Clinical Electrophysiology -> Atrial Fibrillation & Atrial Flutter: -> Left Atrial Appendage Closure D-PO02 - Poster Session II (ID 47) Poster

D-PO02-209 - Virtual Reality Cardiac Segmentation For Prediction Of Left Atrial Appendage Occluder Device Size (ID 1061)


Background: Transoesophageal echocardiography (TEE) & multidetector computed tomography (CT) currently serve for left atrial appendage (LAA) occlusion planning.
Objective: The study objective was to assess the feasibility of CT scan based virtual reality (VR) segmentation for LAA device size prediction.
Methods: LAA occlusion was performed according to clinical need. Dedicated ECG gated CT (IV contrast, systolic phases of the R-R interval (30-40%)) prior to the procedure, VR segmentation (based on the performed CT for each patient) and TEE (during the procedure) were utilized for LAA evaluation in all patients and compared to the inserted device size. LAA maximal and minimal diameters and orifice perimeter were measured and calculated for each modality (CT, VR, TEE). LAA perimeter was calculated using the Ramanujan approximation / π . Statistical analysis included: Lin’s Concordance Correlation Coefficient (ρc), average bias (AB) in millimeters (mm) and the mean square error (MSE) in mm2 were calculated.
Results: The study cohort included 14 patients which underwent LAA occluder insertion (Amplatzer N=5, Watchman N=9). CT, VR and TEE maximal diameter ρc were 0.506 (95% CI 0, 0.80), 0.78 (95% CI 0.5, 0.92) and 0.65 (95% CI 0.38, 0.82), respectively. CT, VR and TEE maximal diameter AB were -0.72 ±4.73 mm, 1.23±2.64 mm and 4.07±1.97 mm, respectively. CT, VR and TEE maximal diameter MSE were 22.91 mm2, was 8.52 mm2 and 20.49 mm2. CT, VR and TEE calculated perimeter ρc were 0.44 (95% CI, -0.02, 0.747), 0.53 (95% CI, 0.19, 0.76) and 0.42 (95% CI, 0.18, 0.62), respectively. The CT VR and TEE calculated perimeter AB were 1.7± 4.32 mm, 3.07±3.24 mm and 5.85±2.21 mm, respectively. CT VR and TEE calculated perimeter MSE were 21.58 mm2, 19.22 mm2, and 38.64 mm2, respectively.
Conclusion: The maximal LAA diameter measured by VR proved to have the highest concordance (ρc) and the lowest mean square error when compared to the inserted device size. The calculated CT and VR LAA perimeter also demonstrated relative correlation (although lower) with the device size. VR opens new exiting imaging frontiers for cardiac non-invasive procedures. VR may be used for better and more accurate pre-procedural planning.