records to predict how long a patient will stay in the hospital, when he or she will leave, and when he or she will die. AIs have also been developed to diagnose heart disease and lung cancer, sometimes more accurately than human doctors.
