CHROMATIN STRUCTURE-ANALYSIS BASED ON A HIERARCHICAL TEXTURE MODEL

The quantification of chromatin structures is an important part of nuc lear grading of malignant and premalignant lesions. in order to achiev e high accuracy, computerized image analysis systems have been applied in this process. Chromatin texture analysis of cell nuclei requires a suitable texture model. A hierarchic model seemed to be most compatib le for this purpose. It assumes that texture consists of homogeneous r egions (textons). Based on this model, two approaches to texture segme ntation and feature extraction were investigated using sections of cer vical tissue. We examined the reproducibility of the measurement under changing optical conditions. The coefficients of variations of the te xture features ranged from 2.1% to 16.9%. The features were tested for their discriminating capability in a pilot study including 30 cases o f cervical dysplasia and carcinoma. The overall classification accurac y reached 65%. This study presents an automated technique for texture analysis that is similar to human perception.