Noninvasive Temperature Estimation Using Real-valued Gabor Transform

A new method for noninvasive temperature estimation is proposed in this paper. Noninvasive temperature estimation is one of key problems of high-intensity focused ultrasound (HIFU). In tissue, temperature changes, caused by HIFU, cause gray variation of the B-mode ultrasonic image, which can be used to estimate temperature. In this paper, the relation between temperature and the energy of real-valued Gabor transform of difference image is studied. In 2D real-valued discrete Gabor transform (RDGT), the information, which is response to the temperature, can be detached from the other information such as noise. It is advantageous to estimating temperature. The in-vitro fresh pork tissue is used in the experiments. Experimental results show the relation between temperature and energy of RDGT is approximately linearity, and noninvasive estimation based on RDGT has a high temperature resolution. The RDGT can be applied to estimate temperature.