首页> 外文会议>5th Kuala Lumpur international conference on biomedical engineering 2011 >Computer-Aided Diagnosis System for Pancreatic Tumor Detection in Ultrasound Images
【24h】

Computer-Aided Diagnosis System for Pancreatic Tumor Detection in Ultrasound Images

机译:超声图像中胰腺肿瘤检测的计算机辅助诊断系统

获取原文
获取原文并翻译 | 示例

摘要

In this study, a computer-aided diagnosis (CAD) system for pancreatic tumors detection in ultrasound images has been developed to provide a physician some diagnosis information and reduce the error rate in diagnosis for patients. After reducing noises, enhancing contrast, and detecting boundary of a tumor in the original ultrasound image, its texture features and morphological features were analyzed. The statistically effective features were selected and served as inputs in the self-organizing map (SOM) to classify the ultrasound images after evaluating the results. The diagnostic efficiency of the CAD system was evaluated after comparing the classified results of ultrasound images with the pathological results of patients. The primary results showed that morphological features had a better performance than texture features for pancreatic tumor classification in an ultrasound image. According to the results, 8 features were proved to effectively classify normal pancreas images and pancreatic tumor images, and 4 features of them were proved to effectively classify benign tumor images and malignant tumor images at the same time. The area of a pancreatic tumor seemed to be the most important morphological feature for image classification. A benign pancreatic tumor usually had a smaller area and a smoother contour than a malignant one.
机译:在这项研究中,已经开发了一种用于在超声图像中检测胰腺肿瘤的计算机辅助诊断(CAD)系统,以为医生提供一些诊断信息并降低患者诊断的错误率。在减少噪声,增强对比度并检测原始超声图像中的肿瘤边界之后,分析了其纹理特征和形态特征。选择统计有效特征并将其用作自组织图(SOM)中的输入,以在评估结果后对超声图像进行分类。在将超声图像的分类结果与患者的病理结果进行比较之后,评估了CAD系统的诊断效率。主要结果表明,对于超声图像中的胰腺肿瘤分类,形态特征比纹理特征具有更好的性能。根据结果​​,证明了8个特征可以有效地对正常胰腺图像和胰腺肿瘤图像进行分类,并且其中的4个特征可以同时对良性肿瘤图像和恶性肿瘤图像进行有效分类。胰腺肿瘤区域似乎是图像分类最重要的形态特征。良性胰腺肿瘤通常比恶性肿瘤具有更小的面积和更平滑的轮廓。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号