Artificial Intelligence is better than a human doctor at predicting health threats: Study

There have been many instances in medical science where we have gained an upper hand by utilising Artificial Intelligence.  Now, taking this further, scientists have created an Artificial Intelligence (AI) system that outpaces doctors in predicting mortality and even risks of heart attacks. After an analysis for six years in a study comprising 85 variables in 950 patients with chest pain who were processed as per the centre’s usual protocol when looking for coronary heart disease, scientists put together an algorithm. This was designed to identify warning signs, especially in patients who could die or have heart attacks. This algorithm turned out to be 90 percent accurate at predicting risk. This study was presented at The International Conference on Nuclear Cardiology and Cardiac CT 2019 on 14 May.

The author of this study, Dr Luis Eduardo Juarez-Orozco of the Turku PET Centre, Finland believes that though there have several advances in the field of medicine, but that we’re still not making full use of all the existing data. Doctors generally make use of risk scores to predict such cases to make treatment decisions. After the repetitive running of similar scores, with machine learning, we can get more complex patterns to work with. Juarez-Orozco further explained that humans usually struggle when it comes to thinking of more than four dimension. His study shows that high dimensional patterns are more useful to predict the outcomes in individuals and this is where machine learning can help us.