论文部分内容阅读
目的提出一种基于HSV空间的免疫组化阳性目标自动提取分析方法。方法首先,依据阳性目标在HSV颜色空间的分布规律,联合使用了H分量、S分量和V分量作为阈值条件,结合最大熵单阈值分割算法,建立阳性目标分割提取模型;然后,提取阳性目标区域的特征参数,定量分析阳性目标的密度、强度等指标参数。结果实现了血管、细胞核、细胞浆和细胞膜等阳性目标的自动提取分析。结论试验结果可为建立免疫组化标准化的判断指标提供量化基础。
Objective To propose an automatic extraction and analysis method for positive immunohistochemistry based on HSV space. Methods Firstly, according to the distribution of positive targets in the HSV color space, H-component, S-component and V-component are jointly used as the threshold condition. In combination with the maximum-entropy single-threshold segmentation algorithm, a positive target segmentation extraction model is established; then, a positive target region is extracted. The characteristics of the parameters, quantitative analysis of the positive target density, strength and other indicators. As a result, automatic extraction and analysis of positive targets such as blood vessels, nucleus, cytoplasm, and cell membrane were achieved. Conclusion The test results can provide a quantitative basis for the establishment of standardized indicators for immunohistochemistry.