KNOWLEDGE-BASED MEDICAL IMAGE-ANALYSIS AND REPRESENTATION FOR INTEGRATING CONTENT DEFINITION WITH THE RADIOLOGICAL REPORT

Technology breakthroughs in high-speed, high-capacity, and high perfor mance desk-tip computers and workstations make the possibility of inte grating multimedia medical data to better support clinical decision ma king, computer-aided education, and research not only attractive, but feasible. To systemically evaluate results from increasingly automated image segmentation it is necessary to correlate them with the expert judgments of radiologists and other clinical specialists interpreting the images. These are contained in increasingly computerized radiologi cal reports and other related clinical records. But to make automated comparison feasible it is necessary to first ensure compatibility of t he knowledge content of images with the descriptions contained in thes e records. Enough common vocabulary, language, and knowledge represent ation components must be represented on the computer, followed by auto mated extraction of image-content descriptions from the text, which ca n then be matched to the results segmentation is essential to obtain t he structured image descriptions needed for matching against the exper ts descriptions. We have developed a new approach to medical image ana lysis which helps generate such descriptions: a knowlege-based object- centered hierarchical planning method for automatically composing the image analysis processes. The problem-solving steps of specialists are represented at the knowledge level in terms of goals, tasks, adn doma in objects and concepts separately from the implementation level for s pecific representations of different image types, ang generic analysis methods. This system can serve as a major functional component in inc rementally building and updating a structured and integrated hybrid in formation system of patient data. This approach has been tested for ma gnetic resonance image interpretation, and has achieved promising resu lts.