The latest groundbreaking usage for AI comes in the field of medicine. Researchers from the University of Nottingham have developed an algorithm using public data which can radically improve predictability of cardiovascular diseases. In simple terms, they’ve made an algorithm that can predict heart diseases.
The system currently in usage comes courtesy of American College of Cardiology (ACC) and American Heart Association (AHA). Their guidelines have the odds of correctly predicting heart diseases at 72.8%, through measuring eight factors: gender, age, smoking status, systolic blood pressure, blood pressure treatment, total cholesterol, HDL cholesterol, and diabetes.
Doesn’t Account for All Heart Disease Factors
However, as the researchers found, the system failed to account for a bunch of factors with a strong correlation to heart disease, such as mental illness. Using the AI developed by Nottingham University researchers, the chances of predicting heart diseases improves to around 74.5 to 76.4 percent, with the neural network model beating the traditional system by 7.6 percent at its peak. It also managed to curb false positives by 1.6 percent.
There’s a lot of interaction in biological systems. That’s the reality of the human body. What computer science allows us to do is to explore those associations.
Already Proving Itself
The researchers first used 295,000 patients’ records to develop a prediction model, and then tested it on 83,000 further records. The resulting outcome predicted 355 further patients having a risk of cardiovascular disease than the existing guidelines.
As the researchers claimed, the ACC guidelines tend to “oversimplify” the factors which judge cardiovascular risk. The algorithm used by the researchers makes it easier to add further medical, lifestyle and genetic factors in the mix, further boosting predictability.
The study is already garnering praise based on its outcome. With 20 million people dying of cardiovascular diseases each year, the research can prove ground-breaking in helping identify further potential victims to it.
