The general objective of the Rutgers Resource is to apply advanced methods of computer science, particularly artificial intelligence, to problems of biomedical research and practice. The Resource promotes the development and use of computer systems for expert consultation in medical diagnosis and management, and for research assistance in processes of scientific experimentation and theory formation. Organizing and representing biomedical knowledge at different levels (descriptive, ruled-based mathematical submodels), and relating the levels through generalized strategies for communication and control with a user, constitutes a major component of our proposed reseach. We continue the present structure of the Resource, with a set of core projects concentrating on general artificial intelligence investigations, and a set of collaborative projects that provide the context within which the methods can be tested. The major areas of biomedical application are in rheumatology, where we are collaborating with Drs. D. Lindberg and G. Sharp at the National Library of Medicine and University of Missouri-Columbia, in clinical pathology, where we are working with Dr. Robert Galen of the Cleveland Clinic Foundation, and in ophthalmology, where we have an active collaboration with Dr. C. Dawson at the University of California-San Francisco. An important aspect of our research is technology transfer, producing prototyes that will be useful to clinical researchers and practitioners. We have pioneered developments in this field by putting an expert system in a widely used clinical instrument: a serum protein electrophoresis analyzer. We are proposing to continue these micro-based innovations for two applications: the interpretation of CPK/LDH isoenzymes for the diagnosis of heart attacks and a primary eye care program for developing countries. We also propose to continue our research in rule refinement and empirical analysis of expert system knowledge bases. In addition to the research activities of the Resource, we continue that highly successful dissemination activities of the AIM Workshop series and the Rutgers Resource participation in national AIM collaborative research.
|Kapouleas, I (1990) Automatic detection of white matter lesions in magnetic resonance brain images. Comput Methods Programs Biomed 32:17-35|
|Garfinkel, L; Cohen, D M; Soo, V W et al. (1989) An artificial-intelligence technique for qualitatively deriving enzyme kinetic mechanisms from initial-velocity measurements and its application to hexokinase. Biochem J 264:175-84|
|Garfinkel, D (1989) Constraint matching as a means of designing biochemical experiments in multi-enzyme systems. J Theor Biol 137:221-34|
|Soo, V W; Kulikowski, C A; Garfinkel, D et al. (1988) Theory formation in postulating enzyme kinetic mechanisms: reasoning with constraints. Comput Biomed Res 21:381-403|
|Van Lente, F; Galen, R S; Castellani, W et al. (1987) Sequential investigation of aspartate aminotransferase elevation in outpatients. Cleve Clin J Med 54:171-5|
|Van Lente, F; Castellani, W; Chou, D et al. (1986) Application of the EXPERT consultation system to accelerated laboratory testing and interpretation. Clin Chem 32:1719-25|
|Kingsland 3rd, L C; Lindberg, D A; Sharp, G C (1986) Anatomy of a knowledge-based consultant system: AI/RHEUM. MD Comput 3:18-26|