Innovative implementation of artificial intelligence techniques for enhanced management of aortic abdominal aneurysms - ABSTRACT Abdominal aortic aneurysms (AAAs) are a significant healthcare challenge, particularly among the elderly. The presence of intraluminal thrombus (ILT) is common in AAAs, yet its precise role in aneurysm progression remains a topic of ongoing debate. Emerging evidence suggests that ILT substantially contributes to AAA growth by creating an environment that promotes proteolytic activity, leading to the degradation of elastin and collagen in the arterial wall. Despite recognition of the role of blood flow patterns in ILT formation and AAA progression, a systematic approach to understanding these relationships is lacking. This study aims to bridge this gap by investigating both gross and local blood flow characteristics to uncover the mechanisms underlying ILT initiation and subsequent AAA rupture risk. Leveraging my expertise in computational modeling, along with new training in deep-learning-based image processing and statistical analysis, this research will examine the spatiotemporal dynamics of hemodynamics and their correlations with ILT development. Machine learning (ML) algorithms will be utilized to analyze global flow indices and near-wall flow characteristics, enhancing our ability to predict ILT formation and AAA rupture risk. The study will first identify general blood flow characteristics linked to ILT development, followed by an in- depth analysis of local hemodynamics in distinct sub-regions of the aneurysm to reveal their specific roles in localized ILT formation and AAA expansion. Finally, both global and local flow characteristics will be integrated using ML to establish a relationship between hemodynamics, ILT accumulation, and rupture risk. Through this comprehensive exploration, the study seeks to provide a more coherent understanding of AAA severity and the interplay between blood flow, ILT, and rupture risk. Overall, this project aims to bridge the gap between aneurysmal hemodynamics, ILT formation, and AAA rupture risk, offering new insights for clinical management and risk assessment of AAA patients. This fellowship will further enhance my expertise in AI-driven medical image processing and statistical analysis while fostering collaboration with clinical experts to advance my career in vascular research.
