| 胰腺内镜超声图像纹理特征提取与分类研究 |
| |
| 引用本文： | 蔡哲元,余建国,张敏敏,金震东.胰腺内镜超声图像纹理特征提取与分类研究[J].上海生物医学工程,2008(3):141-145. |
| |
| 作者姓名： | 蔡哲元 余建国 张敏敏 金震东 |
| |
| 作者单位： | 复旦大学电子工程系;长海医院消化科; |
| |
| 基金项目： | 上海市重点学科建设项目(B112) |
| |
| 摘 要： | 提出了胰腺内镜超声图像的纹理特征提取与分类方法，可应用于胰腺癌内镜超声图像的计算机辅助诊断。对胰腺内镜超声图像采用数字图像处理算法提取9大类共69个纹理特征。使用类间距作为可分性判据，实现特征的初步筛选，之后使用顺序前进搜索算法进一步筛选特征，并由支撑向量机实现分类。对216例病例随机选取训练集和测试集，通过多次随机实验表明。本文提出的算法实现了较高的分类准确率，为胰腺癌的临床诊断提供有价值的参考意见。
|
| 关 键 词： | 胰腺癌 内镜超声 纹理特征 模式分类 |
Texture Feature Extraction and Classification of Pancreatic Endoscopic Ultrasonography Images |
| |
| CAI Zhe-yuan YU Jian-guo.Texture Feature Extraction and Classification of Pancreatic Endoscopic Ultrasonography Images[J].Shanghai Journal of Biomedical Engineering,2008(3):141-145. |
| |
| Authors: | CAI Zhe-yuan YU Jian-guo |
| |
| Institution: | Shanghai 200433;Shanghai 200433 |
| |
| Abstract: | The method of texture feature extraction and classification of pancreatic endoscopic ultrasonogra- phy(EUS)images is proposed,which can be applied to the computer-aided diagnosis of pancreatic EUS images. 69 texture features of 9 feature sets are extracted from the pancreatic EUS images according to the digital image pro- cessing algorithm.Distance between classes is chosen as the separability criterion to select the features in the first step,then the Sequential Forward Selection algorithm is performed to ... |
| |
| Keywords: | Pancreatic Cancer EUS texture feature pattern classification |
| 本文献已被 CNKI 维普 等数据库收录！ |
