摘要
本文在综合介绍目前基于医学影像的计算机辅助诊断(MICAD)研究工作的基础上,重点分析了MICAD主要使用的图像分析和处理技术。作为例子重点介绍了MICAD对图像分割的要求,介绍了目前我们正在开展的最小近邻算法(KNN)和模糊最小紧邻算法(FKNN)进行图像分割的工作,并提出了如何用相邻像素的信息,进一步提高分割的准确性的思路,为后面的计算机自动识别提供依据。而计算机自动识别是基于医学影像的计算机辅助诊断过程自动化的基础。
Based on the present investigation of the Medical Imaging Based Computer Aided Diagnosis (MICAD), main methods of the imaging analysis and process technologies used in MICAD are discussed in this paper. As an example, the imaging segmentation for MICAD is introduced in little more detail for the k-Nearest Neighbor (KNN) and Fuzzy k-Nearest Neighbor (FKNN) methods. The idea how to increase the segmentation accuracy was presented, and the mixture-based segmentation maybe a solution. Accurate segmentation is an important technology for the computer automatic pattern recognition, which inversely as the main fundamental role in intelligent MICAD.
出处
《中国医学物理学杂志》
CSCD
2003年第2期83-86,共4页
Chinese Journal of Medical Physics
基金
北京市重点自然科学基金项目(编号3011002)
