免疫组化彩色细胞图像自动分割的研究

A Study on Automatic Segmentation for Immunohistochemical Color Cell Image

目的:免疫组化彩色细胞图像中阳性细胞的自动分割提取有着重要的临床意义。本文结合三种分割算法的特点,研究实现免疫组化彩色细胞图像的自动分割,提取图像中的阳性细胞。方法:(1)采用OTSU方法在灰度的基础上对免疫组化彩色细胞进行预分割,去除无关背景。(2)使用K-聚类算法,对彩色细胞图像进行彩色分类筛选出阳性细胞和阴性细胞,并对所得阳性细胞图像进行腐蚀,以获取阳性细胞图像的种子。(3)使用区域生长算法对阳性细胞种子进行区域增长,获取完整的阳性细胞图。结果:准确分割出图像中的阳性细胞。结论:该自动分割方法可用于后续的阳性细胞自动计数,辅助医生诊断疾病。

Objective： The automatic segmentation of positive cell ofimmunohistochemical color cell image plays an important role in clinic. This work study on how to segment irnmunohistochemical color cell image and extract positive cell automatically based on three algorithms. Methods： （1） An adaptive algorithm of OTSU is applied to pre-separate immunohistochemical color cell image to eliminate background. （2） An algorithm of K-means is used to classify positive and negative cells based on color, and then erosion algorithm is used to obtain seeds of two kinds of cells. （3） Region growing algorithm is utilized to select positive and negative cells based on seeds. Results： Experimental results prove that positive cell can be extracted accurately, Conclusions： The automatic segmentation can be applied in counting positive cells and assisting doctor to make a diagnosis.