SnapShot Freeze 2.0, a second-generation motion-correction algorithm, improves mitral valve image quality in preoperative cardiac computed tomography for mitral regurgitation

Preoperative cardiac CT for mitral valve assessment is crucial, and improvement of image quality using motion-correction algorithms remains unreported. We retrospectively analyzed 141 datasets from 47 patients with mitral regurgitation, reconstructing images acquired using a CT scanner with 16-cm z-axis coverage with prospective electrocardiography gating utilizing standard (STD), SnapShot Freeze 1.0 (SSF), and SnapShot Freeze 2.0 (SSF2) algorithms. Two radiologists blinded to reconstruction methods, cardiac cycle, and patient information evaluated image quality for mitral valve leaflets (A1-P3) using a five-point Likert scale, assessing 20 cardiac phases for the A2/P2 and systolic phases for all leaflets. Image quality scored significantly higher using SSF2 compared to STD or SSF reconstruction for all 20 cardiac phases of the A2/P2 and the systolic phase of all leaflets (P < 0.05). STD and SSF scores did not differ significantly but were higher in atrial fibrillation than sinus rhythm, a difference not observed with SSF2. No reconstruction method showed significant correlation between heart rate and image quality. The mean systolic phase contributing most to mitral regurgitation diagnosis was 27.7% of the R-R interval. Interobserver agreement was excellent (weighted kappa = 0.956 for A2/P2; 0.915 for A1/P1, A2/P2, and A3/P3). Mitral valve image quality in preoperative cardiac CT for mitral regurgitation improved significantly using SSF2 reconstruction compared to STD and SSF. Reflected across different cardiac phases and influenced by neither heart rhythm nor rate, findings suggest the robustness of SSF2 for use in various clinical scenarios and to aid surgical planning.