Personalized diagnosis and treatment of aortic diseases driven by artificial intelligence: Applications, challenges and future prospects

Aortic aneurysm (AA) and aortic dissection (AD) are major forms of aortic disease and remain associated with substantial morbidity and mortality worldwide. Traditional diagnostic and treatment strategies largely depend on anatomical indicators, particularly maximum aortic diameter; however, such parameters are often inadequate for accurately reflecting patient-specific disease progression and rupture risk. With the rapid development of artificial intelligence (AI), especially machine learning (ML) and deep learning (DL), new opportunities have emerged to overcome these limitations. This review focuses on the role of AI in advancing personalized diagnosis and management of aortic diseases. Specifically, it summarizes recent progress in four major areas. First, DL-based approaches have shown high accuracy in automated image segmentation and computer-aided diagnosis, including segmentation of the aortic lumen, intraluminal thrombus, and true and false lumens. Second, ML models incorporating biomechanical and multi-physics features have demonstrated superior performance over conventional diameter-based criteria in rupture risk assessment. Third, unsupervised learning has enabled the identification of clinically and biologically relevant molecular-imaging subtypes beyond traditional anatomical classifications. Fourth, AI has been increasingly applied to therapeutic decision-making and postoperative outcome prediction, including target identification, endovascular procedural guidance, and early detection of complications. In addition, this review discusses key barriers to clinical implementation, such as limited data quality, model overfitting, poor interpretability of DL models, and uncertainties in regulatory approval. Future advances in multicenter standardized datasets, multimodal fusion strategies, and explainable AI (XAI) are expected to accelerate the transition of aortic disease management toward AI-supported precision medicine.