Artificial intelligence: it would be able to predict the risk of death according to the result of an ECG

Artificial intelligence it would be able to predict the risk

  • News
  • Published on

    Reading 2 mins.

    Canadian researchers have just developed an artificial intelligence program capable of predicting the risk of death from the analysis of examinations, in particular the ECG, carried out in the hospital. Explanations.

    The electrocardiogram (ECG) is a routine examination carried out in the hospital, during a visit to the emergency room or during hospitalization. It allows, via electrodes glued to the thorax, to check the heart rate and the electrical activity of the heart.

    ECGs performed on more than a million patients

    In this study, Canadian researchers worked on an artificial intelligence algorithm to predict patients’ risk of death. For this, scientists presented 1.6 million electrocardiograms performed on 244,077 patients in northern Alberta, a province of Canada, between 2007 and 2020.

    An AI capable of predicting the risk of death

    Result: the program was able to predict the risk of death of these patients at one month, one year and five years, all causes combined and with an accuracy estimated at 85%. Artificial intelligence is also able to classify these patients into five categories according to this risk, from the lowest to the highest.

    The scientists also added other elements to the program, such as information related to the age or sex of the patients or even the results of blood tests. This has made it possible to further increase the accuracy of the predictions made by artificial intelligence.

    According to Padma Kaul, professor of medicine and lead author of this study: “We wanted to know if we could use new methods such as artificial intelligence and machine learning to analyze data and identify patients at higher risk of mortality.“ she explains. “These results illustrate how machine learning models can be used to convert data routinely collected in clinical practice into insights that can be used to improve decision-making..”

    Refine the model for more details

    Now, Dr. Kaul and his team hope to refine these models for particular subgroups of patients. They also plan to focus forecasts beyond all-cause mortality to look specifically at heart-related causes of death. “We want to take the data generated by the healthcare system, convert it into knowledge and feed it back into the system to improve care and outcomes. This is the definition of a learning healthcare system” she concludes.