Pattern analysis of cell micronuclei images to evaluate their use as indicators of cell damage

Early biological changes that can be associated with disease are important indicators or biomarkers for the development of preventive screening strategies. Epidemiological studies have shown that the presence of chromosome damage or instability in human lymphocytes could be considered as an indicator of cancer risk. Chromosome damage can also be estimated using the micronuclei (MN) assay. MN are nuclear bodies originated by chromosome breakage or chromosome segregation during cell division. MN assay can be performed in epithelial tissues in direct contact with xenobiotics and carcinogens, becoming an indicator of cancer risk. Cell MN are thus geometric configurations that appear in the cell cytoplasm as small round bodies near the cell nucleus after cell division. Scoring micronuclei requires a trained individual to detect and count MN in approximately 3000-5000 cells from images in a microscope, becoming a tedious, subjective and error-prone task. In this paper we describe automated detection and counting techniques using digital image processing and pattern recognition, allowing automated detection and quantification of the cellular micronuclei configurations, making this technique much more effective for fast and reliable assessing of DNA damage by exposure to radiation and toxic substances.