基于核磁共振图像的人体面部软组织的分割

摘要

面部软组织的分割是医学诊断和临床治疗最重要的数据基础之一,对提高医生诊断的准确性和保证外科整形手术的质量有重要意义.为实现面部软组织的准确分割,结合MICO算法和区域生长法设计了一种面部脂肪组织的分割方法,该方法通过灰度不均匀处理、初分类移除背景和二分割提取组织实现了较传统算法更为准确的脂肪组织的分割.与此同时,还提出了一种新的咬肌组织分割方法,该方法通过设计一个新的边缘停止函数改进了 DRLSE方法.提出的方法将相位一致性引入构造边缘停止函数,结合图像梯度信息及相位信息解决分割问题,得到了平均准确率达92.41% 的分割结果.%Facial soft tissue segmentation is one of the most important bases for medical diagnosis and clinical treatment.It is of great significance to improve the accuracy of diagnosis and ensure the quality of plastic surgery.In order to realize the accurate segmentation of facial soft tissue,a method of facial adipose tissue segmentation is designed by combining MICO algorithm and region growing method.This method realizes the more accurate segmentation results through intensity inhomogeneity processing,preliminary classification to remove the background and second segmentation to obtain the adipose tissue.At the same time,a new method of masseter muscle segmentation is provided,w hich modifies the DRLSE with a new adaptive edge indicator function.T he provided method introduces the phase congruency into the edge indicator function, combines the image gradient with phase information together to solve the segmentation problem and achieves an average accuracy reached 92.41%.