Computer graphic displays for microscopists assisting in evaluating radiation-damaged cells

Computer analysis of digitized images of cells and tissues has developed into a highly refined methodology. In particular, the nuclear chromatin has numerous features that allow reliable recognition and classification of cells. This was hardly a surprise to cytopathologists, since chromatin texture and staining properties have always provided valuable clues for visual diagnosis. However, two aspects of computer assessment came as a surprise. First, there was the consistency with which chromatin distribution patterns express the state of the cell, its differentiation, and the sensitivity with which a change in the cell's environment is reflected in the chromatin. Second, chromatin texture provides highly discriminating, objectively measurable features that are not perceived by human visual assessment. Numerous studies have shown that computer assessment is capable of making clear-cut discriminations in cases where human cytodiagnostic evaluation is not effective