Amir Rouhollahi, PhD

Pronouns

He/Him/His

Job Title

Research Fellow

Academic Rank

Research Fellow

Department

Surgery

Authors

Amir Rouhollahi, James N. Willi, Jonathan Brown, Hoda Javadikasgari, Tsuyoshi Kaneko, Elazer R. Edelman, Farhad R. Nezami

Principal Investigator

Dr. Farhad R. Nezami

Research Category: Digital Health, Imaging, and Informatics

Tags

Leveraging deep learning algorithms for risk stratification and therapy planning in patients with aortic stenosis

Scientific Abstract

Aortic valve stenosis (AS) is the most prevalent valvular heart disease in developed countries. Surgical aortic valve replacement (SAVR), the traditional standard of care for patients with severe symptomatic AS, is increasingly being complemented by transcatheter aortic valve implantation (TAVI). Several challenges (particularly regarding intervention timing and valve durability) need to be addressed before expansion to lower risk and younger patients including non-circular transcatheter valve deployment, crimp-induced leaflet damage, paravalvular leak, thrombosis, conduction abnormalities, prosthesis-patient mismatch, coronary occlusion, and aortic aneurysm. Several in vivo studies have reported the critical role of calcium morphology and distribution on AS severity and therapy outcome. Yet, there are no phenotype-related criteria for calcium as to which approach to use and when, and how, to balance benefits and risks. Abundant, inefficiently harvested information within the imaging data, if leveraged by the state-of-the-art routines of deep learning, could help characterize AS cases individually, predict clinical events, and plan for efficient therapy. In this study, a novel calcification scoring is proposed based on which a deep learning model is being developed to predict the clinical outcomes using the patient-specific CT images.

Lay Abstract

One of the most common heart valve problems in the developed world is aortic Valve Stenosis (AS) which occurs when the valve openings of the heart are thickened with a buildup of calcium. AS can be treated with surgery or a less invasive method called TAVI where a new, mechanical valve is inserted without removing the old one. As with any medical operation, there are risks involved. Many studies show parallels between the shape of the calcium buildup and the events after the TAVI procedure, but there is no tool in place to measure this correlation. We propose that Images from CT scans can be used with the help of computer models to predict what happens to the patient after surgery. Knowing the outcome in advance leads to safer decisions and reduces the risk of adverse side effects after surgery.

Clinical Implications

Utilizing this AI-powered toolbox as an assistive tool for clinical decision-making will enhance therapy. The patient-specific predictions made by this toolbox will greatly help physicians to assess treatment options with more quantitative insight and be aware of the clinical outcomes before the operations.