DIGITAL MAMMOGRAPHY FOR NEEDLE GUIDANCE AND SPOT VIEWS

点击翻译按钮获取中文摘要

This research is aimed at developing a 'digital mammographic cassette,' in a collaborative effort between the University of Massachusetts Medical Center, Loral Fairchild Imaging Sensors, and the NASA/Goddard Space Flight Center, with consultants from the Jet Propulsion Laboratory, BBN Corporation and the Louisiana Space Consortium at Louisiana State University. This revised application is submitted in response to PA #PA- 94-020 (NIH Guide V.22, No.45, Dec. 1993), co-sponsored by NCI and NASA. The main objective is to develop a high performance module which will bring high quality digital imaging capability to the average mammographic facility without major modification of existing equipment. Our approach is very unique. A very large CCD (62 mm x 62 mm, 2048 x 2048 pixels) with 30 micron square pixels will be incorporated into an enclosure which is the same size as a standard film cassette. There will be no fiberoptic taper between the scintillator and the CCD which might reduce the contrast; instead, a 1-2 mm thick non-tapering fiberoptic plate will be used between the scintillator and the CCD. The use of this type of fiberoptic will dramatically increase the signal impinging on the CCD. This will enable us to increase the spatial resolution from 8-9 lp/mm (in the current CCD-based mammographic cameras) to 13-14 lp.mm in the proposed system without a loss in the signal-to-noise ratio and attainable contrast. The potential increase in the spatial resolution will be explored by running the CCD in the 15-micron pixel mode with a better scintillator. The cassette will be connected to the computer via a thin electrical cable. Unlike other mammographic CCD cameras, the proposed camera is intended to be used without vacuum for thermal isolation or cooling. The CCD will be read out at 8 MHz for a total readout of 0.5 sec. The design of such a system presents special engineering challenges, but it is likely to have a far-reaching impact in the practice of mammography. Currently, some small-field CCDs are attached to stereotactic devices, but they are not easily adaptable for spot views. Moreover, their image quality has been questioned when imaging very subtle lesions. It is anticipated that this digital mammography cassette will deliver diagnostic quality images not only for localization but for diagnostic spot views. This removable cassette will be compatible with all existing mammographic equipment, and it will be inserted into the film cassette slot. The characterization of this imaging module will be critical for the design of a similar full-field cassette which will cover the entire breast. The engineering and testing of this cassette will generate important data, not only for us but for other investigators designing digital mammography systems. Our proposed module is an 'engine' which will also be adaptable to the slot scanning techniques used by other investigators.