Onder Devrim, Zengin Selen, Sarioglu Sulen
Department of Pathology, Faculty of Medicine, Dokuz Eylul University, Izmir, Turkey.
Appl Immunohistochem Mol Morphol. 2014 Nov-Dec;22(10):713-9. doi: 10.1097/PAI.0000000000000003.
The histopathologists get the benefits of wide range of colored dyes to have much useful information about the lesions and the tissue compositions. Despite its advantages, the staining process comes up with quite complex variations in staining concentrations and correlations, tissue fixation types, and fixation time periods. Together with the improvements in computing power and with the development of novel image analysis methods, these imperfections have led to the emerging of several color normalization algorithms. This article is a review of the currently available digital color normalization methods for the bright field histopathology. We describe the proposed color normalization methodologies in detail together with the lesion and tissue types used in the corresponding experiments. We also present the quantitative validation approaches for each of the proposed methodology where available.
组织病理学家受益于多种染色剂，从而获得有关病变和组织组成的大量有用信息。尽管有其优点，但染色过程在染色浓度和相关性、组织固定类型以及固定时间段方面存在相当复杂的变化。随着计算能力的提高和新型图像分析方法的发展，这些不足之处催生了几种颜色归一化算法。本文综述了目前用于明场组织病理学的数字颜色归一化方法。我们详细描述了所提出的颜色归一化方法以及相应实验中使用的病变和组织类型。在可行的情况下，我们还介绍了每种所提出方法的定量验证方法。
