Eun-Ah Park MD, Abstract Co-Author: Nothing to Disclose

Whal Lee MD, Presenter: Nothing to Disclose

Hyung-Kwan Kim, Abstract Co-Author: Nothing to Disclose

Jin Wook Chung MD, Abstract Co-Author: Nothing to Disclose

Jae Hyung Park MD, Abstract Co-Author: Nothing to Disclose

To evaluate the accuracy and reproducibility of three different methods of cavity delineation in patients with hypertrophic cardiomyopathy (HCM), relative to trabeculations and papillary muscles in cardiovascular magnetic resonance (CMR) ventricular volume analysis.

Twenty patients with HCM underwent 1.5-T CMR. Endocardial and epicardial contours were traced manually using three methods: with left ventricular (LV) trabeculae included in the cavity volume (methods A and B), or excluded (C). In terms of methods A and B, an imaginary endocardial border was drawn to include the observed trabecular boundary seen on short-axis cine images (A) or based solely on the cavity and immediate trabeculae (B) at end-systole (ES) phase. Ascending aorta effective flow (AAEF) was measured as a reference of LV stroke volume (SV). We compared ejection fraction (EF) and myocardial mass (MM) among the three methods. Interobserver reproducibility was assessed using intraclass correlation coefficients (ICCs).

Compared to AAEF, LV SV calculated in all methods were significantly larger: mean difference, 7.7 ± 12.2 mL for A (p=0.01), 25.7 ± 17.8 mL for B (p<0.001), and 7.7 ± 14.5 mL for C (p=0.03), and methods A and C were more accurate than B (p<0.001). EF was 55.7% ± 6.9 for A, 68.6% ± 8.4 for B, and 71.7% ± 7.0 for C and showed significant differences in each method (p<0.001). Mean MM was also significantly different between all methods: 164.6 ± 47.4 g/m2 for A, 176.5 ± 50.5 g/m2 for B, and 199.6 ± 53.2 g/m2 for C (p<0.001). There was no difference in MM between ES and end-diastole phases in methods A and C: mean difference, 3.6 ± 8.4 g/m2 for A (p=0.1), 22.3 ± 13.6 g/m2 for B (p<0.001), and 2.8 ± 9.0 g/m2 for C (p=0.5). Method C showed excellent interobserver agreement: ICCs for SV were 0.86 for A, 0.84 for B, and 0.90 for C; ICCs for MM were 0.97 for A, 0.96 for B, and 0.98 for C.

The method of LV cavity delineation significantly affects volume and MM measurements in patients with HCM; exclusion of trabeculae from the LV cavity volume provides the highest accuracy and reproducibility of LV analysis.

The method of LV cavity delineation used in patients with HCM affects the results of CMR ventricular volume and myocardial mass analysis during data Interpretation and longitudinal follow-up.

Park, E, Lee, W, Kim, H, Chung, J, Park, J, Determination of the Endocardial Border in Hypertrophic Cardiomyopathy: Effect of Trabeculae on Evaluating the Left Ventricle on Cardiac MRI.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11002341.html