New Indexes on Heart Motion Evaluation

Abstract

Our research group has developed the tracking method of regional myocardium from ultrasonic radio frequency (RF) signal. The tracking method employs a hierarchical correlation method. In addition, we regarded the correlation coefficient as the confidence coefficient. If it is not high, the tracking position is corrected by neighbor points. The tracking method can detect the movement of myocardium very well if the RF signal is clear. In this research, we investigate the moving distance based on the tracking position that includes less error. As a result, the contributions of the distance are different between normal example and disorder example. We propose three indexes defined by amounts of statistics for the evaluation of a heart motion. The error rates for diagnosis with the proposed index (FP=4.8[%], FN= 9.5[%]) is lower than that with thickening rate (FP=13.6[%], FN=14.3[%]).