Catalyst Awardee

Project Description

Application of Physics-Informed Neural Network and Organoid Vascular Model in Thrombotic Risk Evaluation in Abdominal Aortic Aneurysm

Xiaoning SUN, MD |Chinese Academy of Medical Sciences and Peking Union Medical College; Siting LI, MD; Yuehong ZHENG, MD; Zijian WANG, BS; Zenghan CAO, MD Candidate
Competition Sponsor: Chinese Academy of Medical Sciences
Awardee Year: 2024

The incidence rate of abdominal aortic aneurysms is increasing in recent years. Intra-luminal thrombus (ILT) in abdominal aortic aneurysms (AAAs) is associated with aneurysm progression, rupture, and peri-operational embolic complications. Acute in-stent thrombus formation after endovascular aortic repair (EVAR) is a catastrophic threat to the affected limb and/or viscera. Conventionally, the management strategy of AAA is simply based on AAA diameter and/or subjective empirical evaluation of clinicians. Individualized thrombosis assessment are urgently needed to evaluate postoperative outcomes. Computational fluid dynamics (CFD) simulation is capable of individualized assessment of aortic lesions, but due to 1) the complex workflow and tremendous computational burden, and 2) the simulation results yet lack the support of basic experiments, aortic blood-flow simulation has not been widely used in clinical practice. To address the above issues, this project plans to 1) construct a clinical-imaging database of AAA; 2) use physics-informed neural network (PINN) to foster a real-time aortic blood-flow field simulation tool, and to calibrate the model with CFD analysis; 3) apply micro-fluidic vascular microchips to validate the mechanism of thrombus formation under individualized hemodynamic environment in vitro; (4) construct a real-time intraoperative thrombosis risk prediction model by integrating PINN flow analysis, branch vessel flow conditions, hemodynamic thrombotic indicators, and real-world clinical outcomes. This project provides a new perspective for the assessment of thrombosis-related complications in AAA. As the major product to be expected, the surgical simulation software and algorithms have valuable potential in aiming clinical decision-making and promoting the innovation and transformation of surgical instruments.

Sign up for updates