ECHOCARDIOGRAPHIC ASSESSMENT OF SYSTOLIC AND DIASTOLIC LEFT-VENTRICULAR FUNCTION USING AN AUTOMATIC BOUNDARY DETECTION SYSTEM - CORRELATION WITH ESTABLISHED INVASIVE AND NON INVASIVE PARAMETERS

Systolic and diastolic left ventricular function was assessed using an echocardiographic automatic boundary detection system (ABD) in 50 unselected patients undergoing left cardiac catheterisation. Automatic boundary detection system derived parameters (fractional area change [FAC], peak positive rate of area change [- dA/dt] and peak negative rate of area change [-dA/dt]) were compared with invasively (left ventricular angiography and pressures) and non invasively (Doppler mitral filling velocities and isovolumic relaxation time) acquired conventional indices of ventricular function. Adequate detection of endocardial boundaries and subsequent measurements using the ABD system were achieved in 40/50 (80%) patients in the short axis parasternal view, in 41/50 (82%) in the apical four chamber view and in 34/50 (68%) in both views. For the whole group of patients the FAC (maximal left ventricular diastolic area- minimal left ventricular systolic area - maximal left ventricular diastolic area) estimated in the short axis view correlated with the angiographic ejection fraction (EF) measured in the right oblique projection (r = 0.51, p < 0.001);There was only a weak correlation of the FAC estimated in the apical four chamber view with the EF (r = 0.36, p < 0.01). The mean FAC (mean value of the FAC in the short axis and apical four chamber views) correlated reasonably with the EF (r = 0.62, p < 0.0001). There was no correlation between ABD derived parameters and left ventricular end diastolic pressure (LVEDP) in these patients. In a subgroup of patients with normal coronary arteries and left ventricular function (n = 17), although there was no correlation between EF and FAC, there was a strong positive correlation between FAC (apical four chamber and mean) and LVEDP (r = 0.77, p < 0.01 and r = 0.87, p < 0.01 respectively). No correlation was found in these patients between EF and LVEDP. In a further subgroup of patients with angiographically abnormal left ventricular function (EF < 45%), there was a positive correlation between FAC (short axis, apical four chamber and mean) and EF (r = 0.52, p < 0.05, r = 0.83, p < 0.0001 and r = 0.80, p < 0.001 respectively) and a negative correlation between FAC (short axis and mean) and LVEDP (r = -0.52, p < 0.05 and r = -0.60, p < 0.01 respectively). There was also a negative correlation between LVEDP and EF in the same subgroup of patients (r = -0.65, p < 0.01). None of the ABD derived parameters correlated with non invasively acquired indices of diastolic ventricular function (peak early left ventricular diastolic filling blood velocity [Emax], peak late diastolic velocity [Amax], E/A ratio and isovolumic relaxation time [IVRT], but there was a consistent positive correlation between - dP/dt and + dA/dt estimated in the four chamber view (r = 0.5, p < 0.01, all patients). Therefore, although ABD derived parameters cannot be used in an interchangeable way with ejection fraction, they do provide a rapid, bedside method for the assessment of left ventricular function. FAC and dA/dt do appear to reflect left ventricular performance both in patients with normal ventricles and in patients with impaired left ventricular function.