Machine Learning, Deep Learning and Convolutional Neural Networks on Improving Patient and Learner Outcomes for Endovascular Aneurysm Repair in Infrarenal Abdominal Aortic Aneurysm Patients — A Review

Abstract

Integration of artificial intelligence (AI), machine learning (ML) and deep learning (DL) has ushered in new era of intraoperative treatment and life expectancy predictions with respect to endovascular aneurysm repair (EVAR) in abdominal aortic aneurysms (AAA). Objective was to systematically review recent studies applying AI/ML/DL in EVAR and aortic repairs. A systematic review was conducted in accordance with PRISMA 2020 guidelines. Five peer-reviewed studies (2022–2025) were included based on DL-based intraoperative stent graft segmentation, ML random forest based 3-year survival prediction, AI-driven sac volume measurement, deep learning for CT angiography (CTA) artifact reduction. All five studies demonstrated improved performance with ML, DL and CNN in patient and learner centric outcomes. Small sample size, single-center validation, software dependent aspects for patient and learner related outcomes needing further validation for improved result generalizability. AI, ML, DL show strong potential of improvement across EVAR workflows.

Ähnliche Arbeiten