Abstract / Summary
Object: To test the hypothesis that a reduced-slice geometric model can maintain left ventricular (LV) volumetry, thereby providing a rapid and reliable clinical alternative to conventional full-stack planimetry. Materials and Methods: In a retrospective cohort of 342 patients with full-stack cardiovascular magnetic resonance (CMR), nine geometric combinations (3, 4, or 5 short-axis (SAx) slices and a 2, 3, or 4 chambered long-axis (LAx) slice) were evaluated by decimating the original acquisitions. Volumes were calculated using trapezoidal mid-cavity approximations and a generalized prismoidal model for the basal region. This was compared to the full-stack reference using concordance correlation coefficient (CCC), repeated measures ANOVA, pairwise tests and Bayes factor log10(BF10) analysis. Results: The choice of LAx view influenced the accuracy approximately 50 times more than SAx slice count. For a 2ch + 5 SAx protocol, EF bias was just 0.83% (LoA: -6.18 to 7.84%), with a minimum detectable change (MDC) of 7.01%, compared to 8.7% reported for expert human readers, suggesting strong concordance (CCC>0.90) with full-stack. Bayesian analysis decisively supported 5 over 3 SAx slices (log10(BF10) = 6.42). Conclusion: This proposed model successfully minimizes scan time and breath-holds while maintaining clinical accuracy. Functional metric bias remains lower than scan-rescan variability. Keywords: Left Ventricle; Volumetry; Stroke Volume; Ejection Fraction; Scan Time Reduction; Global Function