Project 17: Knowledge discovery in perioperative databases Organization: Vanderbilt University, Biomedical Abstract Engineering/Anesthesiology Primary Contact: Russ Waitman Knowledge Discovery in Perioperafive Databases: Address: 504 Oxford House Hypothesis Testing, Decision Support, and Care Vanderbilt University Pathway Assessment. Previously merged C4.5 with Nashville TN 37232-4125 expert guided inductive learning. Currently researching Email: waitman@vuse.vanderbilt.edu the use of BRUTE force induction to discover URL: relationships in perioperative data. Will be analyzing http://www.vuse.vanderbilt.edu/-waitman/persinfo.htm physiological data from patients in the operating room as well as information contained in a perioperative Duration: 3 years database. Number of People: 4 Tools Developed: MIKE Keywords: anesthesiology, pefioperative medicine, Academic Disciplines: Computer Science, Anesthesiology, inductive learning Biomedical Engineering, and Electrical Engineering Funding Sources: Department of Anesthesiology Project Related Publications: Tsai YS. King PH. Higgins MS. Pierce D. Patel NP, "An Expert-Guided Decision Tree Construction Stlategy: An Application in Knowledge Discovery with Medical Databases", Proceedings of AMIA Annual Fall Symposium, 208212, 1997. Russ Waitman's Ph.D. dissertation (in progress). Project 18: Prediction and knowledl~e mining with in vitro fertilization data Organization: University of Toronto, Faculty of Abstract Information Studies Primary Contact: I. Jurisica TA3 is a case-based reasoning system with knowledge Address: University of Toronto discovery (KD) capabilities. Faculty of Information Studies TA3 uses efficient and flexible variable context 140 St. George Street similarity-based retrieval. Toronto, Ontario M5S 3G6 Canada KD component is used for 3 main tasks: Email: j uris@ai.utoronto, ca 1. TA3 optimization: locating descriptors relevant for URL: http://www.cs.utoronto.ca/-juris a given context & task; and organizing case base into context-based clusters Duration: 2.5 years 2. Case base and domain knowledge evolution: Number of People: 6 (at present) adding descriptors to assist case discrimination Tools Developed: TA3 during prediction and classification; removing redundant cases and descriptors; creating Academic Disciplines: computer and information sciences, hierarchies of descriptors and their values; finding medicine Funding Sources: Natural Sciences and Engineering ass6ciations 3. Evidence-based medicine: analyzing created Research Council of Canada clusters, hierarchies and associations to identify (NSERC); Communications and underlying principles in the domain Information Technology Ontario (CITO) Keywords: in vitro fertilization; prediction; image analysis; knowledge evolution; Project Related Publications: J. Jurisica, J. Mylopoulos, J. Glasgow, H. Shapiro, and R. F. Casper, "Case-Based Reasoning in IVF: Prediction and Knowledge Mining," Journal ofArtificial lntelligence in Medicine, 12, 1-24, 1998. J. Glasgow and I. Jurisica, "Integration of Case- Based and Image-Based Reasoning," Presented atA,~L41'98 Workshop on Case-BasedReasoning, Madison, WI, 1998.

/lp/association-for-computing-machinery/prediction-and-knowledge-mining-with-in-vitro-fertilization-data-Tj8tB6Aobx