Doctors may have spent years long years studying and specializing to become cardiologists, but they haven’t come close to accurately predicting when a person will suffer from a heart attack. But now, a bold development using artificial intelligence has started predicting cardiac incidents with amazing accuracy. Could this be the life-saving technology that the world has been waiting for?
“scientists have shown that computers capable of teaching themselves can perform even better than standard medical guidelines, significantly increasing prediction rates”
Over 20 million people a year suffer from cardiac arrest, around 800,000 of these are Americans, the Center for Disease Control and Prevention reported. Additionally, one person in the US has a heart attack every 43 seconds. For the longest time, the medical community has accepted that heart attacks are almost impossible to predict given the different factors and confluence of situations that lead to it.
However, Science Magazine reports that computers equipped with artificial intelligence “can be trained” to recognize heart attack risk factors such as blood pressure, cholesterol levels, and age among other factors. “Now, scientists have shown that computers capable of teaching themselves can perform even better than standard medical guidelines, significantly increasing prediction rates. If implemented, the new method could save thousands or even millions of lives a year,” the article read.
But while this development is seen as a game changer in the medical world, it’s not surprising that it is met by experts with raised eyebrows. This is why Vascular Surgeon Elsie Ross from the Stanford University in Palo Alto, California is appealing to doctors to “start to embrace the use of artificial intelligence to assist us in care of patients.”
University of Nottingham Epidemiologist Stephen Weng explains that the human body is a complex interaction of biological systems and computer science explores these associations in order to come up with predictive tools. Weng’s program sorted through the medical record of 378,256 patients from the UK with the aim of finding patterns that could predict cardiovascular events.
After testing and creating base patterns, the AI algorithms were able to come up with predictions on which patients would experience cardiovascular events over the next 10 years. Since the data the computers were fed came from 2005, researchers were able to match and counter check if their predictions did come true. According to Weng, “the best one—neural networks—correctly predicted 7.6% more events than the ACC/AHA method, and it raised 1.6% fewer false alarms. In the test sample of about 83,000 records, that amounts to 355 additional patients whose lives could have been saved. That’s because prediction often leads to prevention.”
“the best one—neural networks—correctly predicted 7.6% more events than the ACC/AHA method, and it raised 1.6% fewer false alarms”
If more computing power and training data were allocated for similar programs, University of Manchester data scientist Evangelos Kontopantelis says Weng’s project could have achieved bigger gains. There are plans to refine the data and include more predictors in the algorithm such as lifestyle and genetic factors, including diabetes.
But before this technology is hailed as the bold discovery of the century, the data scientist noted that there is a big limitation to the use of Artificial Intelligence. He stressed that these machine learning algorithms can hardly be tweaked by humans. Kontopantelis explained that they can see the data that comes in and comes out, but scientists are left guessing what happens in between.
The medical world is slowly inching towards digitization and automation. The use of robots in the health profession has truly brought medicine to the next level. The ability to predict incidences such as heart attacks with a high degree of accuracy could save millions of lives through prevention. This certainly changes everything and leaves a profoundly deep impact in our quality of life.
