Automated MSCT Analysis for Planning Left Atrial Appendage Occlusion Using Artificial Intelligence.

Background: The number of multislice computed tomography (MSCT) analyses performed for planning structural heart interventions is rapidly increasing. Further automation is required to save time, increase standardization, and reduce the learning curve. Objective: The purpose of this study was to investigate the feasibility of a fully automated artificial intelligence (AI)-based MSCT analysis for planning structural heart interventions, focusing on left atrial appendage occlusion (LAAO) as the selected use case. Methods: Different deep learning models were trained, validated, and tested using a cohort of 583 patients for which manually annotated data were available. These models were used independently or in combination to detect the anatomical ostium, the landing zone, the mitral valve annulus, and the fossa ovalis and to segment the left atrium (LA) and left atrial appendage (LAA). The accuracy of the models was evaluated through comparison with the manually annotated data. Results: The automated analysis was performed on 25 randomly selected patients of the test cohort. The results were compared to the manually identified landmarks. The predicted segmentation of the LA(A) was similar to the manual segmentation (dice score of 0.94 ± 0.02). The difference between the automatically predicted and manually measured perimeter-based diameter was -0.8 ± 1.3 mm (anatomical ostium), -1.0 ± 1.5 mm (Amulet landing zone), and -0.1 ± 1.3 mm (Watchman FLX landing zone), which is similar to the operator variability on these measurements. Finally, the detected mitral valve annulus and fossa ovalis were close to the manual detection of these landmarks, as shown by the Hausdorff distance (3.9 ± 1.2 mm and 4.8 ± 1.8 mm, respectively). The average runtime of the complete workflow, including data pre- and postprocessing, was 57.5 ± 34.5 seconds. Conclusions: A fast and accurate AI-based workflow is proposed to automatically analyze MSCT images for planning LAAO. The approach, which can be easily extended toward other structural heart interventions, may help to handle the rapidly increasing volumes of patients.

FEops HEARTguide Patient-Specific Computational Simulations for WATCHMAN FLX Left Atrial Appendage Closure: A Retrospective Study.

Background: Three-dimensional transesophageal echocardiography (3D-TEE) is the primary imaging tool for left atrial appendage closure planning. The utility of cardiac computed tomography angiography (CCTA) and patient-specific computational models is unknown. Objectives: The purpose of this study was to evaluate the accuracy of the FEops HEARTguide patient-specific computational modeling in predicting appropriate device size, location, and compression of the WATCHMAN FLX compared to intraprocedural 3D-TEE. Methods: Patients with both preprocedural and postprocedural CCTA and 3D-TEE imaging of the LAA who received a WATCHMAN FLX left atrial appendage closure device were studied (n = 22). The FEops HEARTguide platform used baseline CCTA imaging to generate a prediction of device size(s), device position(s), and device dimensions. Blinded (without knowledge of implanted device size/position) and unblinded (implant device size/position disclosed) simulations were evaluated. Results: In 16 (72.7%) patients, the blind simulation predicted the final implanted device size. In these patients, the 3D-TEE measurements were not significantly different and had excellent correlation (Pearson correlation coefficient (r) ≥ 0.90). No patients had peridevice leak after device implant. In the 6 patients for whom the model did not predict the implanted device size, a larger device size was ultimately implanted as per operator preference. The model measurements of the unblinded patients demonstrated excellent correlation with 3D-TEE. Conclusions: This is the first study to demonstrate that the FEops HEARTguide model accurately predicts WATCHMAN FLX device implantation characteristics. Future studies are needed to evaluate if computational modeling can improve confidence in sizing, positioning, and compression of the device without compromising technical success.

Validation of a computational model aiming to optimize preprocedural planning in percutaneous left atrial appendage closure.

BACKGROUND: Percutaneous left atrial appendage (LAA) closure can be optimised through diligent preprocedural planning. Cardiac computational tomography (CCT) is increasingly recognised as a valuable tool in this process. A CCT-based computational model (FEops HEARTguide™, Belgium) has been developed to simulate the deployment of the two most commonly used LAA closure devices into patient-specific LAA anatomies. OBJECTIVE: The aim of this study was to validate this computational model based on real-life percutaneous LAA closure procedures and post-procedural CCT imaging. METHODS: Thirty patients having undergone LAA closure (Amulet™ n = 15, Watchman™ n = 15) and having a pre- and post-procedural CCT-scan were selected for this validation study. Virtually implanted devices were directly compared to actual implants for device frame deformation and LAA wall apposition. RESULTS: The coefficient of determination (R2) and the difference in measurements between model and actual device (area, perimeter, minimum diameter, maximum diameter) were ≥0.91 and ≤ 5%, respectively. For both device types, the correlation coefficient between predicted and observed measurements was higher than 0.90. Furthermore, predicted device apposition correlated well with observed leaks based on post-procedural CCT. CONCLUSION: Computational modelling accurately predicts LAA closure device deformation and apposition and may therefore potentiate more accurate LAA closure device sizing and better preprocedural planning.

Patient-Specific Computer Simulation of Transcatheter Aortic Valve Replacement in Bicuspid Aortic Valve Morphology.

BACKGROUND: A patient-specific computer simulation of transcatheter aortic valve replacement (TAVR) in tricuspid aortic valve has been developed, which can predict paravalvular regurgitation and conduction disturbance. We wished to validate a patient-specific computer simulation of TAVR in bicuspid aortic valve and to determine whether patient-specific transcatheter heart valve (THV) sizing and positioning might improve clinical outcomes. METHODS: A retrospective study was performed on TAVR in bicuspid aortic valve patients that had both pre- and postprocedural computed tomography imaging. Preprocedural computed tomography imaging was used to create finite element models of the aortic root. Finite element analysis and computational fluid dynamics was performed. The simulation output was compared with postprocedural computed tomography imaging, cineangiography, echocardiography, and electrocardiograms. For each patient, multiple simulations were performed, to identify an optimal THV size and position for the patient's specific anatomic characteristics. RESULTS: A total of 37 patients were included in the study. The simulations accurately predicted the THV frame deformation (minimum-diameter intraclass correlation coefficient, 0.84; maximum-diameter intraclass correlation coefficient, 0.88; perimeter intraclass correlation coefficient, 0.91; area intraclass correlation coefficient, 0.91), more than mild paravalvular regurgitation (area under the receiver operating characteristic curve, 0.86) and major conduction abnormalities (new left bundle branch block or high-degree atrioventricular block; area under the receiver operating characteristic curve, 0.88). When compared with the implanted THV size and implant depth, optimal patient-specific THV sizing and positioning reduced simulation-predicted paravalvular regurgitation and markers of conduction disturbance. CONCLUSIONS: Patient-specific computer simulation of TAVR in bicuspid aortic valve may predict the development of important clinical outcomes, such as paravalvular regurgitation and conduction abnormalities. Patient-specific THV sizing and positioning may improve clinical outcomes of TAVR in bicuspid aortic valve.

Fluid-Structure Interaction Simulation of Prosthetic Aortic Valves: Comparison between Immersed Boundary and Arbitrary Lagrangian-Eulerian Techniques for the Mesh Representation.

In recent years the role of FSI (fluid-structure interaction) simulations in the analysis of the fluid-mechanics of heart valves is becoming more and more important, being able to capture the interaction between the blood and both the surrounding biological tissues and the valve itself. When setting up an FSI simulation, several choices have to be made to select the most suitable approach for the case of interest: in particular, to simulate flexible leaflet cardiac valves, the type of discretization of the fluid domain is crucial, which can be described with an ALE (Arbitrary Lagrangian-Eulerian) or an Eulerian formulation. The majority of the reported 3D heart valve FSI simulations are performed with the Eulerian formulation, allowing for large deformations of the domains without compromising the quality of the fluid grid. Nevertheless, it is known that the ALE-FSI approach guarantees more accurate results at the interface between the solid and the fluid. The goal of this paper is to describe the same aortic valve model in the two cases, comparing the performances of an ALE-based FSI solution and an Eulerian-based FSI approach. After a first simplified 2D case, the aortic geometry was considered in a full 3D set-up. The model was kept as similar as possible in the two settings, to better compare the simulations' outcomes. Although for the 2D case the differences were unsubstantial, in our experience the performance of a full 3D ALE-FSI simulation was significantly limited by the technical problems and requirements inherent to the ALE formulation, mainly related to the mesh motion and deformation of the fluid domain. As a secondary outcome of this work, it is important to point out that the choice of the solver also influenced the reliability of the final results.