A Two-Dimensional CVIB Imaging System with a Snake-Based Tracking Algorithm

摘要

Quantitative ultrasound tissue characterization based on integrated backscatter (IB) has shown great potential in detecting myocardial ischemia. The magnitude of the cyclic variation in IB (CVIB) has recently been considered as one of the most promising parameters in assessing regional myocardial contractile performance. Our laboratory previously developed a novel ultrasonic fusion imaging method based on the CVIB. However, the major problem for clinical applications of this technique and other existing analytical methods based on IB is that the myocardial tissue can''t be traced effectively without the cardiologist''s intervention. In order to solve this problem, this paper presents a snake-based tracking algorithm to trace myocardial tissue automatically. A mathematical method is also introduced to extend the application of the snake model for detecting non-closed contours. The system developed in our previous research was redesigned to synchronize the radio frequency signal, the electrocardiographic signal, and the video signal, which allows verification of the system. Our results suggest that the system using the auto-tracking method increases the accuracy of detecting myocardial ischemia.