Limitations
As with most imaging studies, our study is limited by inter-observer
bias. Interestingly, our MDCT readers were less experienced than the TEE
readers and still were able to generate similar measurements. Despite
our high-quality scanner there was sometimes systolic artifact on MDCT
which may reduce accuracy. Additionally, in cases where leaflet
pathology involves more than one area, interpretation on MDCT may be
limited. Since obtaining preoperative MDCT is not standard practice for
DMR we had a limited number of patients included and the potential for
selection bias of those included. Most commonly MDCT was obtained in
patients with complex TEE anatomy to further assess the mitral valve
apparatus. Lastly, the imaging was performed on different days and the
hemodynamics may vary among TEE (done under sedation) and MDCT
(performed with IV contrast volume load).