Download group13(AI in medicine)

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Personal knowledge base wikipedia , lookup

Video game rehabilitation wikipedia , lookup

History of artificial intelligence wikipedia , lookup

Clinical decision support system wikipedia , lookup

Knowledge representation and reasoning wikipedia , lookup

Transcript
AI IN MEDICINE
Pic: Google
Tejashree Aher (06011011)
Akhil Deshmukh (06D05007)
Anshul Maheshwari (06D05009)
Narendra Kumar (06D05008)
Overview


Motivation
AIM
◦ What is AIM?
◦ Goals of AIM
◦ Applications of AIM

Clinical expert system : MYCIN
◦
◦
◦
◦
◦
◦
Introduction
How it works
Specification of the therapy selection problem
Representation of Goals
Certainty factor
Partial derivation of the algorithm
Motivation

‘Doctors in a box’ to diagnose diseases.

Community of computer scientists and
healthcare professionals set a research
program - Artificial Intelligence in
Medicine (AIM) with the aim of
revolutionize medicine.
What is AIM?
Clancey and Shortliffe :’Medical artificial
intelligence is primarily concerned with the
construction of AI programs that perform
diagnosis and make therapy
recommendations.
Medical AI programs are based on symbolic
models of disease entities and their
relationship to patient factors and clinical
manifestations’
 AI specialized to medical applications
 Employ human- like reasoning methods in
the programs

AIM use ‘machine learning’ to use and
create knowledge.
 Machine learning - computers that can
learn from experience
 Use – stored data used in diagnosis
 Creation - analyse the relationship within
the data to come up with new results
 Used in Drug discovery

Goals of AIM?

Expert computer programs for clinical use

Dissemination of the best medical expertise
to geographical regions where that expertise
is lacking

Making consultation help available to nonspecialists not within easy reach of expert
human consultants.

To formalize medical expertise
Applications of AIM

Knowledge based systems

Diagnostic and educational systems
Knowledge based systems




Use the medical knowledge stored for reasoning
Store information about a specific task
Knowledge represented in the form of set of rules
Support healthcare workers in the normal course of
their duties -manipulation of data and knowledge
Examples: Generating alerts and reminders -warn
changes in a patient's condition (in less critical cases,
through a email)
Agents for information retrieval- software
agents are sent to search for and retrieve information
Diagnostic and educational systems
Most research systems were developed to
assist clinicians in the process of diagnosis.
 Expert System

◦ A program that contains a large amount of
knowledge in one specific area.
◦ Rules for organizing and expressing its
knowledge
◦ Approaches to integrate the recommendations
MYCIN
•Created in the mid-1970s,helps doctors choose the
correct antibiotics for patients with severe infections (and
the best ones !!!!)
•It is given large amounts of information on meningitis and
bacteremia
•This information represented as -“if A and B are
true(evidently), then there is evidence that C is true”.
•Dynamic computation
•Same recommendation with different certainty factors,
MYCIN integrates them by means of a numerical function.
How MYCIN Works ???
1. Diagnose for infectious diseases.
2. Identify infection that requires therapy,
3. What is the identity of the organism(s) by clinical
and laboratory evidence.

primary, secondary.
4. What are the potentially useful drugs





chloramphenicol (0.95)
clindamycin (0.95)
erthromycin (0.77)
tetracycline (0.41)
carbenicillin (0.25)
5. Which will be best ? (yes, it suggests the best one!)
Fig by: M. Chandra and V. K. Sonkar
Example
Joe shows the following disorders
•Headache
•Bodyache
•Nausea
What exactly is wrong with Joe??
MYCIN has the answer.
Pic: Google
Organization of MYCIN
MYCIN
PATIENT
Physician user
RULE BASE
Consultation program
Dynamic
patient
data
Explanation program
Static
knowledge
base
Knowledge acquisition
program
PATIENT
DB
Infectious disease expert
Fig by: M. Chandra and V. K. Sonkar
MEDICAL
EXPERT
The Knowledge Base
Inferential knowledge: stored in decision rules
•If Premise then Action (Certainty Factor [CF])
•If A&B then C (0.6)
•The CF represents the inferential certainty
Static knowledge:
•Natural language dictionary
•Lists (e.g., Sterile Sites)
•Tables (e.g., primary, gram stain, morphology, aerobicity)
Dynamic knowledge stored in the context tree:
•Patient specific
•Hierarchical structures: Patient, cultures, organisms
Fig by: Yuval Shahar
Specification of the therapy selection problem
Given a diagnosis (one or more organisms suspected of infecting
the patient), choose the therapy (set of drugs) that best satisfies
the following medical goals:
1. Maximize drug sensitivity.
2. Maximize drug efficacy.
3. Continue prior therapy.
4. Minimize number of drugs.
5. Give priority to covering likelier organisms.
6. Maximize number of suspected organisms covered.
7. Don’t give two drugs from the same general class.
8. Avoid contraindications for the patient.
How to choose the best therapy???
It subject all the therapies to the following three
tests 


Coverage test.
Classes of selected drugs in a therapy.
Contra-Indication.
A therapy is suggested or rejected , Explanation !!!
•Representation of goals:
•Set of axioms
•Partial ordering
•Preference order
•Linear ordering
•Metric representation
•Partition
•Yes/no predicate
•Certainty Factor.
Certainty Factor:
•What is Certainty factor?
•How does it combine?
A
•Proceeds as:
A
C
A

B
B
B
•Several rules
single hypothesis.
•Several propositions together.
•Following the chaining rule.
C
•Measure of belief: MB[h, e].
•Measure of disbelief: MD[h, e].
•Certainty factor: CF[h, e] = MB[h, e] – MD[h, e].
•Combination of evidences:
MB[h, s1 ^ s2] = 0
if MD[h, s1 ^ s2] = 1
MB[h, s1] + MB[h, s2]*(1- MB[h, s1]) else
MD[h, s1 ^ s2] = 0
if MB[h, s1 ^ s2] = 1
MD[h, s1] + MD[h, s2]*(1- MD[h, s1]) else
•Combination of hypothesis:
MB[h1  h2,e] = min(MB[h1,e] ,MB[h2,e] )
MB[h1  h2,e] = max(MB[h1,e] ,MB[h2,e] )
Certainty Factors
Certainty Factor (CF) with its conclude functions,
Conclude functionSay the current CF value is x, and a new evidence with CF y is
supporting the same hypotheses comes, then
F(x,y) =
x+y(1--x)
if x, y ≥0,
=
x+y(1 +x)
if x, y<0, |x|, |y|≤ 1.
=
(x + y)/(1 - min(|x|,|y|))
else.
Conclude derives a conclusion including the CF of the result
E.g., “There is suggestive evidence (0.7) that the identity of the organism is
streptococcus”.
•It is always true that -1 ≤ CF ≤ +1
•If CF = +1 then all other hypotheses are rejected
Joe has a disease A
• bodyache ^ headache->yes (0.7) ...e1
•headche^ weakness -> yes (0.8) ...e2
•no weakness -> no (0.6) ….e3
•weakness ^ nausea -> yes (0.6) ....e4
Joe comes to doctorheadache? yes
bodyache ? yes
weakness ? no
nausea ? yes
CF(headache (Joe, yes)) = 0.7
CF(weakness (Joe, yes)) = 0.65
CF(nausea (Joe, yes)) = 0.4
CF(bodyache (Joe, yes)) = 0.8
Pic: Google
MD(joe, e3)= CF(e3)* max(0, CF(weakness))
= 0.6 * (1-0.65) =0.210
get MB(joe, e1) = CF(e1)* max (0, min(CF(bodyache), CF(headache)))
= 0.7 * 0.7 =0.49
MB(joe, e2)
=CF(e2)* max (0, min(CF(weaknes8), CF(headache)))
= 0.8 * 0.65 =0.52
MB(joe, e4)
=CF(e3)* max (0, min(CF(weaknes8), CF(nausea)))
= 0.6 * 0.4 =0.24
=CF(e4)* max (0, min(CF(no weaknes)))
= 0.4 * 0.6 = 0.24
MB(joe, e3)
MB(joe, {e1,e2})= 0.49+ 0.52 *(1-0.49) = 0.7552
MB(joe,{e1,e2,e4})= 0.7552 + 0.24 *(1- 0.0.7552) = 0.813
MD(joe,e3)= 0.6 * 0.24 = 0.144
CF (joe, fever) = MB(joe, fever) - MD(fever)
= 0.813-0.144 = 0.669
………. Chances of Joe having fever !!
Partial derivation of the algorithm
•Representing Goals:
•Linear ordering: <fewer
•Matric scale: 100-1000
•Considering the above example:
•Drug (A) <fewer Drug (B)
Preference ordering and Partition
Preference ordering: CONDENSE, a many to one function F(x).
F(x)<F(y) => x<y
PARTITION: M(x) -> F(x)
F(x)= λ(x) { i | ti-1 <p M(x) <p ti}
t0 ≤p M(x) ≤p tn+1
‘->’: re-formulation of constraint.
Drawback of CONDENSE
F(x) < F(y): significant difference
EXTENDsion and CONJOINing
•
EXTEND: An ordering on individual items to an ordering on bags of items, follows
•{x} < {y} iff x < y.
•If X < Y and X’ < Y’, then X+X’ < Y+Y’, where + denotes bag union. For exampl
1<2 implies {1} < {2} and {l, l} < {1,2}.
• CONJOIN: We combine the preference <fewer for fewer drugs with the
preference <effective for more effective therapy by Conjoining them.
•x ≤ p&q Y iff x ≤P y and x ≤q y //x is atleast as good as y
•x <p&q iff (x ≤p y and x <q y) OR (x <p y and x ≤q y) //x is preferable
•(Note that A <effective B means therapy A is more is more effective than therapy B,
ie. More preferable with respect to the effectiveness.)
Combine coverage preferences
The therapy goals listed in above include maximizing the number of organisms
covered and giving priority to those the patient is likelier to have. Let’s see how
these two goals are integrated:
1. Classify organisms as “most likely” or “less likely.”
2. Relax the coverage goal by ignoring “less likely” organisms.
3. Reformulate the coverage goal as the constraint that all the “most likely”
organisms be covered.
Domination of Preferences
1. Letting one preference -- <primary , <secondary – using <secondary only to resolve ties .
X < primary Y Or (X =primary Y and X <secondary Y).
2. A preference can simply be IGNORED. For example, ignoring <secondary
<primary;secondary to < primary ' This particular case of IGNORE is appropriate if ties
with respect to <primary are too rare to worry about, or if violating <secondary in the
event of such a tie wouldn’t do much harm.
It is unlikely for two therapies to be equally effective on the likeliest organisms but
different on the less likely ones, so it is reasonable to ignore the less likely organisms
altogether.
3. The Condensed preference compares therapies based on the number of “most
likely” organisms covered. This preference is now reformulated into a constraint by
THRESHOLDING.
THRESHOLD (tmin): M(x)
µ(X) ,
(M(x) ≥ tmin),
Maximizing therapy effectiveness appears more important
than minimizing the number of drugs, in the sense that
increasing therapy effectiveness by 1 rank is considered
more desirable than reducing the number of drugs by 1.
-Why MYCINAddresses the problems of reasoning.
Provide clear and logical explanation of reasoning.
Explore how human experts make these rough (but
important) guesses.
Useful for junior or non-specialized doctors.
-MYCINDoes it always thinks like an Expert??
•But not always good to use drugs with high effectiveness .
•So it is always preferred by professional doctors to start with low
concentration ( low mg) drugs, than increase it step by step if effects are not
significant.
•At the time of the first study, MYCIN rules included only bacteremia
(meningitis and endocarditis were added later), thus never tested in a real
clinical environment with general infections
Summary






Reduction in Medication Errors and Adverse
Drug Events.
Computerassisted - fewer errors than
handwritten prescriptions and to be five times
less likely to require pharmacist clarification
Prompt to use a cheaper generic drug when a
more expensive drug was initially ordered;
Cannot model common-sense
Cannot be completely relied upon ( loss of
confidence !! )
The knowledge-acquisition bottleneck remained
significant (additional effort from already busy
individuals !!!)
Contd.
Rely on human knowledge
 The program acts as advisor to a person
 Medical practitioners serve as a critical
layer of interpretation between an actual
patient and the expert systems
 Limited ability of the program to make a
few common sense inferences is enough
to make them usable and valuable

References
Peter Szolovits , Artificial Intelligence and Medicine, Westview Press,1982.

Towards Explicit Integration of Knowledge in Expert Systems: An Analysis
of MYCIN’s Therapy Selection Algorithm, Bill Swartout, Jack Mostow,
AAAI-86 ,1986.

http://www.openclinical.org/gmm_ardensyntax.html.

Peter Szolovits, William J. Long, The Development of Clinical Expertise in the
Computer, Westview Press,1982.
Athanasios K. Tsadiras*, Konstantinos G. Margaritis, “The MYCIN certainty
factor handling function as uninorm operator and its use as a threshold
function in artificial neurons”, Fuzzy Sets and Systems 93,1998.
 Yuval Shahar, Diagnostic Systems (I),Medical Decision support systems,
Stanford Univarcity,2007.
