Kayser Klaus, Hoshang Sabah Amir, Metze Konradin, Goldmann Torsten, Vollmer Ekkehard, Radziszowski Dominik, Kosjerina Zdravko, Mireskandari Masoud, Kayser Gian
Institute of Pathology, The International Union Against Cancer-Telepathology Consultation Center, Charite, Charite Platz 1, D-10117 Berlin, Germany.
Anal Quant Cytol Histol. 2008 Dec;30(6):323-35.
To create algorithms and application tools that can support routine diagnoses of various organs.
A generalized algorithm was developed that permits the evaluation of diagnosis-associated image features obtained from hematoxylin-eosin-stained histopathologic slides. The procedure was tested for screening of tumor tissue vs. tumor-free tissue in 1,442 cases of various organs. Tissue samples studied include colon, lung, breast, pleura, stomach and thyroid. The algorithm distinguishes between texture- and object-related parameters. Texture-based information-defined as gray value per pixel measure--is independent from any segmentation procedure. It results in recursive vectors derived from time series analysis and image features obtained by spatial dependent and independent transformations. Object-based features are defined as gray value per biologic object measured.
The accuracy of automated crude classification was between 95% and 100% based upon a learning set of 10 cases per diagnosis class. Results were independent from the analyzed organ. The algorithm can also distinguish between benign and malignant tumors of colon, between epithelial mesothelioma and pleural carcinomatosis or between different common pulmonary carcinomas.
Our algorithm distinguishes accurately among crude histologic diagnoses of various organs. It is a promising technique that can assist tissue-based diagnosis and be expanded to virtual slide evaluation.
创建能够支持各种器官常规诊断的算法和应用工具。
开发了一种通用算法，可用于评估从苏木精-伊红染色的组织病理切片中获得的与诊断相关的图像特征。该程序在1442例各种器官的病例中进行了肿瘤组织与非肿瘤组织筛查测试。所研究的组织样本包括结肠、肺、乳腺、胸膜、胃和甲状腺。该算法区分基于纹理和基于对象的参数。基于纹理的信息（定义为每个像素测量的灰度值）独立于任何分割程序。它产生从时间序列分析以及通过空间相关和独立变换获得的图像特征得出的递归向量。基于对象的特征定义为每个测量的生物对象的灰度值。
基于每个诊断类别10例的学习集，自动粗略分类的准确率在95%至100%之间。结果与所分析的器官无关。该算法还可以区分结肠的良性和恶性肿瘤、上皮性间皮瘤和胸膜癌转移或不同类型的常见肺癌。
我们的算法能够准确区分各种器官的粗略组织学诊断。它是一种有前景的技术，可以辅助基于组织的诊断并扩展到虚拟切片评估。
