The objective of this work is to design and construct computing systems for the real-time reconstruction of computed tomography (CT) and magnetic resonance (MR) images. When coupled to existing acquistion systems capable of rapidly acquiring CT or MR data, the possibility of real-time CT and or real-time MR imaging is realized. The clinical significance of these developments lies in the ability to monitor in real-time the dynamic events occuring within the human body, as well as the ability to monitor in real- time the efficacy of procedures, such as the injection of contrast material or the positioning of interventional probes, involved in acquiring the study. This research program represents a continuation and expansion of a research effort that has successfully demonstrated the first real-time reconstruction of CT imagery. A major portion of the proposed research is directed at improving the existing system. This system performs the necessary filtering and back-projection operations optically. Experience has indicated that a digital electronic implemention of the filtering step is preferred and will be developed. The optical back-projector works admirably. Its capability will be extended to include fanbeam projection data. An interface will be developed to allow fast scanning systems to access the system. Another major aspect of the research will be directed at demonstrating the real-time reconstruction of MR images. Real- time MR imaging is possible using a fast dataacquistion method that supplies the Fourier transforms of projections. To reconstruct the images, an optical computer that employs the standard back-projector and a modified filtering system will be constructed. The more standard rapidscanning method for MR is the Echoplanar technique. Realtime reconstruction can be accomplished using a number of different acusto-optic architectures that we propose to explore. For both methods of real-time MR, the poor quality inherent in the data makes optics a reasonable technology to investigate. A final area of research involves the initiation of a collaborative effort to explore the potential of VLSI technology for performing some of these computational tasks. It is believed that the one-dimensional operations such as filtering and Fourier transformation are amenable to a VLSI solution. Existing expertise and facilities will allow for the design, layout, and fabrication of these circuits.
