离散Walsh变换在超声治疗无损测温中的应用  

Application of Noninvasive Temperature Measurement with Discrete Walsh Transform in Ultrasound Therapy

作　　者：李勇[1] 丁亚军[1,2] 钱盛友[1] 段先知 邹孝[1] 

机构地区：[1]湖南师范大学物理与信息科学学院,长沙410081 [2]湖南师范大学图像识别与计算机视觉研究所,长沙410081

出　　处：《计算机工程》2015年第1期227-230,共4页Computer Engineering

基　　金：国家自然科学基金资助项目(11174077;11474090);湖南省自然科学基金资助项目(11JJ3079);湖南省教育厅基金资助项目(12C0237)

摘　　要：利用高强度聚焦超声对新鲜的离体猪肉进行辐射,并通过B超获取辐射前后的超声图像。采用二维离散Walsh变换法,从辐射前后的超声图像的差图像中分离出低频分量,作为反映温度信息的参数。选取确定系数作为衡量误差的参数,通过曲线拟合,发现二维离散Walsh变换直流分量与温度呈近似线性关系,且多组数据的斜率非常接近,与传统的灰度均值测温法相比,实验结果表明,该方法误差更小,温度分辨率更高,使用正交矩阵变换,符合计算机语言的逻辑,仿真检测时间较短,在实际应用中能为超声无损测温提供准确可靠的依据。Irradiating fresh pork in vitro by High Intensity Focused Ultrasound（HIFU）,subtraction image of B ultrasonic image is obtained before and after irradiation. The low-frequency components as the parameters about the temperature information,are separated from the subtraction image by two-dimensional discrete Walsh transform. There is an approximate linear relationship between Discrete Walsh Transform Direct Component（DWTDC）and temperature,and the slopes of multiple sets of data are very close proximity using R-square as a parameter to measure the error and curve fitting. The error is smaller. Temperature resolution is higher than the traditional temperature estimation method by using gray average. This method using the orthogonal transform matrix conforms to the logic structure of computer language,can shorten detection time and provide reliable basis for noninvasive temperature estimation in practical application.

关 键 词：高强度聚焦超声 二维离散Walsh变换 低频分量 无损测温 差图像 曲线拟合 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]