Biomedical Diagnostic Centre, Hospital Clínic, University of Barcelona, Barcelona, Spain.
Department of Mathematics, Barcelona Est Engineering School, Technical University of Catalonia, Barcelona, Spain.
Int J Lab Hematol. 2018 May;40 Suppl 1:54-61. doi: 10.1111/ijlh.12832.
Morphological review of the peripheral blood smear is still a crucial diagnostic aid as it provides relevant information related to the diagnosis and is important for selection of additional techniques. Nevertheless, the distinctive cytological characteristics of the blood cells are subjective and influenced by the reviewer's interpretation and, because of that, translating subjective morphological examination into objective parameters is a challenge.
The use of digital microscopy systems has been extended in the clinical laboratories. As automatic analyzers have some limitations for abnormal or neoplastic cell detection, it is interesting to identify quantitative features through digital image analysis for morphological characteristics of different cells.
Three main classes of features are used as follows: geometric, color, and texture. Geometric parameters (nucleus/cytoplasmic ratio, cellular area, nucleus perimeter, cytoplasmic profile, RBC proximity, and others) are familiar to pathologists, as they are related to the visual cell patterns. Different color spaces can be used to investigate the rich amount of information that color may offer to describe abnormal lymphoid or blast cells. Texture is related to spatial patterns of color or intensities, which can be visually detected and quantitatively represented using statistical tools.
This study reviews current and new quantitative features, which can contribute to optimize morphology through blood cell digital image processing techniques.
外周血涂片的形态学检查仍然是一种重要的诊断辅助手段，因为它提供了与诊断相关的信息，对选择附加技术也很重要。然而，血细胞的独特细胞学特征是主观的，受检查者解释的影响，因此，将主观的形态学检查转化为客观参数是一个挑战。
数字显微镜系统在临床实验室中的应用已经得到扩展。由于自动分析仪在检测异常或肿瘤细胞方面存在一些局限性，因此通过数字图像分析来识别不同细胞的形态特征的定量特征很有趣。
主要使用了三类特征：几何特征、颜色特征和纹理特征。几何参数（核/浆比、细胞面积、核周长、细胞质轮廓、RBC 接近度等）为病理学家所熟悉，因为它们与细胞的视觉形态有关。不同的颜色空间可用于研究颜色可能提供的丰富信息，以描述异常的淋巴细胞或原始细胞。纹理与颜色或强度的空间模式有关，可以使用统计工具进行视觉检测和定量表示。
本研究综述了当前和新的定量特征，这些特征可以通过血细胞数字图像处理技术来优化形态学。
