Automated detection of aortic valve closure in apical tissue doppler images

Abstract

Ultrasound tissue Doppler imaging (TDI) can be used to measure velocities of moving cardiac tissue during the cardiac cycle. Aortic valve closure (AVC) can be seen as a notch occurring after ejection, but before early relaxation in velocity/time curves from apical TDI images of the base of the left ventricle. The timing of AVC may be determined by manually looking for this event. An automated algorithm first detecting the timing of early relaxation and mitral valve opening, is however also able to determine the timing of AVC by searching in both space and time. The automated algorithm was tested on the apical four-chamber, two-chamber and long axis views of 16 healthy subjects. In 88% of the cine-loops the algorithm estimated the timing of AVC within 20 msec off the start of the second heart sound as visible in simultaneously recorded calibrated phonocardiograms. Automated detection of AVC might save manual effort, and provide a marker separating ejection and diastole for further automated analysis.