Breast cancer is presently the most common form of malignancy in women. X-ray mammography is the most effective and widely used approach for its early detection. It is also one of the most demanding medical X-ray imaging techniques, because it deals with highly radiosensitive tissue structures of very low contrast and small size. In order to effectively diagnose clinically occult malignancies, it is necessary to resolve tiny (< 0.5 mm), dense, irregularly shaped microcalcifications. The quality and clinical utility of mammograms is currently limited by the film-screens and phosphors used to convert X-rays to visible light. There exists a fundamental tradeoff between screen thickness (and therefore efficiency) and spatial resolution due to lateral light spreading. Modern digital imaging techniques overcome some of the limitations inherent to film-screens, but are still subject to this fundamental tradeoff. In Phase I we fabricated a prototype of a novel imaging sensor based on a structured CsI scintillating plate coupled to a large area charge- coupled device (CCD). This microstructured X-ray sensor suppresses lateral spreading of the scintillation light even when the structure is made very thick. In comparison with standard phosphors using mammographic X-rays, a significant improvement in image resolution is observed with no loss in efficiency. Based on these Phase I results, we have prepared a detailed Phase II plan to conduct the research and development necessary to construct a commercial mammography system which will have the capability of providing dramatic improvements in image quality and reduced patient dose compared to currently available technology.
