Along with tasks that require reasoning with medical knowledge, AI systems also have a very different role to play in the process of scientific research. In particular, AI systems have the capacity to learn, leading to the discovery of new phenomena and the creation of medical knowledge. For example, a computer system can be used to analyse large amounts of data, looking for complex patterns within it that suggest previously unexpected associations. Equally, with enough of a model of existing medical knowledge, an AI system can be used to show how a new set of experimental observations conflict with the existing theories. - from Artificial Intelligence in Medicine, by Enrico Coiera

Good Places to Start

New Blood Test Spots Cancer - Could Be Available as Early as 2004. By Charlene Laino. WebMD Medical News (December 13, 2002). "In what's being called one of the biggest advances in cancer research in years, scientists have developed a blood test that can detect cancer with a greater than 90% accuracy. This artificial intelligence -- already tested for cancers of the breast, ovary, and lung -- could one day be used to detect many types of cancer. ... 'All that's needed [for the quick fingerstick test] is a single drop of blood,' [Emanuel] Petricoin says. 'The computer does the rest.' ... In tests on several hundred blood samples, some taken from women with ovarian cancer and others from healthy women, the test proved 'an astonishing' 100% accurate in detecting cancer, even at the earliest stages, Petricoin said."

Expert Systems - Make a Diagnosis. Part of It's Alive! - From airport tarmacs to online job banks to medical labs, artificial intelligence is everywhere. By Jennifer Kahn. Wired Magazine (March 2002; Issue 10.03). "Intuition may seem like a human trick, but machines can be pretty good at it, too. Underlying a hunch are dozens of tiny, subconscious rules - truths we've learned from experience. Add them up and you get instinct: a doctor's sense that a patient's stomachache might really be appendicitis, for example. Program those rules into a computer and you get an expert system - one of many that can screen lab tests, diagnose blood infections, and identify tumors on a mammogram."

Artificial Intelligence in Medicine. By Enrico Coiera, M.D., PhD. An online chapter of his 1997 book Guide to Medical Informatics, the Internet, and Telemedicine.

• The related page, Artificial Intelligence Systems in Routine Clinical Use, which offers "summaries of 38 Artificial Intelligence systems that have been in routine use in medical settings," is now part of the OpenClinical site.

Robots in the OR -- Stat! Penelope the robot may free nurses to do more "human" tasks. By Josh Chamot. National Science Foundation Discoveries (April 28, 2005). "As the decade unfolds with its shortage of nurses, the sheer volume of patients each nurse must care for is leading to a critical burden for each of these professionals. While nurses will always be crucial to the care of patients, certain jobs may soon be accomplished by sophisticated robots. Surgeon Michael R. Treat and his team at Robotic Surgical Tech, Inc. have developed a robotic surgical assistant, named 'Penelope,' to perform tasks usually assigned to the scrub nurse. All of Penelope's talents are made possible by the innovative application of artificial intelligence to surgical situations." See the related video: Surgeon and "Penelope" in the operating room.

• Related article: For Surgery, an Automated Helping Hand. By Marc Santora. The New York Times (January 18, 2005; registration req'd.). "'Meet Penelope,' Dr. [Michael R.] Treat said, motioning toward a robotic arm poised over a set of surgical tools. ...She is meant to replace the scrub nurse, the person in the operating room who hands the surgeon the tools of surgery. Responding to the ever-widening shortage of nurses in the country, and looking to deal with a problem that frustrated him as a working surgeon, Dr. Treat and his team of tech whizzes are working feverishly to get Penelope ready for her public debut. New York-Presbyterian Hospital has agreed to test Penelope in March in the operating room on a simple removal of a benign cyst. ... Some of Penelope's technology is off the shelf, like the voice recognition software. Dr. Treat said that this way, as others develop better software, they can update Penelope with relative ease. The major innovation is in Penelope's visual recognition, the ability to distinguish between surgical tools. Currently, Penelope can recognize 12 tools and will soon be able to recognize twice that many. That is harder then it might sound, because the tools often look very much alike."
• Visit Robotic Surgical Tech, Inc., developers of Penelope™, "the world's first vision-guided, autonomous surgical robot. The Penelope system will save hospitals money and improve quality of care in the operating room."

Meet Adele, a pedagogical agent from the Center for Advanced Research in Technology for Education (CARTE). "Adele helps the faculty of USC's Department of Family Medicine deliver Continuing Medical Education (CME) courses on the Web. Simulations created for the course in diagnostic skill development will present the physician/student with actual cases, including patient history, results of exams, lab tests, x-rays, CT scans and other diagnostic imaging methods. By questioning and examining the virtual 'patient' and studying clinical data, the student will be able to practice diagnostic skills. Adele will provide feedback and a review of the student's progress, referencing diagnostic best-practice and cost-analysis criteria."

• A demo is offered on Adele's page, with others to be found on the main demo page.

Relational Agent Research being conducted by Timothy Bickmore, Assistant Professor of Medicine, Medical Information Systems Unit, Boston University School of Medicine. "I'm interested in building computer agents that help people through health behavior change and chronic disease self management interventions, and that assist in health education." Be sure to see his Relational Agents page.

Robots Make the Rounds To Ease Hospitals' Costs - VA Experience May Herald New Uses for 'Droids.' By Susan Okie. Washington Post (April 3, 2002). "Stationary robots and those that roll along tracks or wires are used in many industries, but independently mobile robots that interact with human co-workers or the general public are still relatively uncommon. Yet 'service robots,' designed to perform mundane jobs such as delivering drugs, food trays and laboratory specimens, are increasingly being employed in hospitals, which must operate 24 hours a day and face severe labor shortages and high costs for personnel."

Readings Online

AI in Medicine: The Spectrum of Challenges from Managed Care to Molecular Medicine. By Russ B. Altman, M.D., Ph.D. (1999). AI Magazine 20(3): 67-77. "AI has embraced medical applications from its inception, and some of the earliest work in successful application of AI technology occurred in medical contexts. Medicine in the twenty-first century will be very different than medicine in the late twentieth century. Fortunately, the technical challenges to AI that emerge are similar, and the prospects for success are high."

Meet your new therapist - your understanding computer. An innovative new computer program called Beating the Blues claims to be able to help patients suffering from depression and anxiety. But can a machine really be a replacement for face-to-face treatment? By Steve Boggan. The Independent (January 16, 2002).

The Effect of Patient and Physician Reminders on Use of Screening Mammography in a Health Maintenance Organization: Results of a Randomized Controlled Trial. By R. C. Burack, P.A. Gimotty, J. George, et al. (1996). Cancer 78 (8): 1708-1721. Link to article abstract.

Effectiveness of Computer-Generated Reminders for Increasing Discussions about Advance Directives and Completion of Advance Directive Forms. By Paul R. Dexter, Fredric D. Wolinsky, Gregory P. Gramelspacher, et al. (1998). Annals of Internal Medicine 128 (Janurary 15): 102-110.

Safe and Sound: Artificial Intelligence in Hazardous Applications. By John Fox and Subrata Das. AAAI Press. The following excerpt is from the Preface which is available online: "This book is about the nature of cognition, both natural and artificial. It has grown out of a program of research into intelligent functions like reasoning, problem solving and decisionmaking. These are well-established research topics, but our program is unusual in its focus on the integration of these and related cognitive processes. Many cognitive scientists seek a unified theory of their subject matter but, as in many other fields of scientific enquiry, the discipline tends to fragment into more and more specialist areas and unification eludes us. Our long-term aim is to develop intellectual and methodological tools that will foster a unified cognitive science. ... Medicine provides the specific context for our discussions and practical demonstrations. As medical researchers, as well as cognitive scientists, our aim has been to create technologies which can help nurses, doctors and other medical professionals make clinical decisions when managing the diagnosis and treatment of life-threatening conditions, such as cancer."

Medicine on Mars - How sick can you get during three years in deep space? By Jerome Groopman. The New Yorker (February 14, 2000: pages 36 - 41; available at the author's web site). "There might also be a 'virtual mentor': a computerized database whose artificial intelligence will assist the onboard physician in diagnosis and treatment. 'For example, astronaut John Doe has a fever, and pain in his right lower abdomen,' [Jon] Bowersox says. 'The virtual mentor instructs the doctor to perform an ultrasound, with voice and image overlay, all the while comparing the new ultrasound data with baseline data. Appendicitis, which is the likely diagnosis, would be best treated with intensive antibiotics, given the difficulties of operating in zero G. Surgery is a last resort.' When surgery could not be avoided, it would be performed by the doctor, coached by the virtual mentor and aided by robotics."

Doctors' Orders - As expert systems become more expert, physicians see the advantages of taking on CPOE. By Mark Hagland. Healthcare Informatics (January 2003). "If the future is a clinically driven, elegantly managed, intelligent system--one that helps physicians optimize patient care and helps hospitals and medical groups constantly improve care delivery--then the folks at Ohio State University (OSU) Health System, Columbus, have at least glimpsed the future. They're in the vanguard of patient care organizations using computerized physician order entry (CPOE) systems which, authorities say, best represent the legacy of years of expert-systems development work in healthcare."

LifeCode: A Deployed Application for Automated Medical Coding. By Daniel T. Heinze, Mark Morsch, Ronald Sheffer, Michelle Jimmink, Mark Jennings, William Morris, and Amy Morsch. AI Magazine 22(2): 76-88 (Summer 2001). This paper is based on the authors' presentation at the Twelfth Innovative Applications of Artificial Intelligence Conference (IAAI-2000). "LifeCode is a natural language processing (NLP) and expert system that extracts demographic and clinical information from free-text clinical records."

Machines and medicine. Tim Lougheed. University of Ottawa Gazette (February 2, 2001). "Consider a health care assistant who is on duty 24 hours a day, seven days a week. One who can stay by the bed of a patient and collect information accurately, tirelessly and in comprehensive detail. One who can record and compile those details, comparing them with thousands or even millions of other cases, looking for clues to the common causes of illnesses and recommended courses of action. Such assistants exist. They are ready, willing, but not quite able to carry out such duties. They are the increasingly powerful computer systems that can interact with equipment used to monitor patients in hospitals. Doctors in all branches of medicine are increasingly eager to enlist these computers as assistants in their work. The only problem has been training those assistants for the job. Dr. Robin Walker, a professor in the Department of Pediatrics and chief of the division of neonatology at the Children's Hospital of Eastern Ontario, is one of the leaders in a new project to carry out this sort of training. ... [H]e will collaborate with Monique Frize, an electrical and computer engineering professor cross-appointed to the University of Ottawa and Carleton University. ... They will apply artificial intelligence technology to the clinical challenges of diagnosing and treating patients, in partnership with the Canadian Neonatal Intensive Care Unit Network."

• See a related article from the Ottawa Citizen (June 24, 2004).

Technology in Clinical Practice: Computer-Based Therapy and Radio Frequency Identification. By John Luo, M.D. Psychiatric Times (October 2005; Vol. XXII, Issue 12). "[T]here are significant numbers of patients who will not seek mental health care. These patients often turn to self-help materials available online or in the bookstore. With today's sophisticated computers and the large amount of storage available on DVDs, computer-based assessment and treatment software is becoming more commonplace. These programs are quite sophisticated, utilizing specialized heuristic techniques to produce a more natural response to patient-entered information. Natural language processing allows software to determine the nature of queries and responses. ... There are numerous advantages to computer-based therapy. Patients can work at their own pace in the privacy of their own home. ... In addition, software programs are not susceptible to the frustrations of traditional treatment such as changed appointments, therapist boredom, absences and ethical misconduct. ... Computer-based therapy programs offer many benefits with limited risks. While they are not touted as replacements for traditional therapy, they are very useful supplements."

Medical Diagnostic Decision Support Systems - Past, Present, and Future: A Threaded Bibliography and Brief Commentary. By Randolf A. Miller (1994). JAMIA (Journal of the American Medical Informatics Assn.) 1 (1): 8-27.

Stanford Medical Informatics: Uncommon research, common goals. By Mark A. Musen. MD Computing (January/February 1999). "The DENDRAL project continued into the 1970's, when a Stanford medical student names Ted Shortliffe became interested in applying the knowledge-based paradigm to clinical problems. In collaboration with Bruce Buchanan (who at the time was a research associate on the DENDRAL project) and several prominent clinical faculty members, Shortliffe built MYCIN, the first convincing demonstration of the power of the rule-based approach in the development of robust clinical decision-support systems."

A Randomized Trial of "Corollary Orders'' to Prevent Errors of Omission. By J. Marc Overhage, William M. Tierney, Xiao-hua (Andrew) Zhou, et al. (1997). JAMIA (Journal of the American Medical Informatics Assn.) 4 (5): 364-375.

Use of proteomic patterns in serum to identify ovarian cancer. By Emanuel F. Petricoin III, Ali M. Ardekani, Ben A. Hitt, et al. (2002). Lancet 2002; 359: 572-77.

• See the press release from The Lancet
• Ovarian Cancer - A New Test May Catch the 'Silent Killer' Before It Spreads. Reuters / available from ABCNEWS.com (February 8, 2002).
• Scientists Report Initial Success With a Blood Test for Ovarian Cancer. By Denise Grady. The New York Times (February 8, 2002; free reg. req'd.).

Computer-Generated Informational Messages Directed to Physicians: Effect on Length of Hospital Stay. By Steven Shea, Robert V. Sideli, William DuMouchell, et al. (1995). JAMIA (Journal of the American Medical Informatics Assn.) 2 (1): 58-64.

Artificial Intelligence in Medicine. Peter Szolovits, editor. 1982. (AAAS Selected Symposia Series, Volume 51) "This book has been out of print since the early 1990's, though it is still often available through book search services, including the usual on-line ones. The original book was relatively popular for a symposium book; it was re-published in paperback and went through several printings. From the vantage point of nearly twenty years after its publication, I believe that many of the ideas in the chapters are still vibrant. ... I have reconstructed this volume in HTML and made it available on the Web in the hope that it will inspire new researchers to learn about some of the elegant older work and to take up the challenges not yet met."

Speech recognition technology finds a voice. Discussion at SCAR [Symposium for Computer Applications in Radiology] gets down to nuts, bolts, and templates in latest element of radiology's technology arsenal. By Merlina Trevino. SCAR Conference Reporter (2002). "It's been one of the most popular Diagnostic Imaging PACSpoll questions: Are you ready to replace traditional transcription services with speech recognition technology? More than 50% of the poll respondents answered with a resounding yes."

Evaluation of a Decision Support System for Initiation and Control of Oral Anticoagulation in a Randomised Trial. By V. Vadher and D.L.H. Patterson 1997. BMJ (British Journal of Medicine 314: 1252-1256. The CDSS improved the quality of initiation and control of warfarin treatment by trainee doctors.

Related Web Sites

AI Systems in Clinical Practice. From OpenClinical. "Archive originally developed and administered by Enrico Coiera ... Now maintained by OpenClinical." The AI systems have been organized into the following categories: Acute care systems, Decision support systems, Educational systems, Laboratory systems, Medical Imaging, and Quality assurance and administration.

• Also see: Artificial Neural Networks - An Introduction which includes a link to Margarita Sordo's article, Artificial Neural Networks in Healthcare.

AIM Program: Artificial Intelligence in Medicine at Cedars-Sinai Health System.

AIME 05: The European Society for Artificial Intelligence in Medicine's (AIME) 10th Conference on Artificial Intelligence in Medicine. 23 - 27 July 2005. Aberdeen, Scotland.

American Medical Informatics Association. Conferences, publications, etc.

• Search the Journal of the American Medical Informatics Association. Content: Full Text: Jan/Feb 1997 - Present ; Abstracts only: Jan/Feb 1994 - Nov/Dec 1996.

Artificial Intelligence in Medicine. This journal, published by Elsevier, contains "original articles from a wide variety of interdisciplinary perspectives concerning the theory and practice of artificial intelligence (AI) in medicine, human biology, and health care." You can access the tables of contents and abstracts, but articles are available only to subscribers.

Artificial Neural Networks in Medicine World Map [of projects, labs, articles, people and more]. Maintained by Daniel Ruhe.

Biomedical Security Institute. "BMSI has a two-pronged approach: 1.) Development of a prototype computer-based surveillance, analysis and communication systems infrastructure to provide early warning of naturally occurring disease outbreaks and terrorist attacks employing biological pathogens. It will perform continuing real-time data mining and analysis of selected data streams (such as electronic medical records and microbiology laboratory results) for sentinel events or situations. ... 2.) A research program that integrates the latest research and technology in public health, biomedical sciences and biomedical informatics, including: ... Intelligent systems, data mining, artificial intelligence, bioinformatics and early detection/early warning computer-based surveillance systems...."

The Center for Human Simulation (CHS) at the University of Colorado. Home of the digitalized knee.

"The Clinical Decision Making Group at the MIT Laboratory for Computer Science is a research group dedicated to exploring and furthering the application of technology and artificial intelligence to clinical situations."

Decision Systems Group at Harvard Medical School and Brigham and Women's Hospital. Research and development work includes Representation of Medical Knowledge, Decision Support Systems, Machine Learning Models, Geometry and Image Based Reasoning, and Bioinformatics.

Division of Clinical Decision Making, Informatics and Telemedicine at the Tufts - New England Medical Center: "Research - Applying decision analysis, utility assessment, literature synthesis, medical informatics, and artificial intelligence in medicine, the Division focuses on clinical decision analysis, cost-effectiveness and health policy analysis at institutional and policy levels. ... The Division uses the following techniques for decision analysis in medical decision making: decision tree construction, Markov model development, Monte Carlo simulation, Bayesian interpretation of diagnostic tests, the measurement of patient preferences, cost-effectiveness analysis, literature review, meta-analysis and discrete event simulation. ... Members of the Division are actively involved in exploring and developing platforms for decision support and decision analysis, as well as techniques for assessing patient preferences and incorporating them into clinical decision making. ... Decision Maker - This program for performing decision analyses on MS-DOS computers has been in use for over a decade. The current version is 7.07. Using Decision Maker, the decision analyst can analyze simple decision trees or Markov cohort models. Monte Carlo simulation is also supported."

• Decision Maker^© home page.

Guardian: "intelligent agents for monitoring medical patients". From Adaptive Intelligent Systems in the Knowledge Systems Laboratory at Stanford. "The Guardian project was completed in 1996. This page contains references to some of the original Guardian materials and group members."

HealthDoc Project. "Tailored Patient Education Systems. The HealthDoc project is developing natural language software systems for producing, on demand, health-information and patient-education brochures that are tailored to the medical condition and personal requirements of the individual patient."

"HelpMate [from Pyxis Corporation] is a trackless, robotic courier system designed to perform material transport tasks throughout the hospital environment. Twenty-four hours a day, 365 days a-year, HelpMate transports pharmaceuticals, lab specimens, equipment and supplies, meals, medical records and radiology films back and forth between support departments and nursing floors."

"ISIS is a decision support system that helps physicians select the most cost-effective diagnostic imaging studies. From MIDAS, The Medical Informatics and Decision Science Consortium Milwaukee, Wisconsin.

The Informatics Review, the E-journal of the Association of Medical Directors of Information Systems. Be sure to check out their collection of links.

KMeD. Knowledge-Based Multimedia Medical Distributed Database System from UCLA. The project's 5 objectives are: "1.Query medical multimedia distributed databases by both image and alphanumeric content, 2.Model the temporal, spatial, and evolutionary nature of medical objects, 3.Formulate queries using conceptual and imprecise medical terms and support cooperative processing, 4.Develop a domain-independent, high-level query language and a medical domain user interface to support KMeD functionality, and 5.Provide analysis and presentation methods for visualization of knowledge and data models."

• Be sure to check out the screenshots since they actually show what the system can do.

Medical Informatics FAQs. Maintained by Aamir Zakaria, M.D.

Medical Informatics Journal Club Classics. Maintained by Dean F. Sittig.

Medical Robotics and Computer Assisted Surgery (MRCAS) Lab at The Robotics Institute, School of Computer Science, Carnegie Mellon University. "Our research involves both planning aspects of computer-assisted surgery, e.g., medical image computing and surgical simulation, and execution aspects, e.g., intraoperative sensing, registration, and actuation." One of their many projects is "HeartLander - An inchworm-like mobile robot for minimally invasive beating-heart cardiac surgery."

Medical Vision Group. "The goal of our group is to develop new algorithms for medical image analysis and visualization of medical imagery, as well as to build vision based systems for surgical navigation and surgical planning. Our group at the MIT AI Lab has been collaborating closely for several years with the Surgical Planning Laboratory of Brigham and Women's Hospital. This page contains links to various projects carried on by the group, as well as the web pages of the group members."

Nursebot Project. "The goal of our project is to develop mobile, personal service robots that assist elderly people suffering from chronic disorders in their everyday life. We are currently developing anautonomous mobile robot that "lives" in a private home of a chronically ill elderly person."

Phibot. A research project of the University of Mainz, the German Research Center of Artificial Intelligence (DFKI) and brainbot technologies AG. "Phibot is an intelligent internet information retrieval tool for scientists."

"TheraDoc, Inc. is a medical informatics company that designs, markets and implements expert systems for clinical decision support. TheraDoc’s software is designed to help meet today’s health care challenges by providing clinicians with tools that enhance their abilities to deliver high quality, evidence-based care which aids to effectively reduce medical errors, lower treatment costs and improve the quality and safety of patient care."

The Third Workshop on Agents Applied in Health Care. 30 July 2005 Edinburgh, Scotland; held in conjunction with IJCAI2005. "

"TraumAID is a program developed over the past twelve years to assist physicians with the diagnosis and treatment of penetrating trauma (gunshot and stab wounds) to the chest and abdomen. The project has subsequently expanded to encompass other emergency medicine related endeavors. TraumAID is a cooperative effort between the Computer and Information Science Department of the University of Pennsylvania, and Allegheny University of the Health Sciences (formerly the Medical College of Pennsylvania and Hahnemann University)."

Related Pages

• AI in the news
• Expert Systems
• General Index to AI in the news: Applications
• Machine Learning
• Public Health & Welfare (includes resources about Biosurveillance)

More Readings

Balas, E.A., S.M. Austin, J.A. Mitchell, et al. 1996. The Clinical Value of Computerized Information Services. A Review of 98 Randomized Clinical Trials. Archives of Family Medicine 5 (5): 271-278. Automated information services improve clinical practice.

Buchanan, Bruce G., and Edward H. Shortliffe., editors (1984). Rule-Based Expert Systems --The MYCIN Experiments of the Stanford Heuristic Programming Project. Reading, MA: Addison-Wesley.

Casner, P.R., R. Reilly, and H. Ho 1993. A Randomized Controlled Trial of Computerized Pharmacokinetic Theophylline Dosing vs. Empiric Physician Dosing. Clinical Pharmacology and Therapeutics 53: 684-690.

Chambers, C.V., D.J. Balaban, B.L. Carlson, et al. 1991. The Effect of Microcomputer Generated Reminders on Influenza Vaccination Rates in a University-Based Family Practice Center. Journal of the American Board of Family Practice 4: 19-26.

Clancey, William J., and Edward H. Shortliffe, editors. 1984. Readings in Medical Artificial Intelligence : The First Decade. Reading, MA: Addison-Wesley.

Coiera, Enrico. 1997. Guide to Medical Informatics, the Internet and Telemedicine. New York: Chapman and Hall Medical.

Fihn, S.D., M.B. McDonell, and D. Vermes 1994. A Computerized Intervention to Improve Timing of Outpatient Followup: A Multicenter Randomized Trial in Patients Treated with Warfarin. Journal of General Internal medicine 9: 131-139.

Fitzmaurice, D.A., F.D.R. Hobbs, E.T. Murray, et al. 1996. Evaluation of Computerized Decision Support for oral Anticoagulation Management Based in Primary Care. British Journal of General Practice 46: 533-535.

Frame, P.S., J.G. Zimmer, P.L. Werth, et al. 1994. Computer Based vs. Manual Health Maintenance Tracking. Archives of Family Medicine 3: 581-588.

Giger, Maryellen, and Charles A. Pelizzari. 1996. Advances in Tumor Imaging. Scientific American (September 26, 1996): 110-12.

Hobbs, F.D., B.C. Delaney, A. Carlson, et al. 1996. A Prospective Controlled Trial of Computerized Decision Support for Lipid Management in Primary Care. Family Practice 13: 133-137.

Hunt, Dereck L., R. Brian Haynes, Steven E. Hanna, et al. 1998. Effects of Computer-Based Clinical Decision Support Systems on Physician Performance and Patient Outcomes. JAMA 280 (15): 1339-1346. Computer Based Decision Support Systems (CDSS) help doctors access information that specifically matches patient characteristics, a considerable improvement over having medical personnel conduct lengthy searches through electronic databases, and then through individual articles, to find relevant information. It is crucial, though, to evaluate the quality of CDSS's, and the authors find a happy increase in high-quality testing of systems in the last few years.

Lathrop, Richard H., Nicholas R. Steffen, Miriam P. Raphael, et al. 1998. Knowledge-Based Avoidance of Drug-Resistant HIV Mutants. In Proceedings of the Tenth Annual Conference on Innovative Applications of Artificial Intelligence, ed. Buchanan, Bruce and Ramasamy Uthurusamy, 1071-1078. Menlo, CA: AAAI. This system connects scientific AIDS literature describing specific HIV drug resistances to the Customized Treatment Strategy for specific HIV patients. The possible drug treatment regimens currently approved by the US Food and Drug Administration (FDA) are considered and ranked by their estimated ability to avoid identified current and nearby drug resistant mutants. The highest-ranked treatments are recommended to the attending physician, yielding precise treatment of individual patients, and a decreased tendency to select for drug resistant genes in the global HIV gene pool. The application is currently in use in human clinical trials on HIV patients. From an AI viewpoint the case study demonstrates the extensibility of knowledge-based systems, illustrating how previously encoded knowledge can support unanticipated applications.

Little, J. P., and Mark Gingrich. 1992. Adjudipro. In Innovative Applications of Artificial Intelligence 4, ed. Scott, A. Carlisle and Phillip Klahr, 249-264. Menlo Park, CA: AAAI.

Marmorstein, Howard, Jayesh Ghis, Sandeep Sathay, et al. 1997. PST: The Provider Selection Tool. In Proceedings of the Ninth Annual Conference on Innovative Applications of Artificial Intelligence, ed. Senator, Ted and Bruce Buchanan, 867-874. Menlo, CA: AAAI. Oxford Health Plans, Inc. is a managed healthcare organization. Oxford Members' satisfaction with their Primary Care Physician (PCP) is important to their relationship with the company. People traditionally choose a doctor by asking family and friends for recommendations. While this will most likely remain the most trusted method, Oxford aims to complement and also confirm the advice of family and friends with the Provider Selection Tool (PST). PST is a case-based reasoning tool that elicits search criteria from a Member to assess and evaluate a roster of Providers that meet those criteria. These Providers are then presented in ranked order.

McDonald, C.J., S.L. Hui, and W. M. Tierney 1992. Effects of Computer Reminders for Influenza Vaccination on Morbidity During Influenza Epidemics. MD Computing 9: 304-312.

McPhee, S. J., J. Adair Bird, D. Fordham, et al. 1991. Promoting Cancer Prevention Activities by Primary Care Physicians. JAMA 266: 538-544.

McPhee, S.J., and W.M. Detmer 1993. Office-Based Interventions to Improve Delivery of Cancer Prevention Services by Primary Care Physicians. Cancer 72: 1100-1112.

Miller, R. , and et. al. 1982. INTERNIST-1: An Experimental Computer-Based Diagnostic Consultant for General Internal Medicine. New England Journal of Medicine 307: 468-476.

Mungall, D., D. Anbe, and P. L. Forrester 1994. Clinic Trials and Therapeutics: A Prospective randomized Comparison of the Accuracy of Computer-Assisted vs GUSTO nomogram-directed heparin therapy. Clinical Pharmacology & Therapeutics 55: 591-596.

Ornstein, S. M., D.R. Garr, R.G. Jenkins, et al. 1991. Computer-Generated Physician and Patient Reminders: Tools to Improve Population Adherence to Selected Preventive Services. Journal of Family Practice 32: 82-90.

Overhage, J.M., W.M. Tierney, and C.J. McDonald 1996. Computer Reminders to Implement Preventive Care Guidelines for Hospitalized Patients. Archives of Internal Medicine 156: 1551-1556.

Poller, L., D. Wright, and M. Rowlands 1993. Prospective Comparative Study of Computer Programs Used for management of Warfarin. Clinical Pharmacology & Therapeutics 46: 299-303.

Pringle, M., S. Robins, and G. Brown. 1990. Computer Assisted Screening: Effects on the Patient and His Consultation. BMJ (British Journal of Medicine) 290: 1709-1712.

Rosser, W.W., I. McDowell, and C. Newell. 1991. Use of Reminders for Preventive Procedures in Family medicine. CMAJ (Canadian Medical Assn. Journal) 145: 807-812.

Rossi, R.A., and N.R. Every 1997. A Computerized Intervention to Decrease the Use of Calcium Channel Blockers in Hypertension. Journal of General Internal Medicine 12: 672-678.

Shortliffe, Edward H. 1995. Consultation Systems for Physicians: The Role of Artificial Intelligence Techniques. In Computation and Intelligence: Collected Readings, ed. Luger, George F., 511-527. Menlo Park/Cambridge, MA/London: AAAI Press/The MIT Press.

Tape, T.G., and J.R. Campbell 1993. Computerized medical records and Preventive Health Care: Succes Depends on Many Factors. American Journal of Medicine 94: 619-625.

Thomas, J.C., A. Moore, and P.E. Qualls. 1983. The Effect on Cost of Medical Care for Patients Treated with an Automated Clinical Audit System. Journal of Medical Systems 7: 307-313.

Turner, R.C., J.G. Peden, and K. O'Brien 1994. Patient-Carried Card Prompts vs. Computer-Generated Prompts to Remind Private Practice Physicians to Perform Health Maintenance Measures. Archives of Internal Medicine 154: 1957-1960.

Vadher, B.D., D.L.H. Patterson, and M. Leaning 1997. Comparison of Oral Anticoagulant Control by a Nurse Practitioner Using a Computer - Decision -Support - System with that by Clinicians. Clinical Laboratory Haematology 19: 203-207.

Verner, D., H. Seligman, and S. Platt 1992. Computer Assisted Design of a Theophylline Dosing Regimen in Acute Bronchospasm: Serum Concentrations and Clinical Outcomes. European Journal of Clinical Pharmacology 32: 29-33.

Wellwood, J., S. Johannessen, and D.J. Spiegelhalter 1992. How Does Computer-Aided Diagnosis Improve the management of Acute Abdominal Pain? Annals of the Royal College of Surgeons of England 74: 40-46.