【作者】 王浩军； 郑崇勋； 闫相国； 张新宇；

【Author】 Wang Haojun Zheng Chongxun Yan Xiangguo Zhang Xinyu(Institute of Biomedical Engineering ,Xi’an Jiaotong University ,Xi’an 710049)

【机构】 西安交通大学生物医学工程研究所；

【摘要】 为了开发血及骨髓涂片中血细胞自动分类及计算机辅助诊断系统,提出了一个基于自适应多尺度阈值和流域变换的彩色混合分割算法完成对有核血细胞(胞核和胞浆)的检测和分割。该方法有效地利用了图像中血细胞的颜色信息及细胞的一些先验知识,快速地完成对有核细胞的检出和粘连细胞的分离。结果显示胞核和胞浆的分割成功率分别为91.9%和88.4%。它必将提高对血细胞的自动分类性能。更多还原

【Abstract】 To develop an automatic classification and diagnostic system for the hematopoietic cells from the blood and bone marrow smears,a hybrid technique which combines adaptive multi-scale thresholding and watershed transform is proposed for the detection and segmentation of nucleated hematopoietic cells (both nucleus and cytoplasm).The segmentation strategy uses the color information and prior knowledge to detect nucleated cells quickly and separate overlapped cells effectively.The result shows that the average corrected rate of segmentation is 91.9% for nuclei and 88.4% for cytoplasm.It will improve the automatic classification of cells.更多还原