A Study of Multidimensional Multicolor Images

TLDR: A machine vision system for automatic characterization of sections of biopsied tissue capable of recognizing normal liverBiopsies and liver biopsies showing alcoholic hepatitis and acute viral hepatitis is described.

Abstract: This paper describes a machine vision system for automatic characterization of sections of biopsied tissue. Color images are first segmented by using an original color segmentation procedure. A structural analysis is then performed over the segmented images to extract the same features that a pathologist would use to make a diagnosis. Alternatively, monochromatic images can be analyzed by using cellular logic techniques. These two methods were used to implement a machine vision system capable of recognizing normal liver biopsies and liver biopsies showing alcoholic hepatitis and acute viral hepatitis. The color segmentation procedure leads to 100 percent classification success in distinguishing between the normal and acute viral hepatitis classes. By using cellular logic techniques, no overlap was found between the normal, acute viral hepatitis, and alcoholic hepatitis classes at the 90 percent confidence interval.