Accurate assessment of tissue specimens is a key requirement in the diagnosis of breast cancer. This requires images with high spatial and contrast resolution. A compact, high-resolution x-ray system for imaging of excised tissue specimens will be developed for use in the operating room. Currently, specimen radiography is performed in a cabinet x-ray unit where the image is acquired either on radiographic film or using a small area digital detector. Typically, the digital detector includes an x-ray absorbing phosphor, a fiber-optic coupler and an optical CCD camera. Such a detection system is limited in spatial and contrast resolution related to the use of phosphor technology and distortion and losses in the fiber-optic components. The new system will employ a novel direct conversion selenium CCD detector that will be used to form a slot scanning x-ray imaging system. It offers high x-ray detection and conversion efficiency, high limiting spatial resolution, low noise, linear response and high dynamic range. In the proposed system, the incoming x-rays are absorbed in a layer of selenium where their energy causes electric charge to be liberated. The selenium is directly evaporated onto a CCD readout chip. The charge is integrated in the readout chip and the signal is then digitized and sent to a computer for image processing and display. In Phase I the research plan consists of five goals. First, a small version of the CCD chip will be used in an experimental scanning system to simulate the full-scale specimen radiography system. To emulate a full-scale system, composite or """"""""stitched"""""""" images will be made from multiple scans with the small- format chip. Imaging will be performed on tissue-equivalent plastic test objects. Second, the optimal geometry for image acquisition will be determined. For example, depending on the angle and trajectory of the detector, distortions associated with parallax will potentially limit resolution. Third, the optimal x-ray spectral energy will be determined. Tube output, required imaging time, contrast and dynamic range for high quality images will be considered in the optimization. Once this is achieved, the fourth goal will be to evaluate the importance of scattered radiation on image quality through modeling of signal and noise transfer and through experimental imaging using tissue-equivalent phantoms. If necessary, an anti-scatter technique will be used to reduce scatter. Finally, based on the results of these experiments we will prepare a preliminary design of a prototype digital specimen imager system to be built during Phase II. During Phase II, the full-size optimized CCD readout chip will be fabricated and tested and then integrated into a selenium detector with optimized thickness. The new detector's imaging capability and performance will be characterized. The system is expected to provide improved definition of fine tissue structures and aid in determining that the targeted focus of disease has been appropriately removed in biopsy or surgery. This should help ensure improved accuracy of resection of disease and provide a better link between pre- operative imaging and tissue histology.
