Interaction between respiration and beat-by-beat ventricular parameters from acoustic quantification

In 8 patients with cardiac dilative myopathy and pseudo-normalized mitral Doppler pattern (DP, age 60/spl plusmn/2, mean/spl plusmn/SE), and in 11 age-matched control subjects (NL, age 61/spl plusmn/4), beat-by-beat end diastolic (EDV), end systolic (ESV), stroke (SV) ventricular volume (VV) measurements were obtained at rest for periods of at least 4 min by echocardiographic acoustic quantification (AQ) signal. RR intervals and respiration were recorded as well. In all subjects, spectral analysis showed a prevalence of high frequency (HF) components in all VV series, with an increased absolute power in DP (53/spl plusmn/12 ml/sup 2/ of HF SV variability compared to 16/spl plusmn/5 in NL). Folded scatter diagrams of the VV values vs the respiratory phase revealed consistent patterns. Parameters of a linear prediction model of SV were estimated by considering as inputs:1) three samples of respiration; 2) the previous sample of EDV slow trends, 3) the previous RR interval with a goodness of fit of 47/spl plusmn/4% in NL and 43/spl plusmn/6% in DP.