Title :
Real-time motion tracking for non-invasive ultrasound cardiac therapy using histotripsy
Author :
Miller, Ryan M. ; Kim, Yohan ; Lin, Kuang-Wei ; Owens, Gabe ; Cain, Charles A. ; Xu, Zhen
Author_Institution :
Dept. of Biomed. Eng., Univ. of Michigan, Ann Arbor, MI, USA
Abstract :
Cardiac and respiratory motion are significant concerns for achieving high treatment efficiency in cardiac ultrasound therapy. This paper develops a fast target motion tracking algorithm integrated with real-time ultrasound imaging for histotripsy-mediated cardiac therapy. Motion tracking is performed by finding the best match in an ultrasound image to a reference target image. Block matching using diamond search produces a position estimate using the criteria of sum of absolute difference, which is refined by a weighted average with the target´s expected position (Kalman filter). For proof of concept this algorithm was configured to use with a 1 MHz annular therapy array that is capable of steering its focus in the axial direction. The motion tracking system operated at peak frame rate of 62 Hz, tracking target displacements up to 20 mm at peak velocities up to 80 mm/second. The system is capable of sub-mm tracking accuracy with maximum observed error less than 3 mm at the highest velocity (80mm/sec). Testing in tissue phantoms showed no significant reduction in treatment efficiency measured by erosion rate for displacements up to 12.5 mm, compared to significant reduction in erosion rate without motion tracking for displacements as small as 3 mm. This was validated in ex vivo tissue, where for 10 mm displacement a mean erosion rate with motion tracking was maintained at 97% of the erosion rate for the control case (stationary target), compared to only 42% with no tracking. The system was tested in vivo in an open chest canine model targeting the atrial septum, where the system provided 24 minutes of motion corrected therapy in the live beating heart, generating significant damage in the atrial septum. These results show that our algorithm has sufficient speed to track cardiac and respiratory motion for histotripsy cardiac therapy, providing significantly higher accuracy and efficiency than therapy without motion tracking. By replacing the annular array with a 3D- steerable phased array, this system can be expanded to 3D tracking for histotripsy therapy of cardiac or other fast moving tissue.
Keywords :
Kalman filters; biomedical ultrasonics; ultrasonic imaging; 3D steerable phased array; 3D tracking; Kalman filter; atrial septum; block matching; cardiac ultrasound therapy; fast moving tissue; frequency 62 Hz; histotripsy cardiac therapy; histotripsy therapy; histotripsy-mediated cardiac therapy; motion corrected therapy; motion tracking system; non-invasive ultrasound cardiac therapy; real-time motion tracking; real-time ultrasound imaging; reference target image; respiratory motion; size 10 mm; size 20 mm; target motion tracking algorithm; time 24 min; ultrasound image; velocity 80 mm/s; Algorithm design and analysis; Imaging; Medical treatment; Real time systems; Target tracking; Ultrasonic imaging; cardiac therapy; histotripsy; motion tracking;
Conference_Titel :
Ultrasonics Symposium (IUS), 2011 IEEE International
Conference_Location :
Orlando, FL
Print_ISBN :
978-1-4577-1253-1
DOI :
10.1109/ULTSYM.2011.0229