Accuracy of Devereux and Teichholz formulas for left ventricular mass calculation in different geometric patterns: comparison with cardiac magnetic resonance imaging

Figures

• 

  Figure 1. Short historical overview of the LV mass calculation (upper panel). Exemplary measurement of LV EDD, IVS and PW in the parasternal long axis of an Echo. The formula and corresponding myocardial mass calculations for the DEV and TEICH method as well as the results from CMR imaging are shown below, for the same person (lower panel).

• 

  Figure 2. The contour plot represents the differences (Δ) of modeled LV mass between the DEV and TEICH equation in grams (g), elaborating the systematic overestimation of the DEV equation. Mathematical calculations are based on incremental continuous increase of LV EDD and of mean wall thickness [(IVS + PW)/2].

• 

  Figure 3. Correlation plots (left panels) with line of identity (dotted) and Bland–Altman plots (right panels) with mean bias (dashed) and limits of agreement (dotted horizontals) for the DEV and TEICH method with CMR as the reference imaging in the derivation cohort (n = 1276).

• 

  Figure 4. Multivariable regression fits of the inter-method bias of LV mass index (g/m2) as the dependent variable for the DEV (solid) and TEICH (dashed) method vs. CMR imaging as the reference method for mean wall thickness, LV EDD, age and sex as the independent variables in the derivation cohort (n = 1276).

• 

  Figure 5. Subgroup analyses represented by boxplots showing the measurement differences of LV mass (Δ) for the derivation cohort (n = 1276, top row) and the validation cohort (n = 226, bottom row), respectively, as well as for the DEV (left panels) and the TEICH method (right panels), and illustrating the systematic overestimation of LV mass with the DEV equation compared to CMR as the reference imaging method for each tier of LV geometry.