Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article


Master of Science




Dr. Michael W.A. Chu


Purpose: Multi-phase computed tomography angiography (CTA) for the pre-procedural planning of TAVR presents a unique opportunity to assess 3D myocardial biomechanics. This study aimed to assess the feasibility and predictive utility of 3D myocardial deformation analysis (3D-MDA) based principal strain to predict heart failure or death following TAVR using pre-procedural, multi-phase computed tomography angiography (CTA) datasets.

Methods: Two hundred and five patients undergoing pre-TAVR multi-phase gated CTA followed by successful TAVR were retrospectively identified. Whole heart 3D mesh chamber models were generated followed by 3D-MDA of the left ventricle (LV) to determine global LV minimum principal strain (minPS) for endocardial, epicardial and transmural layers.

Results: Of the 205 patients, 196 (96%) had analyzable CTA data for 3D-MDA [median (IQR) age of 85 (79.588) years (55% male); STS-PROM score: 3.10 (2.104.55); and echocardiographic LVEF 60.0 (55.965.0)%]. At a median 25 (11–36) months following TAVR, 55 patients (28%) experienced a composite clinical outcome of heart failure hospitalization or death. Patients with a global minPS below a 23.7% experienced a 3-fold higher rate of the primary outcome (p

Conclusions: CT derived 3D-MDA is feasible and delivers novel deformation markers strongly and independently predictive of future cardiovascular outcomes in patients undergoing TAVR.

Summary for Lay Audience

Transcatheter aortic valve replacement (TAVR) represents one of the most impactful technical advancements for patients with aortic valve disease. As a minimally invasive therapeutic alternative to open heart surgery, this technique has established an important role in the management of severe aortic stenosis: being initially reserved only for high-risk elderly patients unable to undergo surgical care, and now expanding to meet the needs of broader referral populations. This procedure inherently developed early dependency on advanced imaging for pre-procedural planning and intra-procedural guidance given need to select delivery paths, feasibility and to place valve devices without direct visualization.

This as a result has led to the ubiquitous use of multi-phase, ECG-gated computed tomography for pre-procedural evaluation in these patients. To date this imaging data has been used to confirm procedural eligibility and determine optimal deployment strategies (sizing and position) for the valve device. Despite this, tremendous value from these imaging datasets is currently disregarded. Our team is researching if images routinely obtained from CT scans prior to TAVR can be used to construct 3D “beating heart” models that predict benefit from this procedure. Specifically, our aim is to predict from these images how much undergoing a TAVR procedure will reduce the chances that patients experience future hospitalization or death.

In the second part of this research study, we are asking twenty patients referred for TAVR if they would agree to undergo an additional imaging test called a cardiac MRI (magnetic resonance imaging). This non-invasive test can assess the heart muscle’s health by measuring how much scarring or “fibrosis” is present, an important influence on how the heart will function following TAVR. By measuring this we can better understand how information from the CT scan is allowing artificial intelligence to predict future risk of heart failure and death, an important part of building trust in this technology.