In order to extract quantitative physiological information from cardiac Doppler velocity data, precise contour determination is required. We have previously shown that computation of Parametrized Diastolic Filling (PDF) parameters  from the clinical diastolic Doppler velocity profile (DVP) can be accomplished by manual comparison of PDF model predicted to clinically recorded DVP. However, visual DVP contour estimation as well as comparison of PDF model prediction to the DVP requires hand tracing, is inherently subjective and is prone to interobserver variations. Using the cardiac Doppler ultrasound acoustic signal we developed a computer based method of DVP computation. When applied to the E wave of the human diastolic transmitral DVP, our method allowed automated computation of PDF parameters, equivalent to automated solution of the 'inverse' problem of diastole.