| 邢谦谦,王凯,李穆,陆燕琴,曹蕾.肺部CT图像的结节检测和类型判别准则设计及应用[J].中国医学影像技术,2016,32(7):1130~1134 |
| 肺部CT图像的结节检测和类型判别准则设计及应用 |
| Nodule detection, designation and application of nodule type discrimination criteria in lung CT images |
| 投稿时间：2015-11-30 修订日期：2016-03-25 |
| DOI：10.13929/j.1003-3289.2016.07.036 |
| 中文关键词: 肺肿瘤 图像处理 计算机辅助 检测 类型判别准则 |
| 英文关键词:Lung neoplasms Image processing computer-assisted Detection Type discrimination criteria |
| 基金项目:国家自然科学基金青年基金（61102114） |
|
| 摘要点击次数: 1885 |
| 全文下载次数: 794 |
| 中文摘要: |
| 目的 探讨肺结节自动检测算法和结节类型判别准则对肺结节检测及其类型判别的可行性。方法 首先利用结构分析、高通滤波、高斯拟合实现候选肺结节的检测，然后进行候选结节的假阳性筛除得到肺结节筛查结果。最后，在肺结节检测基础上设计了肺结节类型判别准则自动将肺结节分为磨玻璃样（GGO）、混合型GGO、实性结节三类。结果 本研究对107个肺结节进行实验，其总体检测率为87.85%，将结节分类结果与LIDC数据库的分类进行对比，一致性较高。结论 本文方法可针对不同类型的肺结节进行有效检测并自动分类，进而辅助临床进行肺癌的早期诊断。 |
| 英文摘要: |
| Objective To explore feasibility of nodule type discrimination criteria in lung nodule detection and the type discrimination. Methods First, the nodule structural analysis method followed by high-pass filtering method and Gaussian fitting method was used to accurately detect the lung nodules. Then false-positive screening method was used to complete the detection of nodules. Finally, the type discrimination criteria automatically divided the nodules into three categories:the ground glass opacity (GGO), the mixed GGO, the solid nodules. Results The overall detection rate of 107 pulmonary nodules in the experiments was 87.85%. Nodule classification in this article and LIDC database had high consistency. Conclusion The experiment results show that the method can effectively detect and automatically classify the lung nodules in CT images, and then assist doctors in early diagnosis of lung cancer. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
