Model-based image processing derived parameters of ventricular filling: can they predict exercise capacity in subjects with chronic heart failure?

Model-based image processing (MBIP) of Doppler echocardiographic transmitral flow (E-waves) has been validated as a method of quantitative diastolic function (DF) assessment. MBIP incorporates the mechanical suction-pump role of the heart, uses the E-wave as input, solves the ´inverse problem´ of diastole and generates three unique parameters (x_o, c, k) for each E-wave. The model´s spring constant k is the analogue of (average) chamber stiffness (ΔP/ΔV). Exercising subjects with chronic heart failure (CHF) attaining an oxygen consumption peak VO₂ ≤ 14 ml/kg/min are likely to benefit from transplantation whereas those attaining peak VO₂ > 14 ml/kg/min do not. The relationship between peak VO₂ and DF has not been determined in CHF. Doppler E-waves of 31 pre-transplant subjects were analyzed using MBIP. Least squares linear best tit determined the k vs. peak VO₂ relation. For subjects with VO₂ ≤ 14 ml/kg/min (n = 12) k was linearly proportional to peak VO₂ with r = 0.57. We conclude that: k is inversely correlated with peak VO₂; a clear delineation exists for k at a VO₂ ≤ or > 14 ml/kg/min. These results show that the stiffer the chamber the worse the exercise tolerance, and MBIP facilitates quantitative DF determination in subjects with CHF.