Artificial intelligence predicts heart attacks with ease

A British study has shown how machine learning can anticipate many heart diseases that afflict millions of people

One of the causes of sudden, and in some cases even silent, death is heart attack. Over the years, however, medicine has made great strides in an effort to detect the onset of heart disease well in advance. As in the case of research conducted by the University of Nottingham.

The English academic institution, in fact, has developed a prevention study based on artificial intelligence that overcomes some of the limitations present in current analysis techniques. In medicine, a number of standard criteria are used to assess the risk of developing heart disease: gender, age, smoking, total cholesterol level, HDL cholesterol, diabetes and blood pressure. According to the scientists who developed the survey, these models are not always able to prevent heart failure. Help could come from the use of machine learning, i.e., advanced learning.

The heart-saving study

The idea behind it revolves around the possibility of minimizing human error by more accurately analyzing a set of data about people's health status and using some intelligent algorithms. The 10-year study tasked machine learning with monitoring and predicting the onset over time of cardiovascular disease, identified in medicine as CVD (cardiovascular disease), in apparently healthy subjects. The research was conducted on more than 378 thousand patients from the age of 30 to 84 years: 259 thousand were used to train the algorithms, while the rest of the participants served to validate the machines.

Machine learning predicts many heart diseases

During the study, which began on January 1, 2005 and ended on January 1, 2015, scientists were able to identify 25 thousand cardiovascular diseases, some of them potentially fatal. 75% of the diseases were accurately predicted by machine learning algorithms. And there's more. Compared to doctors, the machines identified 355 more patients suffering from cardiovascular disorders.