Methods To Determine And Enhance The HIFU Effective Treatment Region Based On Mode B Ultrasonic Image Processing

As a new non-destructive cancer treatment method, high intensity focused ultrasound (HIFU) therapy, which can succeed after only one time treatment and has no side-effects, has become a hotspot in international ultrasound treatment field. Real-time monitoring during HIFU treatment is a key problem to assure high efficiency and safety. At present, when treatment process is monitored by B-mode ultrasonic equipment, the degree and area of lesions are determined mainly by the doctor based on B-mode ultrasonic images, which is lack of objectivity and accuracy.Under the condition that HIFU and B-mode ultrasonic device are combined to use in experiments, how to detect and enhance damage regions automatically by image-processing techniques during HIFU therapy is studied in this paper. The main done work is as follows:1) B-mode ultrasonic images obtained before and after HIFU radiation are gray-scale transformed, normalized, cut and subtracted. Then the subtraction image is de-noised by iterative mathematical morphology method, which removes the noise and holds the information of lesions simultaneously. 2) Variety of image segmentation methods are tested to detect damage regions according to the image characteristics of HIFU and requirements of practical application. Compared to LoG operator and OTSU method, the treated region detected by threshold method based on the expected value is more consistent with the actual situation, so it can provide references for the doctor to determine the segmentation threshold in the clinical treatment. Positioning capability of region growing method based on dynamic shades of gray to determine seed points is improved, so its result is good. When lesions in biological tissue are multi-points and denatured regions do not overlap, this method is very fit. The FCM algorithm which takes the spatial information into account has stronger anti-noise ability than the traditional one. The damage region detected by this method has a clear outline (retaining the details), and its size is consistent with the actual situation basically.3) The linear transformation function of traditional spatial gray-color transformation method is modified, and it is applied to enhance the destructive area. This method has higher resolution than the traditional one, and it has maintained the original brightness level. To establish the color plotting of the relationship between gray scale and color, it provides a reliable basis for doctor's semi-quantitative analysis.A lot of research work focused on real-time monitoring during HIFU therapy has been done in this paper, and it is helpful for promoting the application and development of HIFU technology.