Insights into diastolic function analyses using cardiac magnetic resonance imaging: impact of trabeculae and papillary muscles

Figures

• 

  Fig. 1. LV volumetry using CMRtools. The longitudinal LV main axis (black line in A) is defined from three representative short axis slices. Red points along the longitudinal main axis (B) illustrate localization of short axis slices where the longitudinal axis was set (diastole and systole, respectively). The LV endocardial contour (yellow circumscript in C) is set by manual definition of endocardial landmarks (red points in C) and propagating (yellow points along the endocardium in C) in a diastolic (shown) and systolic mid-papillary short axis slice. This contour is propagated for the other cardiac phases and the short axis stack. The resulting LV blood volume (including TPM) is visualized in long axis view (red color in D), and endocardial contour can be controlled (green circles in D). Exclusion of TPM (yellow in E) from the LV blood pool (red in E) is automatically provided by the software. Definition of mitral valve plane is done in 2- and 4-chamber views in diastole and systole (red line with arrows in F, left) to separate blood volumes of the atrium and ventricle. Mitral valve plane is shown (red rectangle in F, right) in a 3D geometric model of the LV. A third cardiac phase…

• 

  Fig. 2. Temporal LV volume and derived diastolic parameters in a healthy volunteer. A Temporal LV volume (squares) by inclusion of TPM into the LV myocardium and derived diastolic parameters (LVpreA, DVR). B Temporally differentiated LV volume (thin line) and after smoothing (circles, dashed line) with derived diastolic early and atrial peak filling rate (EPFR and APFR) indicated by arrows. Early (EFV) and atrial (AFV) peak filling volumes are determined as areas from a Gaussian dual-peak fit (grey, red, and blue line). C Disabling the tissue-blood threshold (= inclusion of TPM into the blood pool) leads to overlapping early and atrial peaks

• 

  Fig. 3. Exponential relationship of early to atrial ratios (E/A) as function of age in healthy volunteers. Gaussian fitted filling volumes (FVR, solid line, 95% confidence interval by dashed lines), and peak filling rates (PFRR, red line) were determined by (A) inclusion of trabeculae and papillary muscles into the myocardium (TPM-myo) and (B) into the blood pool (TPM-blood). With TPM-myo, coefficients of determination were significantly higher for FVR (r2 = 0.89 vs. r2 = 0.66) and PFRR (r2 = 0.85 vs. r2 = 0.56). Intra-individual follow-up measurements (n = 2–4) in four subjects are connected by trend lines