| 基于小波系数Hu矩的生物组织损伤监测方法 |
| |
| 引用本文： | 颜佩,丁亚军,钱盛友,胡强,盛祎,邹孝.基于小波系数Hu矩的生物组织损伤监测方法[J].电子测量与仪器学报,2016,30(7):1062-1067. |
| |
| 作者姓名： | 颜佩 丁亚军 钱盛友 胡强 盛祎 邹孝 |
| |
| 作者单位： | 1. 湖南师范大学物理与信息科学学院长沙 410081;2. 湖南师范大学物理与信息科学学院长沙 410081; 湖南师范大学图像识别与计算机视觉研究所长沙 410081 |
| |
| 基金项目： | 国家自然科学基金(11174077，11474090)，湖南师范大学博士基金(130645) |
| |
| 摘 要： | 提出了一种基于B超图像的小波系数Hu矩特征值并结合支持向量机监测生物组织损伤的方法。利用高强度聚焦超声(HIFU)对新鲜离体猪肉组织进行辐照,实时获取辐照前后的B超图像,并对其进行预处理获取减影图像。提取减影图像的Hu矩、小波系数均值和基于小波系数的Hu矩3个特征值,分别利用支持向量机对生物组织样本进行学习、分类处理。结果表明:小波系数Hu矩特征值比Hu矩和小波系数均值的总辨识率分别高出2.70%和2.05%,从而可以更有效地监测HIFU治疗中生物组织损伤情况。该方法可以帮助临床医生客观地监控HIFU治疗过程,对提高HIFU疗效有实际意义。
|
| 关 键 词： | 高强度聚焦超声 小波系数 Hu矩 支持向量机 生物组织损伤 |
Biological tissue lesion monitoring method based on Hu moment of wavelet transform coefficient |
| |
| Yan Pei,Ding Yajun,Qian Shengyou,Hu Qiang,Sheng Yi and Zou Xiao.Biological tissue lesion monitoring method based on Hu moment of wavelet transform coefficient[J].Journal of Electronic Measurement and Instrument,2016,30(7):1062-1067. |
| |
| Authors: | Yan Pei Ding Yajun Qian Shengyou Hu Qiang Sheng Yi and Zou Xiao |
| |
| Affiliation: | College of Physics and Information Science, Hunan Normal University, Changsha 410081, China,1. College of Physics and Information Science, Hunan Normal University, Changsha 410081, China;
2. Institute of Image Recognition & Computer Vision, Hunan Normal University, Changsha 410081, China,College of Physics and Information Science, Hunan Normal University, Changsha 410081, China,College of Physics and Information Science, Hunan Normal University, Changsha 410081, China,College of Physics and Information Science, Hunan Normal University, Changsha 410081, China and College of Physics and Information Science, Hunan Normal University, Changsha 410081, China |
| |
| Abstract: | A biological tissue lesion monitoring method based on Hu moment of the wavelet transform coefficient of B-mode ultrasound image is proposed in this paper.Irradiating fresh pork in vitro by high intensity focused ultrasound (HIFU ),real-time B-mode ultrasound images before and after irradiation are obtained.B-mode ultrasound images before and after irradiation are pretreated for digital subtraction images, and three characteristics:Hu moment,the mean of the wavelet transform coefficient as well as Hu moment based on the wavelet transform coefficient,are extracted from digital subtraction images.Then,the biological tissue samples are learned,classified and processed by support vector machine (SVM) respectively.The results show that the biological tissue lesion recognition rate of Hu moment of the wavelet transform coefficient is higher compared with Hu moment and the mean of the wavelet transform coefficient.It means that it can monitor the biological tissue lesion in HIFU treatment more effectively.This method can help clinicians to monitoring the HIFU treatment process objectively,and it is significant practically to improve the HIFU therapeutic effect. |
| |
| Keywords: | high intensity focused ultrasound (HIFU ) wavelet transform Hu moment support vector machine (SVM) biological tissue lesion |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《电子测量与仪器学报》浏览原始摘要信息 |
|
点击此处可从《电子测量与仪器学报》下载全文 |
|
