项目名称: 基于CT图像的肺部肿瘤辅助诊断关键技术研究
项目编号: No.61272245
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 自动化技术、计算机技术
项目作者: 刘慧
作者单位: 山东财经大学
项目金额: 81万元
中文摘要: 研制达到或超过专家会诊水平的医学影像计算机辅助诊断系统是人们梦寐以求的事情,它将促使医学影像诊断技术的变革,使医学图像的辅助诊断水平取得极大的提高。本项目研究基于CT图像的肺部肿瘤辅助诊断关键技术,主要研究内容包括:1、二维CT图像的肺肿瘤特征提取与表达;2、三维CT图像的肺肿瘤特征提取与表达;3、基于特征聚类和多变量单值函数理论,建立肺肿瘤特征与肿瘤类型之间的一一对应关系,4、研究基于不同特征描述组合的肺部肿瘤分类算法。研究范围涉及理论与方法、算法与技术以及原型系统。项目的目标是为解决CT图像肺肿瘤的特征提取、特征描述和肺肿瘤分类关键问题提供新思路、新理论和新方法,为研制基于CT医学图像的肺肿瘤辅助诊断系统提供一系列简单、高效和鲁棒的技术。特征提取、表达和分类也是计算机图像学、计算机视觉、虚拟现实和模式识别等中的共性基础问题,因此项目的结果具有理论意义和潜在应用价值。
中文关键词: CT图像;肺部肿瘤;特征提取;分类;计算机辅助诊断
英文摘要: It's important to promote the changes of diagonosis technologies based on medical images by way of developing advanced computer aided diagnosis system that achieves or exceeds the level of expert consultation, which will improve the level of medically assisted diagnosis. This project researches the key technologies of lung tumor's computer aided diagonosis based on CT images, the research contents include: 1.Extract and express the features of 2-D CT images; 2.Extract and express the features of 3-D CT images; 3.Establish the one-to-one corresponding relationship between feature description and tumor types, based on the feature clustering and single-valued function algorithms; 4.Realize the multi-classification of lung tumors based on different features combination. The research scope involves theory and method, algorithm and technology, prototype system and so on. The project goal is to provide new ideas, theories and methods for solving the key technologies of feature extraction of lung tumor's CT images, feature description and tumor classfication, and present a series of simple, efficient and robust techniques for devoloping the lung tumor aided diagnosis system based on CT images. The project results will have theoretical significance and potential application value, because the problems of feature extracti
英文关键词: CT image;Lung Tumor;Feature Extraction;Classification;Computer-aided Diagnosis