首页> 外文期刊>電子情報通信学会技術研究報告. 医用画像. Medical Imaging >Tumor region extraction keeping boundary shape information and its application to benign/malignant discrimination of tumors
【24h】

Tumor region extraction keeping boundary shape information and its application to benign/malignant discrimination of tumors

机译:保留边界形状信息的肿瘤区域提取及其在肿瘤良恶性鉴别中的应用

获取原文
获取原文并翻译 | 示例
           

摘要

This paper presents a method to exactly extract tumor regions for discrimination between benign tumor and malignant tumor from chest X-ray CT images used for a close medical examination. For benign/malignant discrimination using CAD(Computer-Aided Diagnosis) system, tumor regions which have exact edge information is required to define regions for calculation of features and quantification of edge information itself. We applied the method to seventy-eight practical chest X-ray CT images. It was shown that tumor regions in sixty-seven CT images were extracted with detailed shapes of boundary. Furthermore, we tested this method bt applying it to sixty-seven CT images which were extracted tumor regions well. It was confirmed that the performance of discrimination was improved by using the tumor regions extracted by the proposed method.
机译:本文提出了一种方法,可以从用于密切医学检查的胸部X射线CT图像中准确地提取出可区分良性肿瘤和恶性肿瘤的肿瘤区域。对于使用CAD(计算机辅助诊断)系统的良性/恶性鉴别,需要具有精确边缘信息的肿瘤区域来定义用于特征计算和边缘信息本身量化的区域。我们将该方法应用于78例实用的胸部X射线CT图像。结果表明,在67幅CT图像中,肿瘤区域被提取出来并具有详细的边界形状。此外,我们测试了该方法,并将其应用于67个CT图像中,这些CT图像很好地提取了肿瘤区域。证实了通过使用所提出的方法提取的肿瘤区域改善了辨别性能。

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