Physicians and possibly medical students and paramedics.

Similar documents
Overview. cis32-spring2003-parsons-lect15 2

Overview EXPERT SYSTEMS. What is an expert system?

Sample Interaction: Explanation. Sample Interaction (cont d) Sample Interaction: Therapy. Sample Interaction: Diagnosis. How It Worked: Representation

Knowledge is rarely absolutely certain. In expert systems we need a way to say that something is probably but not necessarily true.

Expert Systems: Simple Examples and case study

Knowledge Based Systems

INFERENCING STRATEGIES

Computer-Based Medical Decision Making: From MYCIN to VM

Expert Systems. Artificial Intelligence. Lecture 4 Karim Bouzoubaa

Chapter 2. Knowledge Representation: Reasoning, Issues, and Acquisition. Teaching Notes

Work up of Respiratory & Wound Cultures:

Neurology Diagnosis System

Clinical Decision Support Systems. 朱爱玲 Medical Informatics Group 24 Jan,2003

Bacteriemia and sepsis

Mixed gram positive organisms uti

Survival Skills for Researchers. Study Design

Details of the Revised Therapy Algorithm

Expert Systems Research: Modeling the Medical Decision Making Process

56 The Evolution of MYCIN s Rule Form

Unit II Problem 2 Microbiology Lab: Pneumonia

URINARY TRACT INFECTIONS 3 rd Y Med Students. Prof. Dr. Asem Shehabi Faculty of Medicine, University of Jordan

-Almost one third of cases admitted to medical centers are related to urinary tract infection

CPS331 Lecture: Coping with Uncertainty; Discussion of Dreyfus Reading

Microbial Flora of Normal Human Body Dr. Kaya Süer. Near East University Medical Faculty Infectious Diseases and Clinical Microbiology

MICROBIOLOGY SPECIMEN COLLECTION MANUAL

31.1 Materials and Methods

URINARY TRACT INFECTIONS 3 rd Y Med Students. Prof. Dr. Asem Shehabi Faculty of Medicine, University of Jordan

STIN2103. Knowledge. engineering expert systems. Wan Hussain Wan Ishak. SOC 2079 Ext.: Url:

A FRAMEWORK FOR CLINICAL DECISION SUPPORT IN INTERNAL MEDICINE A PRELIMINARY VIEW Kopecky D 1, Adlassnig K-P 1

Normal Flora PROF. HANAN HABIB DEPARTMENT OF PATHOLOGY COLLEGE OF MEDICINE, KSU

Experimental Research in HCI. Alma Leora Culén University of Oslo, Department of Informatics, Design

Normal Flora. CLS 212: Medical Microbiology

2046: Fungal Infection Pre-Infusion Data

Bacteraemia in patients receiving human cadaveric

N ational STI M icrobiological Surveillance Programme David A. Lewis FRCP(UK) PhD

Inferencing in Artificial Intelligence and Computational Linguistics

How to interpret your urine sample results

UNIT. Experiments and the Common Cold. Biology. Unit Description. Unit Requirements

Harvard-MIT Division of Health Sciences and Technology HST.952: Computing for Biomedical Scientists. Data and Knowledge Representation Lecture 6

How Does Analysis of Competing Hypotheses (ACH) Improve Intelligence Analysis?

Christina Martin Kazi Russell MED INF 406 INFERENCING Session 8 Group Project November 15, 2014

320 MBIO Microbial Diagnosis. Aljawharah F. Alabbad Noorah A. Alkubaisi 2017

SAGE. Nick Beard Vice President, IDX Systems Corp.

Infection Prevention & Control Core Skills Level 2

DR. HUDA ABO- ALEES GRAM-NEGATIVE BACILLI THE ENTERICS:

Small living organism Not visible to the naked eye Must be viewed under a microscope Found everywhere in the environment, including on and in the

Guideline for the Management of Fever and Neutropenia in Children with Cancer and/or Undergoing Hematopoietic Stem-Cell Transplantation

Customary urine test is the dip stick and the mid-stream culture of voided urine. Up to 77% of cystitis cases are cultured

Laboratory diagnosis of spinal tuberculosis: Past and Present. SA Patwardhan, S Joshi Abstract : Spinal tuberculosis often has an indolent course and

Best Practice Guidance 12_Version 1.2. Title: Collection of Biological Samples

Implementation of Inference Engine in Adaptive Neuro Fuzzy Inference System to Predict and Control the Sugar Level in Diabetic Patient

BIOSTATISTICAL METHODS

Anamnesis via the Internet - Prospects and Pilot Results

Analysis of complex patterns of evidence in legal cases: Wigmore charts vs. Bayesian networks

Latent Tuberculosis Infection Reporting Instructions for Civil Surgeons Using CalREDIE Provider Portal

Bacterial Infections of the Urinary System *

CERTIFICATE OF ACCREDITATION

Statistical inference provides methods for drawing conclusions about a population from sample data.

Representation and Analysis of Medical Decision Problems with Influence. Diagrams

Rule Based Expert System for Diagnosis of Neuromuscular Disorders

AIM #4: DEFINE SCIENCE AND EXPLAIN THE PROCESS OF THE SCIENTIFIC METHOD

HOSPITAL INFECTION CONTROL

What Is Science? Lesson Overview. Lesson Overview. 1.1 What Is Science?

Work-up of Respiratory Specimens Now you can breathe easier

Antibiotic Protocols for Paediatrics Steve Biko Academic Hospital

On the Representation of Nonmonotonic Relations in the Theory of Evidence

Communication Research Practice Questions

Principles of Infectious Disease Lecture #13 Dr. Gary Mumaugh

How to interpret your urine sample results

An expert system (Autodoc) for diagnosing diseases and prescribing medication using visual basic.net

EXERCISE. Proteins,Amino Acids, and Enzymes VII: Oxidase Test. Suggested Reading in Textbook. Pronunciation Guide. Materials per Student

Chronic urinary tract infections icd 10

IP Lab Webinar 8/23/2012

Lecture 1: Genito-urinary system. ISK

At the end of this presentation, you will be able to:

GRAM-NEGATIVE BACILLI THE ENTERICS: Family Enterobacteriaceae: Genus Escherichia & Genus Klebsiella

Chapter 9. Tests, Procedures, and Diagnosis Codes The McGraw-Hill Companies, Inc. All rights reserved.

Does amoxicillin help urinary tract infections

Automatic Extraction of Synoptic Data. George Cernile Artificial Intelligence in Medicine AIM

EPG Clinical Guidelines

Sexually Transmitted Diseases. Chlamydial. infection. Questions and Answers

A Computational Theory of Belief Introspection

UNIT 5 MAINTENANCE SYSTEMS Urinary System Test Bank

The Clinical Significance of Blood Cultures. Presented BY; Cindy Winfrey, MSN, RN, CIC, DON- LTC TM, VA- BC TM

4. RED AND BLUE TOP COLLECTION SWABS WILL NO LONGER BE SUPPLIED FOR BACTERIAL CULTURE

The Vine Assessment System by LifeCubby

Staining Technology and Bright- Field Microscope Use

3. Fixed-sample Clinical Trial Design

Principles of Disease and Epidemiology Chapter 14. Biology 1009 Microbiology Johnson-Summer 2003

Artificial Intelligence Programming Probability

Management of Needlestick and Mucous Membrane Exposures To Blood/Body Fluids

NEOSPORIN G.U. Irrigant Sterile (neomycin sulfate polymyxin B sulfate solution for irrigation)

TRICHOMONAS VAGINALIS

PRESENTER: DENNIS NYACHAE MOSE KENYATTA UNIVERSITY

MICROBIOLOGICAL TESTING IN PICU

E-BOOK # E COLI IN WELL WATER EBOOK

Corporate Online. Using Term Deposits

Risk Mediation in Association Rules:

ISSN (PRINT): , (ONLINE): , VOLUME-5, ISSUE-5,

Kyabaggu D, 2 Ejobi F 3* Olila D

Transcription:

MYCIN Introduction Problem domain: Selection of antibiotics for patients with serious infections. Medical decision making, particularly in clinical medicine is regarded as an "art form" rather than a "scientific discipline": this knowledge must be systemised for practical day-to-day use and for teaching and learning clinical medicine. Target Users: Physicians and possibly medical students and paramedics. Originators: Edward Shortliffe, (1972) Department of Medicine and Computer Science, Heuristic Programming Project, Stanford University School of Medicine, Stanford, CALIFORNIA References: Shortliffe, Edward H.(1981), "CONSULTATION SYSTEMS FOR PHYSICIANS: The Role of Artificial Intelligence Techniques". In Webber, Bonnie L.and Nilsson, Nils J. (Eds.). Readings in Artificial Intelligence. pp323-333. Tioga Publishing Company. Palo Alto, California.

Why choose MYCIN? "MYCIN developed at Stanford in the mid-1970s, was one of the first programs to address the problems of reasoning with uncertain or incomp lete information. MYCIN provided clear and logical explanation of the reasoning, used a control structure appropriate to the specific problem domain, and identified criteria to reliably evaluate its performance. Many of the expert system development techniques currently in use were first developed in the MYCIN project."

How does it work? DETAILS OF NEW CASE user ADVICE & EXPLANATIONS user interface knowledge base inference strategy expert system building tools knowledge engineer domain expert

What does it do? Diagnose inf ect ious disease and recommend appropriat e t herapy. Is t here a significant infect ion which requires t reat ment? What is t he ident it y of t he organism? What are t he pot ent ially useful drugs? Which drug or drugs will be best for t he pat ient?

MYCIN -- A typical consultation 1. Patient s name: (first - last) ** FRED BRAUN (Questions 1-3 requestbackground patient data). 2. Sex: ** M 3. Age: ** 55 4. Are there any cultures for Fred Braun which may be related to the present illness, and from which organisms have been grown successfully in the microbiology laboratory? ** Y - - - CULTURE - 1 - - - 5. From what site was the specimen for CULTURE - 1 taken? ** BLOD = BLOOD (System corrects typing error). 6. Please give the date and time when CULTURE-1 was obtained. (mo/da/yr time) ** JUN 20, 1977 The first organism isolated from the blood culture of 20-JUN-77 (CULTURE-1) will be referred to as: - - - ORGANISM - 1 - - - 7. Enter the laboratory - reported identity of ORGANISM-1: ** UNKNOWN (Typically identity is not known yet). 8. The strain (gram or Ziehl-Neelsen acid-fast) of ORGANISM-1: ** NEG (Preliminary lab results give some clues). 9. Is ORGANISM - 1 rod or coccus (etc.): ** ROD 10. What is the form of the individual organisms (e.g. lancetshaped for cocci, fusiform for rods, etc)? ** FUSIFORM (... more questions follow in order to gather sufficient information to infer the identity and significance of the infecting organisms...) 23. What is the suspected portal of entry of the gramneg rod from the blood culture of 20-JUN-77 (ORGANISM-1) into this sterile site? ** GI (6) (User indicates the uncertainty of his response). 24. Is Fred Braun a burn patient? ** N 25. Is Fred Braun a compromised host (e.g. alcoholic, sickle-cell-disease, immunosuppressed)? ** N 26. Is Fred Braun febrile? ** Y (... after about 50-60 questions, MYCIN prints the diagnostic hypotheses on which therapy will be based...)

My therapy recommendations will be designed to treat for organisms that are either very likely, or, although less likely, would have a significant effect on therapy selection if they were present. It is important to cover for the following probable infection(s) and associated organism(s): INFECTION-1 is BACTEREMIA <ITEM-1> E.COLI [ORGANISM-1] <ITEM-2> KLEBSIELLA [ORGANISM-1] <ITEM-3> ENTEROBACTER [ORGANISM-1] <ITEM-4> KLEBSIELLA-PNEUMONIAE [ORGANISM-1] ** WHAT DOES GU MANIPULATION TELL YOU ABOUT THE IDENTITY OF AN ORGANISM? The rules listed below use: whether the patient has had a genito-urinary manipulative procedure to conclude about: the identity of the organism 156, 163, 190 Which do you wish to see? ** 156 RULE 156 IF: 1. The site of the culture blood, and 2. The gram stain of the organism is gramneg, and 3. The morphology of the organism is rod, and 4. The portal of entry of the organism is urine, and 5. The patient has not had a genito-urinary manipulative procedure, and 6. Cystitis is not a problem for which the patient has been treated THEN: There is suggestive evidence (.6) that the identity of the organism is e.coli ** IS BLOOD A STERILE SITE? Yes. Blood is one of those sites that is normally sterile. ** WHAT ARE THE NONSTERILE SITES? Those sites that are normally nonsterile are: CERVIX, CUTANEOUS-ULCER, LOCHIA, NOSE, SKIN, STOOL, THROAT, URETHRA, VAGINA. ** IS BACTEROIDES AEROBIC? No, the aerobicity of bacteroides is anaerobic.

Dynamic knowledge - st ruct ure pat ient cult ure1 cult ure2 operat ion organism1 organism2 organism3 drug5 drug1 drug2 drug3 drug4 Context tree

Implementation and Algorithms How MYCIN works 1. Create patient 'context' tree 2. Is there an organism that requires therapy? 3. Decide which drugs are potentially useful and select the best drug The above is a goal-oriented backward chaining approach to rule invocation & question selection. MYCIN accomplishes the invocation and the selection through two procedures: MONITOR and FINDOUT -- procedures developed by the MYCIN development team:

How MYCIN Works MONITOR (for MYCIN rules) attempts to evaluate the premise of the current rule, condition by condition. If any of the conditions is false, or indeterminate due to lack of information, the rule is rejected, and the next rule on the list of applicable rules pending in the current context is tried. The rule application succeeds when all of the conditions in the premise are deemed to be true, and the conclusion of the rule is added to the record of the current consultation:

Start Consider the 1st condition in the premise of the rule Gather the necessary information using the FINDOUT mechanism no Has all necessary information been gathered to decide if the condition is true yes Consider the next condition in the premise of the rule yes Is the condition TRUE? yes Are there more conditions to check? REJECT THE RULE no (or unknown) no Add the conclusion of the rule to the ongoing record of the current consultation EXIT EXIT

How MYCIN Works FINDOUT (Mechanism) searches for data needed by the MONITOR procedure, particularly the MYCIN 'clinical' parameters referenced in the conditions which are not known. Essentially FINDOUT gathers the information that will count for or against a particular condition in the premise of the rule under consideration. If the information required is laboratory data which the user can supply, then control returns to MONITOR and the next condition is tried. Otherwise, if there are rules which can be used to evaluate the condition, by virtue of the fact that their actions reference the relevant clinical parameter, they are listed and applied in turn using MONITOR:

Start no Retrieve Y = List of Rules which may aid in deducing the value of the PARAMETER Is the PARAMETER a piece of LABORATORY data? yes Ask USER for the Value of the PARAMETER Apply MONITOR to each rule in the List Y RETURN yes Is Value of the PARAMETER known Is Value of the PARAMETER known yes RETURN no Retrieve Y = List of Rules which may aid in deducing the value of the PARAMETER no Ask USER for the Value of the PARAMETER Apply MONITOR to each rule in the List Y RETURN RETURN An example of the kind of reasoning network generated by the MONITOR and FINDOUT mechanisms. Names of the clinical parameters are underlined. When a rule has multiple conditions in the premise, numbers have been included to specify the positions of the associated clinical parameters within the premise condition.

GOAL REGIMEN RULE 092 TREATFOR COVERFOR RULE 090 RULE 149 1 2 3 IDENT INFECTLOC FEBRILE SIGNIFICANCE RULE 038 RULE 042 RULE 044 RULE108 SITE 1 2 3 SITE COLLECT SIGNUM 1 2 3 SITE NUMCLUS NUMPOS RULE 001 SITE RULE 041 RULE 171 1 2 3 1 2 3 SITE SITE ABNORMAL SITE SITE ABNORMAL ASK2 ASK2