Blood Flow Estimation and Its Applications in Ultrasonic Small Animal Imaging

• Main Authors: Chen, Yen-Fu, 陳彥甫
• Other Authors: Li, Pai-Chi
• Format: Others
• Language: zh-TW
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/50161606796309691218

碩士 === 國立臺灣大學 === 電機工程學研究所 === 91 === Small animal models have been used extensively in genomics research, drug development, and developmental biology. Hence, the development of advanced small animal imaging technologies in improving the research quality and efficiency has become increasingly important and is also the main goal of this study. Ultrasound imaging has the advantages of cost-effectiveness, non-invasiveness, rapid imaging and portability. Optimal imaging of small animals, however, requires at least 10-fold improvement in resolution over the conventional ultrasound imaging system. In addition, traditional signal processing techniques are also inadequate to precisely estimate blood flow parameters in small vessels. Because resolution scales directly with frequency, the required level of resolution can be achieved by employing much higher ultrasound frequency. In this study, a high frequency ultrasound imaging system specifically designed for microimaging of the mouse is constructed. Not only limitations in traditional signal processing technique are overcame through the improvement of flow estimation strategies, but also the acquisition time associated with discrete mechanical scanning of a single piston transducer is reduced by utilizing a continuous scanning technique called swept-scan. Simulation and experimental results show that the 2D autocorrelation technique, the wideband maximum likelihood estimator, and the butterfly search technique all can achieve better velocity resolution and are more robust to noise than the conventional 1D autocorrelation approach. The butterfly search technique is the best among these algorithms in terms of the tradeoff between the computation time and the performance. In-vivo experiments on the mouse embryo and tumor model also demonstrate the ability to simultaneously acquire and display B-mode, color-mode, and power Doppler small animal images. Pulsed wave Doppler is also implemented. Experiments on mouse tumors successfully image tumor vascularity with flow velocities on the order of mm/s. This imaging system can be used to provide valuable information for biomedical research in the future.