Artificial intelligence (AI) and machine learning (ML) are being used more and more in healthcare and medicine over the past 20 years. This study aimed to look at all the current uses of AI and ML in the field of vascular surgery and to analyze the related research papers that have been published.
We looked at a large number of research papers from various databases until February 19, 2021. The study was reported according to widely accepted guidelines for research review. The process involved checking the titles and summaries, reading the full text of the papers, and collecting relevant information twice over to ensure accuracy. We collected information about the research topic, the specific area of vascular surgery studied, the kind of AI/ML used, the data used, and how AI/ML was applied. We also classified the journals where the papers were published as either clinical (related to medicine) or technical. The researchers were classified as having a medical background, a non-medical background (like engineering), or both, based on their affiliations.
We found 7,434 research papers, and out of these, 249 were selected for detailed analysis. The number of papers being published is growing rapidly, with 158 (63%) of them published in the last 5 years. Most of the studies were about diseases related to the carotid artery (which supplies blood to the brain), abdominal aortic aneurysms (bulges in the large artery of the abdomen), and peripheral arterial disease (narrowing of arteries other than those supplying the heart or brain). On average, the researchers used 1.5 different AI methods in their studies. AI/ML was mostly used to predict disease outcomes, analyze images, diagnose diseases, or a combination of these. The most common AI/ML methods used were artificial neural networks, support vector machines, the k-nearest neighbors algorithm, and random forests. These methods were applied to various data, including ultrasound and CT scan images, clinical data, or a mix of data types. Only 22 (9%) of the studies were published in journals specific to blood vessel surgery, while most were published in journals related to computer science or engineering. Among the authors, 46% had a medical background, 48% had a non-medical background, and 5% had both.
The use of AI and ML in vascular surgery is growing quickly. There’s a lot of focus on carotid artery disease and abdominal aortic disease, but many other areas of vascular surgery aren’t getting as much attention. Neural networks and support vector machines are the most commonly used AI methods in these studies. As AI/ML is used more and more in this field, it’s important for surgeons to understand its potential and limitations. Also, as AI/ML use increases, there is a need for doctors who are experts in these methods to help bring them into everyday practice.