HST 947: Medical Artificial Intelligence

Fall term 2003

Instructors:       Peter Szolovits, psz@mit.edu, NE43-416, (617) 253-3476,
Lucila Ohno-Machado, machado@dsg.bwh.harvard.edu, (617) 732-8543
Secretary: Fern DeOliveira, fern@medg.lcs.mit.edu, NE43-417, (617) 253-5860.

Meets: 6.034 lectures, recitations and tutorials, plus Wednesdays 3:15-4:30pm (right after the Wednesday 6.034 lecture) in the lounge outside NE43-416.

An intensive introduction to artificial intelligence and its applications to problems of medical diagnosis, therapy selection and monitoring. The class meets with lectures and recitations of 6.034, whose material is supplemented by additional readings and discussion sessions. Students are responsible for completing all homework assignments in 6.034 and for additional problems and/or papers. This class is available for credit only to graduate students in HST. It carries 12 units (5-3-4) of H-LEVEL Graduate Credit.

At each week's meeting of HST947, we will discuss a topic in the medical applications of AI, based on papers that are to have been read before the Wednesday meeting of that class.  We will attempt to have the papers available on-line at this site, usually as Adobe PDF files.  (Click on the Adobe logo to download a copy of the free reader for PDF.) Those papers whose PDF files contain simply a scanned image of the original are, unfortunately, rather large.  You should probably only download them via a high-speed internet connection.  Papers to which we do not hold copyright are in a special subdirectory that will ask you to authenticate your belonging to this class before allowing you access.  The user name and password were announced at the first class meeting.

Schedule of Classes and Readings

Note: You should complete the readings for a class before the corresponding class session. The schedule below is tentative (initially, a copy of last year's), and will change as we move through the semester.

Class Topic Reading Assignments
Sept. 3, 2002 Organizational Meeting  
Sept. 8 Scheme in one gulp; Special class meeting Monday morning, 9-11am. (Optional) Please refer to Learning Scheme.
Sept. 10 Rule-Based Expert Systems

Davis, R., B. G. Buchanan and E. H. Shortliffe (1977). "Production Rules as a Representation for a Knowledge-Based Consultation Program." Artificial Intelligence 8: 15-45.
Scott, A. C., W. J. Clancey, R. Davis, et al. (1984). Methods for Generating Explanations. Rule-Based Expert Systems. B. G. Buchanan and E. H. Shortliffe. Reading, MA, Addison-Wesley: 338-362.

Sept. 17 Explanation Generation

Wallis, J. W. and E. H. Shortliffe (1984). Customized Explanations Using Causal Knowledge. Rule-Based Expert Systems. B. G. Buchanan and E. H. Shortliffe. Reading, MA, Addison-Wesley: 371-388.
Swartout, W. (1983). XPLAIN: a System for Creating and Explaining Expert Consulting Programs. Artificial Intelligence 21:285-325.

Sep. 24 Diagnosis by Pattern Matching and Search

Pople, H. E., Jr. (1982). Heuristic Methods for Imposing Structure on Ill-Structured Problems: The Structuring of Medical Diagnostics. Artificial Intelligence in Medicine. P. Szolovits. Boulder, Colorado, Westview Press: 119-190.

Oct. 1 Hypothetico-Deductive Reasoning

Pauker, S. G., G. A. Gorry, J. P. Kassirer, and W. B. Schwartz (1976). "Toward the Simulation of Clinical Cognition: Taking the Present Illness." American Journal of Medicine 60:1-18.

P. Szolovits and S. G. Pauker. Categorical and probabilistic reasoning in medical diagnosis. Artificial Intelligence, 11:115-144, 1978. (Also in PDF.)

Oct. 8 Causal Reasoning

W. B. Schwartz, R. S. Patil, and P. Szolovits. Artificial intelligence in medicine: where do we stand. New England Journal of Medicine, 316:685-688, 1987.

R. S. Patil, P. Szolovits, and W. B. Schwartz. Causal understanding of patient illness in medical diagnosis. In Proceedings of the Seventh International Joint Conference on Artificial Intelligence, pages 893-899, 1981.

R. S. Patil, P. Szolovits, and W. B. Schwartz. Information acquisition in diagnosis. In Proceedings of the National Conference on Artificial Intelligence, pages 345-348, American Association for Artificial Intelligence, 1982.

A copy of Patil's thesis, which forms the basis of the last two papers, is also available on-line for anyone interested in the full details.

Oct. 15 Causal Reasoning II

Bill Long will join us to lead a discussion of how to model physiologic knowledge and reasoning, using examples from his Heart Disease Program.

W. J. Long. Temporal Reasoning for Diagnosis in a Causal Probabilistic Knowledge Base. Artif. Intell. in Medicine 8:193-215, 1996.

Oct. 22 Evaluation of (Diagnostic) Systems

These papers each report an attempt at a careful evaluation of one or more diagnostic expert systems. The first is the first published attempt to do "head-to-head" comparisons among systems. Please read these papers for methodology more than for particulars of the systems being evaluated.

Berner, E. S., G. D. Webster, A. A. Shugerman, et al. (1994) "Performance of Four Computer-Based Diagnostic Systems." New England Journal of Medicine 330.

Yu, V., Buchanan, B., Shortliffe, E., et al. (1979) Evaluating the performance of a computer-based consultant. Computers and Biomedicine 9:95-102.

Yu, V., Fagan, L., Wraith, S., et al. (1979) Antimicrobial selection by computer: A blinded evaluation by infectious disease experts. JAMA 242:1279-1282.

Long, W. (1980) "Criteria for Computer Generated Therapy Advice in a Clinical Domain." Computers in Cardiology 1980 285-288.

Long, W., Naimi, S., Criscitiello, M. (1994) Evaluation of a new method for cardiovascular reasoning. J. Amer. Med. Informatics Assn. 1:127-141.

Fraser, H., Long, W., Naimi, S. (2003)"Evaluation of a Cardiac Diagnostic Program in a Typical Clinical Setting." J. Amer. Med. Informatics Assn. 10:373-381.

Oct. 28 Machine Learning I

Baolin Wu, Tom Abbott, David Fishman, Walter McMurray, Gil Mor, Kathryn Stone, David Ward, Kenneth Williams and Hongyu ZhaoVol. "Comparison of statistical methods for classification of ovarian cancer using mass spectrometry data."  Bioinformatics 19 (13) 2003, pages 1636–1643

Soumya Raychaudhuri, Jeffrey T. Chang, Patrick D. Sutphin, and Russ B. Altman.  "Associating Genes with Gene Ontology Codes Using a Maximum Entropy Analysis of Biomedical Literature.Genome Research 12:203-214, 2002.

Late addition: Terrin, N., et al. External validity of predictive models: A comparison of logistic regression, classification trees, and neural networks. Journal of Clinical Epidemiology 56:721 (2003).

Nov. 5 Neural Nets

Remzi, M., et al. An Artificial Neural Network to predict the outcome of repeated prostate biopsies. Urology 62:3:456-460, 2003.

A Computer-Aided Diagnostic System to Characterize CT Focal Liver Lesions: Design and Optimization of a Neural Network Classifier. IEEE Transactions on Information Technology in Biomedicine 7:3:153-162, 2003.

Nov. 12 No class; AMIA Fall Symposium, Washington DC  
Nov. 19 Support Vector Machine Iizuka, Norio, et al. Oligonucleotide microarray for prediction of early intrahepatic recurrence of hepatocellular carcinoma after curative resection. The Lancet 361: 923 (2003).

Dreiseitl, S., et al. A Comparison of Machine Learning Methods for the Diagnosis of Pigmented Skin Lesions. J. Biomed. Informatics 1995:34:28-36.

Nov. 26    
Dec. 3 Presentations David Hanauer
Sameer Bade
Dec. 10 Presentations Alex Turchin
Reejis Stephen
Jennifer Sun
John Chuo
... and no more good stuff this term. Happy Holidays!
Not on Schedule Temporal Reasoning

A. Das and M. Musen. A Comparison of the Temporal Expressiveness of Three Database Query Methods. Proc Annu Symp Comput Appl Med Care 1995:331-7.

Y. Shahar and M.A. Musen, Knowledge-Based Temporal Abstraction in Clinical Domains. Artificial Intelligence in Medicine 1996 8(3):267-298.

I. S. Kohane and I. J. Haimowitz. Hypothesis-Driven Data Abstraction with Trend Templates. Proceedings, Annual Fall Symposium of the American Medical Informatics Association; 1993; Washington, DC. p. 444-448. The slides from Zak's presentation of this material are available as Temporal_slides.PDF.

  Past machine learning papers

Dreiseitl, S., et al. A Comparison of Machine Learning Methods for the Diagnosis of Pigmented Skin Lesions. J. Biomed. Informatics 1995:34:28-36.

Khan, J., et al., Classification and diagnostic prediction of cancers using gene expression profiling and artificial neural networks. Nature Medicine 2001 7(6):673-9.

Golub, T., et al.. Molecular Classification of Cancer: Class Discovery and Class Prediction by Gene Expression Monitoring. Science 286: 531-7, 1999.

Weber, G., et al. Building an Asynchronous Web-Based Tool for Machine Learning Classification, 2002.

Ramaswamy, S., et al. Multiclass cancer diagnosis using tumor gene expression signatures. PNAS 2001:98:26:15149-54.

Chan, K., et al., Comparison of machine learning and traditional classifiers in glaucoma diagnosis. IEEE Trans. Biomed. Engr. 2002:49:9:963-74.

Ong, I.M., et al. Modelling regulatory pathways in E. coli from time series expression profiles. Bioinformatics 2002:18 Suppl. 1:S241-S248.

Stephanopoulos, G., et al., Mapping physiological states from microarray expression measurements. Bioinformatics 2002:18:8:1054-1063.


psz@mit.edu; 12/10/2002