Latest Developments in Adapting Deep Learning for Assessing TAVR Procedures and Outcomes

Abstract

Aortic valve defects are among the most prevalent clinical conditions. A severely damaged or non-functioning aortic valve is commonly replaced with a bioprosthetic heart valve (BHV) via the transcatheter aortic valve replacement (TAVR) procedure. Accurate pre-operative planning is crucial for a successful TAVR outcome. Assessment of computational fluid dynamics (CFD), finite element analysis (FEA), and fluid–solid interaction (FSI) analysis offer a solution that has been increasingly utilized to evaluate BHV mechanics and dynamics. However, the high computational costs and the complex operation of computational modeling hinder its application. Recent advancements in the deep learning (DL) domain can offer a real-time surrogate that can render hemodynamic parameters in a few seconds, thus guiding clinicians to select the optimal treatment option. Herein, we provide a comprehensive review of classical computational modeling approaches, medical imaging, and DL approaches for planning and outcome assessment of TAVR. Particularly, we focus on DL approaches in previous studies, highlighting the utilized datasets, deployed DL models, and achieved results. We emphasize the critical challenges and recommend several future directions for innovative researchers to tackle. Finally, an end-to-end smart DL framework is outlined for real-time assessment and recommendation of the best BHV design for TAVR. Ultimately, deploying such a framework in future studies will support clinicians in minimizing risks during TAVR therapy planning and will help in improving patient care.

TAVI-PREP: A Deep Learning-Based Tool for Automated Measurements Extraction in TAVI Planning

“…The complete workflow can be visualized in Figure 1. For the segmentation data, the public MM-WHS dataset [18][19][20] was used. Training was performed on the publicly available CT scans and 15 in-house manual segmentations provided by an expert cardiologist.…”

“…This work used the publicly available datasets MM-WHS available at: https://zmiclab.github.io/zxh/0/mmwhs/data.html accessed on 31 August 2023 [18][19][20]. The private dataset used to enhance the segmentation performance cannot be publicly shared.…”

“…In the specifics of TAVI procedures and outcomes, numerous deep learning applications have been reported in the literature in recent years [18]. Among them, some works have attempted to automatize the measurements of some of the anatomical structures.…”

Fluid–Structure Interaction Aortic Valve Surgery Simulation: A Review

“…This approach does not take into account structural fields and therefore does not allow the assessment of structures such as valve motion. To perform an accurate dynamic analysis that includes modeling of blood flow during the cardiac cycle in combination with the structural mechanics of the valve, FSI analysis is necessary because it considers both the structural and fluid flow domains [125,126]. The Arbitrary Lagrange-Euler method (ALE) [45], smoothed-particle hydrodynamics (SPH) [127], and the IBM (immersed boundary method), as well as the Lattice Boltzmann Method [48] are currently the three most popular methods for solving FSI problems in hemodynamic studies.…”

“…x FOR PEER REVIEW 9 of 24 FSI analysis is necessary because it considers both the structural and fluid flow domains[125,126] …”