CB微核法图像自动分析技术研究进展

Research Progress on Automatic Analytical Technique for Cytokinesis-block Micronucleus Images

淋巴细胞胞质分裂阻断微核法(CB微核法)自动分析是剂量快速估算的关键。本文从微核图像的预处理、自动分割、淋巴细胞区域的分类、双核淋巴细胞的自动识别、粘连细胞的判别与分离、微核的自动识别和计数6个技术环节介绍了CB微核法图像自动分析技术研究进展,为研制适用于CB微核法图像的自动分析系统提供参考。

The cytokinesis block micronucleus test （CB MNT） in binucleated lymphocytes （BNL） can serve as a biological dosimeter after radiation exposure. The automation of the MNT in human BNL has considerably advanced in recent years. The primary techniques for the automatic analysis of the BNL MNT images stained with Giemsa were summarized in detail, including the following contents primarily： （ 1 ） preprocessing of MN images, （2） automatic segmentation of MN images, （3） classification of the regions in MN image, （4） automatic scoring for the BNL, （5） automatic separating for the overlapping cells, and （6） automatic counting of MN, etc.