淋巴瘤病理图像中R-S细胞的自动提取  

Automatic Abstraction R-S Cells in Pathological Images of Lymphoma

作　　者：陈宏方[1] 刘秉瀚[1] 郑智勇[2] 

机构地区：[1]福州大学数学与计算机科学学院,福建福州350002 [2]南京军区福州总医院病理科,福建福州350025

出　　处：《集美大学学报（自然科学版）》2005年第2期142-145,共4页Journal of Jimei University：Natural Science

基　　金：国家自然科学基金项目(60372027);福建省自然科学基金项目(A0310010)

摘　　要：介绍一种淋巴瘤病理图像中R-S细胞自动提取的方法.针对淋巴瘤病理图像中组织灰度分布的特点,采用循环分解和类别方差思想的混合自动分割算法分割病理图像;提出一个基于目标形态、灰度纹理和空间相关性等特征的识别模型,提取R-S细胞.实验结果表明,该方法速度快,能准确地识别提取淋巴瘤病理图像中的R-S细胞,有助于病理医生研究诊断.Amethod of automatically abstracting R-S cells in pathological images of lymphoma is intro- duced. According to the characteristics of tissular gray distribution in pathological images, a hybrid method based on the algorithm of recursive threshold ing to segment pathological images is adopted and a recognition model based on mathematicalmorphology gray texture and spatial relativities to abstract R-S cells are provid- ed. The experiment results show that the efficiency of this method is satisfying, the method can precisely ab- stractR-S cells in pathological images of lymphoma and help pathological doctors to research and diagnose.

关 键 词：R-S细胞 图像分割 自动提取 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]