基于竞争最优原型淋巴细胞协同分类.pdfVIP

2006 11 2 CH INESE JOURNAL OF STEREOLOGY AND MI AGE ANALYSIS Vo.l 11 No. 2 June 2006 71 : 1007- 1 82( 2006) 02- 0071- 06 1 1 2 , , ( 1. , 350002; 2. , 350025) : 提出一种基于竞争最优原型的淋巴细胞协同分类方法 首先, 采用改进的 IS DATA 算法聚类同型 态的淋巴细胞原型集;然后,依据待分类细胞样本,建立协同竞争机制求 出各原型集中的最优原型; 最后, 将待分类样本分别与各型最优原型进行相似匹配, 以最大相似结 果作为分类结果 实现了淋巴滤泡套细胞淋巴滤泡中心细胞淋巴滤泡中心母细胞中心 母细胞性淋巴瘤细胞等 4组类型的细胞分类 实验结果表明该方法分类效果理想, 为下 一 步淋巴细胞定量分析打下基础 : 淋巴细胞; 协同分类; 最优原型 : TP391; R392. 12 : A Lymphocyte synergetic classification based on competition optimum prototype 1 1 2 LIU Binghan, WANG Weizhi, ZHENG Zhiyong ( 1. Fuzhou University, Fuzhou 350002, China; 2. Fuzhou GeneralHospita,l NanjingM ilitary Area, Fuzhou 350025, China) Abstract: bjective To put forward a lymphocyte synergetic classification based on competition opti mum prototype. Method Firstly, according to IS DATA arithm etic, set up the lymphocyte prototype set. Secondly, in accordancew ith the cell swatch, pick out the optmi um prototype based on synergetic competitionmodel. Lastly, match the cell swatch w ith the optmi um prototype and obtain themost smi ilar result. Results The classification of mantle cells, follicular centrocytes cells, follicular centroblastic cells and centroblastic lymphoma cells is achieved. Conclusion The expermi ental results show that the method can be further extended to lymphocyte quantitative analysis. Key words: lymphocyte; synergetic classification; optmi um prototype , , , , , , , ,