Lyon, Ella M.Essafri, IlhamZhang, MengquianLucero, MelanieIchimura, KenzoTruong, UyenTurton, Helena A.Stenmark, Kurt R.Spiekerkoetter, EddaKheyfets, Vitaly O.2025-05-142025-05-142025-04https://hdl.handle.net/11124/180595https://doi.org/10.25676/11124/180595Accurate ventricular volume estimation from cardiac magnetic resonance imaging (CMR) in rodents is often limited by reduced short-axis (SA) coverage due to anesthesia duration, subject health, or slice planning constraints. This study (1) evaluated how reduced SA coverage impacts volume estimates and (2) compared the performance of three reconstruction methods under these constraints: the standard Simpson’s Disk Summation (SD), a short-axis alpha-shape method (SAα), and a combined long-axis plus short-axis alpha-shape method (LA+SAα). The SAα and LA+SAα approaches are novel techniques that use alpha-meshing and Monte Carlo volume estimation to address the geometric limitations of SD. We analyzed CMR data from nine rodents (five Sham, four Pulmonary-Artery-Banded) and compared volume estimates computed with all three methods under full SA coverage (7–8 slices) and reduced coverage (5–6 slices). Incorporating a single long-axis image and mesh-based reconstruction notably improved volume accuracy, particularly for the right ventricle. These findings demonstrate that combining orthogonal imaging views with advanced volumetric estimation enhances the reliability of ventricular volume calculations under constrained imaging conditions. This approach supports more accurate preclinical cardiac assessments and provides a practical solution when full short-axis acquisition is not feasible.posterengCopyright of the original work is retained by the author.Improving CMR ventricular volume estimation in rodents under limited data conditionsText