Abstract: In order to determine effective treatment region more accurately and objectively during the process of HIFU treatment，different image segmentation algorithms are compared based on the characteristics of HIFU treatment.Then the method of combining threshold based on arithmetic mean of gray value with iterative mathematical morphology is used to detect the size of the HIFU treatment regions.Experiments show that the results of this method are consistent with the actual injury，which proves the feasibility and effectiveness of this method.

摘要： 为了在HIFU治疗过程中更客观有效地判定有效治疗区域，根据HIFU治疗的特点，在比较各种图像分割算法的基础上采用基于灰度期望值的阈值法，并与迭代数学形态学方法结合，从图像处理角度检测出了HIFU焦域的大小。实验表明该方法所获得的结果与实际损伤情形基本吻合，从而证明了该方法的可行性和有效性。