Ultrasound Image-guided Tracking Algorithm for Moving-tumor Treatment

摘要

Image-guided technology, such as MR-guided focused ultrasound surgery, has recently proposed as a treatment for moving-tumor. However, the 3 Hz frame rate of MR imaging is still too slow for real-time tracking of the tumor in motion. The fact that a 30 Hz frame rate can easily be achieved with the ultrasound imaging system makes it a very viable option. In this study, an ultrasound imaging tracking algorithm for liver tumor has been developed. The proposed ultrasound imaging tracking algorithm utilized the concepts of pattern matching, local search, and image correlation to track a pre-specified target area in real time. In order to know the resolution of the proposed tracking algorithm, animal experiments were conducted. Three 6 degree-of-freedom electromagnetic tracking sensors, the 3D guidance trakSTAR, were surgically attached to the liver of anesthetized pigs. The liver motion could then be derived from the position signal from the trakSTAR sensors with a resolution of 0.5mm. The in-vivo experiments showed the tracking accuracy of the ultrasound-guided tracking algorithm is better than 1 mm for moving tumors.