A System for Automated Liver Tissue Image Analysis: Methods and Results

Image analysis techniques are described for the automatic detection, segmentation, feature description, and classification of tissue components, in liver tissue images. The result of the analysis is a quantitative description of the tissue, based on statistics of the components. A software package for analysis of these images has been developed by combining some newly developed algorithms with other techniques. The newly developed algorithms include the converging squares algorithm for location of cell nuclei, and the wedge filter technique for nuclei segmentation.