Machine learning, which is a type of artificial intelligence, is being used more and more in healthcare, including in vascular surgery. In this study, we looked at 212 studies that used machine learning in this field. We found that ML it’s being used to diagnose, predict outcomes, and analyze images related to various vascular diseases, including carotid stenosis, aortic aneurysm, peripheral artery disease, diabetic foot ulcer, venous disease, and renal artery stenosis.
There has been a significant increase in the number of these studies over the past few decades, from just one between 1991 and 1996 to 118 between 2016 and 2021. The majority of these studies were based on past data and carried out at a single location, and none of them were randomized controlled trials, which are considered the gold standard in medical research.
The studies generally showed that machine learning was effective, with most demonstrating a good ability to correctly classify patients based on their disease status. In fact, of the 22 studies that compared machine learning to traditional methods or doctors’ judgments, 20 found that machine learning did better and 2 found it did just as well.
However, the researchers also found that nearly 95% of the studies were at high risk of bias, and less than half followed good reporting practices. They suggest that future research in this area should use standardized tools to improve the quality and usefulness of their findings.