Publications

Analyzing aortic biomechanical wall stresses for abdominal aortic aneurysms remains challenging. Clinical applications of biomechanical and morphological image-based analysis protocols have limited adoption owing to the time and expertise required. Our multidisciplinary and multi-institute team has demonstrated the feasibility of expediting advanced aortic image analysis on a single patient tracked longitudinally. We also demonstrate the utility of a previously trained artificial intelligence-based classifier that accurately predicts patient outcomes, a potential alternative to serial surveillance. This paper describes the overall workflow and processes performed in a 70-year-old man who was incidentally diagnosed to have a 5.4-cm juxtarenal aortic aneurysm in 2016 with successful fenestrated endovascular repair in 2023.

Our results reveal a potential avenue for an elastin-targeted small aortic aneurysm therapeutic, acting as a supplement to the currently employed passive monitoring strategy. Additionally, the elastogenesis analysis workflow explored here could guide future mechanistic studies of elastin formation, which in turn could lead to new non-surgical treatment strategies.

Failure of vein grafts via intimal hyperplasia (IH) remains a critical problem, with the 5-year reoperation rate at 60% of all cases[1]. Vein segments transposed to the arterial circulation for use as bypass grafts are exposed to increased bloodflow and intraluminal pressure[2]. Indeed, Liu and Fung showed that the average circumferential wall stress (CWS) in an arterial vein graft (AVG) immediately upon reestablishing flow could be 140 fold that in a vein under normal circumstances[2]. The tissue often responds to this perceived injury by thickening, which is thought to be an attempt to return the stress to venous levels. However, this response is uncontrolled and can over-compensate, leading to stenosis instead of the desired thickening or “arterialization” of the AVG. It has been suggested that this hyperplastic response by AVGs is a direct result of a “cellular shock” related to their abrupt exposure to the harsh new biomechanical environment[3]. We hypothesize that the adverse hyperplastic response by AVGs may be reduced or eliminated by more gradually exposing them to the arterial biomechanical environment. We believe that an adventitially-placed electrospun polymer wrap will allow an AVG ample opportunity to adapt and remodel to the stresses of its new environment in situ, thereby reducing cellular injury and limiting the initiating mechanisms of IH. Previous work has shown that preventing acute distension of AVGs by adding an external support or sheath can improve various pathologic responses[4, 5], but clinical utility of such a wrap is unproven. Liu et al. used a polytetrafluoroethylene external support to reduce IH in AVGs[2]. However, the immunological response to a permanent wrap is unfavorable and led Vijayan et al. to develop a biodegradable polyglactin sheath. These polyglactin sheaths were loose-fitting and allowed the AVG to expand to their maximum diameters under arterial pressure and thus did not offer mechanical support or prevent increased levels of CWS.

The establishment of a new pathway for blood flow immediately following endovascular aneurysm repair (EVAR) results in morphological changes and remodeling of the aneurismal sac. While EVAR is a minimally invasive surgical intervention, failure of the endovascular graft (EVG) may occur in which there is downstream migration and endoleak formation, creating a repressurization of the aneurismal sac and an increased risk of rupture. While the mechanism of aneurysm rupture and EVG failure is fundamental in nature, the factors that most significantly contribute to the end result are not yet fully understood. Mechanically, both are the consequence of an exerted force or disturbance exceeding the strength of a given material, whether it is the aneurismal arterial wall or the interaction that exists between the graft and wall. Embedded within this causal relationship are the contributions of arterial wall remodeling, intraluminal thrombus formation, and the dynamics that exists within the lumen. Several studies have been performed to examine these factors individually as they affect shear stress, the development of vortices, and the mechanical stress experienced along the arterial wall. However, a complete investigation is needed to study an anatomically realistic geometry operating under physiological conditions. The computational analyses conducted in this investigation address the confluence of these factors as they are modeled within an accurate patient-specific abdominal aortic aneurysm (AAA) reconstructed from CT scan data prior to and after EVAR. Our results verify the pressure-dominated characteristic of the flow and the negligible contribution of the dynamic and frictional force components; both are in good agreement with previously published results for analytical estimation of flow-induced forces in EVGs. [1]

The number of medical images taken has continued to increase year over year for an aging population in the United States. It has been shown that patients understand their diagnoses better when shown a 2D or 3D image of their respective diseases. However, clinicians do not regularly show patients their images as it requires additional time and processing. In this experiment, we demonstrate the use of augmented reality to visualize abdominal aortic aneurysms using a previously developed artificial intelligence engine. Our group further expanded the number of cases used for training the stress prediction model to a total of 274 patients (206 used for training or ~ 5.4 million nodes, and 68 for testing or ~1.8 million nodes). Medical images undergo automated segmentation, and wall stresses are predicted on the 3D surface of aneurysms to view a heat map. The pipeline includes introducing elements into the Microsoft HoloLens 2 ecosystem to view models and additional analytics, enabling clinicians and patients to view the biomechanical status without the need for a computational or imaging expert. The proposed clinical workflow would allow a local server to process medical imaging data, generate point clouds, predict wall stresses on individual points, and create a 3D model with a colormap to view in augmented reality. The study revealed that neural networks and ensemble boosted tress models predicted the wall stresses more accurately (when compared to ground truth finite element analysis results). The approach is not limited to the HoloLens 2 ecosystem but can be used with other emerging augmented or virtual reality hardware systems.