RICE: A method for quantitative mammographic image enhancement

• A novel and first of its kind method designed to enhance quantify focal densities in dense breast. The identifies normal parenchyma recursively subtracts it from the breast image, including above below a potential abnormality such as mass. This enables unmask tumors embeded (examples included). quantifies densities, thus enabling measure extent assymetry among bilateral mammograms. is particularly useful when applied time series mammographic images, where can potentially predict (a case study included) laterality cancer. content based image enhancement that uses measurable parenchymal contents breast, unlike existing methods relies on appearance mamogram enhances an by changing dynamic range, or performing histogram equalisation, low-pass filtering, regional contrast stretching, gamma correction etc. works with regular mammograms well other modalities Volpara SAR, synthetic generated DBT stacks. has extend images beyond mammography (example We introduce Region Interest Contrast Enhancement (RICE) identify It aims help radiologists: 1) enhancing images; 2) detecting regions interest (such densities) are candidate masses masked behind parenchyma. Cancer masking unsolved issue, density categories BI-RADS C D. RICE suppresses order highlight densities. Unlike modifying range image; actual tissue composition segments Volumetric Breast Density (VBD) maps into smaller then applies recursive mechanism estimate ‘neighbourhood’ for each segment. updates neighbourhood, encompassing tissue, piecewise constant component image. not only mass but also helps estimating density. In extensive experiments, all types most challenging category Suitably adapted, be used precursor any computer-aided diagnostics detection system.