Machine learning in vascular surgery: a systematic review and critical appraisal

Abstract

Machine learning (ML) is a rapidly advancing field with increasing utility in health care. We conducted a systematic review and critical appraisal of ML applications in vascular surgery. MEDLINE, Embase, and Cochrane CENTRAL were searched from inception to March 1, 2021. Study screening, data extraction, and quality assessment were performed by two independent reviewers, with a third author resolving discrepancies. All original studies reporting ML applications in vascular surgery were included. Publication trends, disease conditions, methodologies, and outcomes were summarized. Critical appraisal was conducted using the PROBAST risk-of-bias and TRIPOD reporting adherence tools. We included 212 studies from a pool of 2235 unique articles. ML techniques were used for diagnosis, prognosis, and image segmentation in carotid stenosis, aortic aneurysm/dissection, peripheral artery disease, diabetic foot ulcer, venous disease, and renal artery stenosis. The number of publications on ML in vascular surgery increased from 1 (1991–1996) to 118 (2016–2021). Most studies were retrospective and single center, with no randomized controlled trials. The median area under the receiver operating characteristic curve (AUROC) was 0.88 (range 0.61–1.00), with 79.5% [62/78] studies reporting AUROC ≥ 0.80. Out of 22 studies comparing ML techniques to existing prediction tools, clinicians, or traditional regression models, 20 performed better and 2 performed similarly. Overall, 94.8% (201/212) studies had high risk-of-bias and adherence to reporting standards was poor with a rate of 41.4%. Despite improvements over time, study quality and reporting remain inadequate. Future studies should consider standardized tools such as PROBAST and TRIPOD to improve study quality and clinical applicability.

Perceptions of Canadian vascular surgeons toward artificial intelligence and machine learning

“… 28 Our group recently reported a systematic review of ML applications in vascular surgery and identified 212 studies using ML techniques for diagnosis, prognosis, and image segmentation in carotid stenosis, aortic aneurysm/dissection, peripheral artery disease, diabetic foot ulcer, venous disease, and renal artery stenosis. 29 AI/ML can be used to achieve the goals of improving efficiency in clinical practice by learning from large amounts of data to make automated predictions to guide clinical decision-making. 30 The advantage of using ML models is that they can quickly analyze large amounts of data, including a patient’s demographic information, medical history, previous clinical encounters, and imaging data.…”

“…32 A systematic review of ML applications in vascular surgery included 212 relevant studies on diagnosis, prognosis, and image segmentation for six major vascular conditions with good predictive value. 29 The participants reported that AI/ML would be most useful in research and education, reflecting that this technology remains in the research and development phase, with limited applications in routine clinical settings. As AI/ML algorithms become deployed clinically, the perceptions of vascular surgeons regarding areas in which the technology could be helpful could evolve.…”

“…In radiology, ML has been applied to medical image segmentation, computer aided diagnosis, and text analysis of radiology reports through natural language J o u r n a l P r e -p r o o f processing, identifying complex patterns automatically and helping radiologists make decisions more effectively and efficiently 28 . Our group recently published a systematic review of ML applications in vascular surgery and identified 212 studies using ML techniques for diagnosis, prognosis, and image segmentation in carotid stenosis, aortic aneurysm/dissection, peripheral artery disease, diabetic foot ulcer, venous disease, and renal artery stenosis 29 . AI/ML can be used to achieve the goals of improving efficiency in clinical practice by learning from large amounts of data to make automated predictions to guide clinical decision making 30 .…”

“…This is supported by the fact that existing literature demonstrates that AI/ML algorithms can be trained to diagnose peripheral artery disease 33 and predict mortality risk 34 , analyze imaging data to predict abdominal aortic aneurysm growth 2 , automate article screening for systematic J o u r n a l P r e -p r o o f reviews 35 , and facilitate surgical simulations 32 . A systematic review of ML applications in vascular surgery included 212 relevant articles on diagnosis, prognosis, and image segmentation for 6 major vascular conditions with good predictive value 29 . Interestingly, participants believed that AI/ML would be most useful in research and education, which reflects the fact that this technology remains in research and development phase, with limited applications in routine clinical settings.…”

e-Health in Vascular Diseases: Integrating Digital Innovation in Everyday Clinical Practice

“…The main domains include telemedicine/ telehealth, mobile applications (also called “m-health”), the development of smart devices such as sensors and wearables, the use of digital technology for healthcare information system or the development of integrated networks, and all are potentially enhanced using AI techniques [ 1 , 6 ]. Digital health represents a considerable potential for improving the management of vascular diseases and has attracted growing interest for clinicians, researchers, patients, companies, institutional and policy makers, as shown by an exponential increase in academic publications [ 7 , 8 ]. A substantial growth of the global market for digital health could be observed over the past decade [ 9 ].…”