Roughness feature of metaphase chromosome spreads and nuclei for automatedcell proliferation analysis

As a step towards automation of mitotic index estimation for cell prolifera tion studies, a roughness feature of surface-intensity images is introduced : the mean depth-width ratio of extrema (MDWRE). This feature allows identi fication of variable-shaped metaphases and interphase nuclei in the presenc e of many artefacts (one metaphase per hundreds of nuclei and thousands of artefacts). The texture of the cytological objects (seen as rough surfaces) is quantified by scanning, in one dimension, the lines contained in a clos ed contour. MDWRE proves to be suitable for image magnifications by a facto r of as low as ten, making faster scanning of slides possible. The use of t his feature gives + 14%, + 65%, + 133% and + 133% better performance figure s than classical textural features derived from co-occurrence matrices, suc h as contrast, energy, entropy and angular second moment, respectively, and + 51% better than the relative extrema density (RED). The MDWRE per object and the shape of the histogram of the depth-width ratio of grey-level roug hs have been shown to be very useful as textural features for the classific ation of metaphase images.