Artificial Intelligence and Vascular Surgery: A Glance into the Crystal Ball

Abstract

INTRODUCTION Since its introduction at the Dartmouth College Conference in 1956, Artificial Intelligence (AI) has emerged as a transformative force in various industries, and its impact on health care, particularly in the field of surgery, is nothing short of revolutionary. The integration of AI technologies into surgical practices is reshaping the landscape of health care, offering improved precision, efficiency, and patient outcomes. This write-up explores the multifaceted role of AI in vascular surgery, delving into its applications, benefits, challenges, and the promising future it holds for the medical community. FUNDAMENTALS OF ARTIFICIAL INTELLIGENCE IN MEDICINE AI corresponds to a broad discipline that aims to understand and design systems that display properties of human intelligence such as reasoning, learning, adaptation, interaction, and sensory understanding. It uses principles and devices from a variety of fields such as computation, mathematics, logic, mechanics, and biology with the aim to replicate and understand intelligence and cognitive processes.[1] The basic foundations of AI are (a) analyzing large amounts of data, (b) recognizing patterns, (c) predicting outcomes, and (d) drawing conclusions to improve workflow. The four core subfields of AI are: Machine learning This enables machines to learn and make predictions by recognizing patterns. Supervised learning is useful to predict a known result or outcome, whereas unsupervised learning is useful for identifying patterns within data. Reinforcement learning is when a system interacts with its environment so as to maximize the acquisition of a predefined outcome. Machine learning (ML) is useful in identifying subtle patterns in large datasets which may be indiscernible to humans. Natural language processing This is a system that is the computer’s ability to understand human language. The better systems have expanded beyond simple word recognitions and have successfully mined data from electronic medical records. Artificial neural networks These are based on biological nervous systems. Deep learning networks have many layers and can recognize many complexities. Computer vision This is a machine analysis of images and videos. The health-care applications include image acquisition and interpretation in axial imaging, image-guided surgery, and virtual endoscopies.[2] AI applied to health care promises the potential to deliver transformative change that will benefit patients, health-care professionals, and economies in the foreseeable future. While there are a large number of publications on AI related to oncology, radiology, cardiology, and neurology, there are very few on vascular disease management. The current applications of AI in vascular surgery are: Disease prognostication and prediction Image analysis and vascular diagnostics “Big Data” mining and analysis Vascular education.[3] DISEASE PROGNOSTICATION AND PREDICTION Abdominal aortic aneurysm The role of AI in the decision-making strategy for abdominal aortic aneurysm (AAA) repair and management has been successfully studied. Even though the initial maximal diameter is a well-established independent risk factor for AAA rupture, other variables such as AAA growth rate, wall stress, wall tension, intraluminal thrombus, female sex, hypertension, and smoking have been identified and the risk of progression is difficult to predict.[4] One ML technique to predict future AAA growth was applied in the Oxford AAA study, a prospective study of AAA patients treated using the routine National Health Service management pathway. The algorithm predicted an individual AAA diameter to within a 2-mm error in 85% and 71% of patients at 12 and 24 months, respectively, highlighting the important role of ML techniques for AAA research in the current and future eras of vascular medicine.[5] There are many studies, which have been done in endovascular aortic aneurysm repair. While it is a well-established treatment modality, p