An abdominal aortic aneurysm is the most commonly diagnosed arterial aneurysm that is an abnormal bulge on the aorta associated with the gradual thinning of the vessel wall. With mortality as high as 80% in cases of ruptures, abdominal aortic aneurysm accounts for more than 10,000 deaths in the United States every year. One of the most common treatments is endovascular aneurysm repair, which redirects blood flow away from the aortic wall and bypasses the weak spots by implanting a covered stent graft in the aneurysm sac via a minimally invasive procedure. Among the stent recipients, 30% of them can experience persistent blood flow into the aneurysm sac, called ‘endoleak,’ leading to aneurysm expansion and rupture. Thus, blood pressure and flow near the stent should be periodically monitored. However, the commonly used imaging technique is highly dependent on patient compliance, and its repeatedly administrated iodinated contrast poses a risk of chronic kidney disease. As such, the overall objective of this research is to create a Smart Stent based on a flexible and battery-less membrane-based sensor and wireless bioelectronics with a deep-learning algorithm to realize automated diagnosis of endoleak. This collaborative research will integrate the scientific findings and discoveries with educational venues for students across disciplines (Electrical and Mechanical Engineering), generations (K-12 to lifelong learners), and two institutes (Temple University and Kansas State University).
The overall objective of this research is to develop a Smart Stent for post-endovascular aneurysm repair surveillance that combines a flexible, and battery-less bioelectronic system with a deep-learning algorithm to realize automated diagnosis of endoleak. The central hypothesis is that Smart Stent, created by conformally weaving piezoelectric porous membrane sensors inside and outside of the conventional stent graft, will bring a novel electromechanical wireless biotelemetry scheme whose sensor data can be directly analyzed by a deep-learning model for classification of complex hemodynamics. The intellectual merits of the proposed research include 1) design of an auxetic porous piezoelectric membrane for the Smart Stent that is optimized for the multi-modal sensing of blood pressure and flow, 2) microfabrication of complex 3D structure and surface-integrated micro coils for near-field magnetic induction communication, 3) a novel electromechanical interrogation scheme that converts energy from physiological information (e.g., blood pressure and flow) into wireless magnetic induction signals, 4) a comprehensive evaluation using a precise aneurysm phantom model to build baseline data for cardiovascular research, and 5) a deep learning-enabled sensing classification algorithm that offers real-time, quantitative, and automated assessment of five different types of endoleak. The research will establish a machine learning-enabled wireless sensing system that will spur new theory and understanding for the next generation implantable biomedical system.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
