Signal averaging helps reliable noninvasive monitoring of left ventricular dimensions based on acoustic quantification

The acoustic quantification (AQ) technique for on-line endocardial border detection provides continuous left ventricular (LV) area or volume signals, which may allow noninvasive quantification of valuable clinical indices of LV function. However, these signals are distorted by noise, so that accurate interpretation on a beat-to-beat basis is not possible. An automated signal averaging algorithm was developed based on the stochastic sign change as a criterion for temporal beat alignment and rejection of artifacts, in order to improve noise characteristics of the AQ signals. The algorithm was evaluated using area signals obtained from 8 healthy volunteers. The estimates of a set of systolic and diastolic LV function parameters obtained using this algorithm were found to be highly consistent. This approach provides a fast and easy method for noise reduction in AQ signals, which may significantly improve the accuracy of this new noninvasive method of assessment of LV function based on ultrasound technology