This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
MYCIN is used for detection of bacteria causing infection and recommend antibiotics for that very particular infection. It also tells the dose according to patient's body weight. It provides users with explanations of its reasoning. This is also used for detection of blood clotting diseases.
MYCIN was an expert system developed in 1970's at Stanford University. It is written in Lisp for doctoral dissertation of Edward Shorliffe under Bruce Buchanan, Stanley N. Cohen and others. It was developed in laboratory which created Dendral expert system and it uses judgmental rules that had elements of uncertainty associated with them. MYCIN is used for detection of bacteria causing infection such as bacteremia and meningitis, and recommend antibiotics for that very particular infection. It also tells the dose according to patient's body weight. Even some antibiotics has suffix-"MYCIN". In short, MYCIN is an expert system to perform at the levels of a human expert and to provide users with explanations of its reasoning. This is also for diagnosis of blood clotting diseases.
METHOD USED BY MYCIN:
MYCIN uses simple engine for its working and knowledge based of approx. 600 rules. It shots the physician with so many questions for which he has to answer in yes and no and MYCIN expert system helps him to identify the bacteria that is responsible for that specific problem and also recommend the drug course of the treatment for that very specific problem.
E.g.: Rules used for specification of problem according to evidence given:
- the gram stain of organism is grampositive, and
- the morphology of organism is coccus and
- the aerobicity of organism is aerobic.
- the growth conformation of the organism is clumps.
THEN: there is suggestive evidence (0.7) that identifies that organism is Staphylococcus.
- the gram stain of the organism is gramnegative and
- the morphology of the organism is rod and
- the organism breath aerobically.
THEN: there is strongly suggestive evidence (0.8) that the class of the organism is enterobacteriaceae
There are two data structures used by MYCIN expert system
- Static data structure
- Dynamic data structures
1) Static data structures:
These store medical knowledge not suitable for storage as inferential rules: includes lists of organisms, knowledge tables having name of bacteria types and parameters.
Parameters = features of patients, bacterial cultures, drugs, etc.
EXPECT- includes range of possible values to bacteria
PROMPT - English.
LOOKAHEAD- lists of rules mentioning the parameter in their premise (e.g. a rule might need to know whether or not a patient is febrile)
UPDATED-BY lists rules mentioning the parameter in their action (i.e. they may draw a conclusion about the value of the parameter, such as the IDENTITY parameter)
2) Dynamic data structures:
It store information about the evolving case - the patient details, possible diagnoses, rules consulted:
Example of dynamic data:
Toevaluate the premise of the rule:
($AND (SAME ORGANISM-1 GRAM GRAMNEG)
(SAME ORGANISM-1 MORPH ROD)
(SAME ORGANISM-1 AIR AEROBIC))
From following data
GRAM = (GRAMNEG 1.0)
MORPH = (ROD 0.8) (COCCUS 0.2)
AIR = (AEROBIC 0.6) (FACUL 0.4)
Total value of whole expression is lowest certainty value = 0.6
(CONCLUDE ORGANISM-1 is enterobacteriaceae 0.48 (= 0.6 (CF of premise) x 0.8 (CF of rule))
RESULT CONCLDED AT STANFORD USING MYCIN:
The research at Stanford Medical School concluded that therapy recommended by MYCIN is accepted in about 69% of cases which is very much exact as compare to recommend by physician who was recommended by taking some criteria but accuracy is low.
HOW MYCIN WORK:
1) Diagnose for infectious disease.
2) Identify infection that requires therapy.
3) What is the rank of organism creating infection
- Aerobic, etc.
4) Which drug is recommended for that particular infection?
- Cloramphenical (0.99)
- Clendamycin (0.99)
- Erthromycin (0.57)
- Tetramyclin (0.28)
5) Which is best out of above?
The system has to interact with the user for giving its consultant. If user want to interact with the system, it is done as user enter the data in English and the system response to the user by using its inference engine which is used for carrying out the reasons while answering to the questions of the users. The role of the inference engine is searching of answers to the questions of the users. It is as, MYCIN match the condition given by the users with the conditions placed in the system and answer the questions by the user. MYCIN also us heuristics search for answering the questions asked by the user to the system.
The interaction is done between two a SYSTEM and a USER and user may be Physicians, medical students or paramedics.
- SYSTEM: What is the name of the patient (first name last name)?
- USER: Jannat Khan.
- SYSTEM: What is the age of the patient?
- USER: 35.
- SYSTEM: What is the gender of the patient(Male(M),Female(F))?
- USER: F.
- SYSTEM: What are symptoms observed till now? USER:
- Upset stomach.
- High Fever.
- Body ache.
- Neck Stiffness.
- SYSTEM: When did these symptoms was observed?
- USER: October 26, 2009.
- SYSTEM: Do you have CSF culture report of the patient?
- USER: Yes.
- SYSTEM: Do think the patient has infection at the site through which specimen is not fond by you?
- USER: No.
- SYSTEM: Has the patient had objective evidence of abnormal neural activities given by the physician?
- USER: Yes.
- SYSTEM: Do the patient have the x-ray of its chest?
- USER: No.
- SYSTEM: Has the patient have any contagious diseases in recent time?
- USER: No.
- SYSTEM: Is a patient has burns on its body?
- USER: No.
- SYSTEM: Do the patient has addicted to drugs?
- USER: Yes.
- SYSTEM: Meningitis is hospital acquired system or not?
- USER: Yes.
The CSF culture will be considered to be associated with meningitis.
Result given by MYCIN:
- I have found evidence of (Rule45 Rule57) which cover the treatment for e-coli causing meningitis.
- I have found evidence of (Rule45 Rule57 Rule78) which cover the treatment for pseudomaonas aeruginosa causing meningitis.
- I have found the evidence (Rule45 Rule57 Rule78) which cover the treatment for Klebsiella Pneumoniae causing meningitis.
- I have found the evidence (Rule 07) which cover the treatment for Diplococcus-pneumoniae causing meningitis.
- pseudomaonas aeruginosa
- klebsiella pneumonia
- SYSTEM: Dose the patient has any type of allergy from any antimicrobial agent?
- USER: No.
- SYSTEM: What is the weight of patient in kilograms?
- USER: 67.
Recommendations by MYCIN:
My preferred therapy is given below:
In order to treat for items<1 2 3 4> make the combinations of drugs given below:
Dose 3.5g q4h IV
Dose 125 mg q8h IV
As penicillin is in high concentration so it can inactive aminoglycosides. So, do not mix the two antibiotics mentioned above.
Meningitis: Meningitis is the disease; in which inflammation occur over the protective members who cover the brain and spinal cord. This is caused by viruses, bacteria and other micro organism. It is also caused to the person who takes drugs. The symptoms observed during this are headache, fever, upset stomach, body ache and neck stiffness. For verify the disease CSF(cerebrospinal fluid) culture is done, which involves inserting of a needle into spinal canal to extract cerebrospinal fluid. This process is known as lumber puncture.
- It always ask question which human forget to ask: There are many questions which human forget to ask but they are necessary to be answer. So, it is very useful.
- It can work continuously: MYCIN woks continuously, no need of human efforts.
- User can use it more frequently: It is easy to use and give accurate answer which is necessary to detect the disease.
- It is a multi-user expert system: It can diagnose for more than one disease at a time. It saves our time.
- Give accurate answers: It ask on questions repetitively to give proper and accurate answer.
- Lack of common sense: No common sense is there in the expert system. They have no understanding it only works on the data added by the developer. If a patient has bleeding over gunshot wound than MYCIN would attempt to diagnose that it is due to some bacteria.
- Domain experts not always able to explain their logic: The knowledge base expert system is small but sometimes it has to work over other but that information is not feed in so, sometimes it does not give proper and accurate answer.
- Sometimes wrong decisions: Sometimes there are errors in the knowledge base so, MYCIN does not able to give accurate report.
- It cannot be adapted to changing environment: It cannot be adapted to the changing environment until the knowledge base is changed.