Real-time 3D echocardiographic transilluminated imaging combined with artificially intelligent left atrial appendage measurement for atrial fibrillation interventional procedures

Abstract

Aims: This study investigated the feasibility and accuracy of real-time three-dimensional (3D) echocardiographic transilluminated imaging (TrueVue Glass) in left atrial appendage (LAA) anatomical morphology and artificial intelligence (AI)-assisted 3D automated LAA measurement (3D Auto LAA) software in the preoperative evaluation of LAA occlusion (LAAO) in patients with atrial fibrillation (AF).Method and results: Thirty-seven patients with AF were selected. Two-dimensional (2D) and real-time 3D transesophageal echocardiography (RT3D-TEE) were performed preoperatively, using conventional 3D, the new 3D TrueVue Glass mode, and cardiac computed tomography angiography (CCTA) to assess and type the morphology of LAA. Physiological parameters were measured using traditional 2D and 3D manual (3D Manual LAA), 3D Auto LAA, and CCTA. TrueVue Glass for LAA outer contour display was compared with CCTA. Comparisons were based on correlation and consistency in measuring the maximum diameter (LZ max), minimum diameter (LZ min), area (LZ area), and circumference (LZ cir) of LAA landing zone (LZ). Times and variabilities were compared. The concordance rate for external shape of LAA was 97.14% between TrueVue Glass and CCTA. 3D Auto LAA and 3D Manual LAA have a stronger correlation and higher consistency in all parameters. 3D Auto LAA showed higher intra- and interobserver reproducibility and allowed quicker analysis (p < 0.05). LAAO was performed in 35 patients (94.59%), and none of which had serious adverse events.Conclusion: TrueVue Glass is the first non-invasive and radiation-free visualization of the overall external contour of LAA and its adjacent structures. 3D Auto LAA simplifies the measurement, making the preoperative assessment more efficient and convenient while ensuring the accuracy and reproducibility. A combination of the two is feasible for accurate and rapid assessment of LAA anatomy and physiology in AF patients and has practical application in LAAO.