Medical Informatics and Medical DecisionSupport Systems: An Introduction

1
Medical Informatics and Medical Decision-Support
Systems An Introduction

• Yuval Shahar, M.D., Ph.D.

2
Medical Informatics (A)

• Medical information science is the science of
  using system-analytic tools . . . to develop
  procedures (algorithms) for management, process
  control, decision making and scientific analysis
  of medical knowledge.
• (E.H. Shortliffe, The science of biomedical
  computing. Medical Informatics 19849185-93.)

3
Medical Informatics (B)

• Medical Informatics comprises the theoretical
  and practical aspects of information processing
  and communication, based on knowledge and
  experience derived from processes in medicine and
  health care.
• (J.H. van Bemmel, The structure of medical
  informatics. Medical Informatics 19849175-80.)

4
Medical Informatics (C)

• In medical informatics we develop and assess
  methods and systems for the acquisition,
  processing, and interpretation of patient data
  with the help of knowledge that is obtained in
  scientific research.
• (J.H. van Bemmel and M.A. Musen, Handbook of
  Medical Informatics, Springer Verlag, 1997.)

5
Medical Informatics is Multidisciplinary

• It is applies methodologies developed in multiple
  areas of scientific endeavor to many different
  tasks
• In turn, it often gives rise to new, more general
  methodologies that enrich these scientific
  disciplines

6
Example of Scientific Areas Relevant to Medical
Informatics

• Medicine/ Biology
• Mathematics
• Information Systems
• Computer Science
• Statistics
• Decision Analysis
• Economics/Health Care Policy
• Psychology

7
Examples of Medical Informatics Areas

• Hospital information systems
• Electronic medical records medical vocabularies
• laboratory information systems
• pharmaceutical information systems
• radiological (imaging) information systems
• Patient monitoring systems
• Clinical decision-support systems
• Diagnosis/interpretation
• Therapy/management
• Bioinformatics Closely related tasks/methods

8
The Diagnostic-Therapeutic Cycle,A Simplified
View
Data collection -History -Physical
examinations -Laboratory and other tests
Information
Data
Decision making
Patient
Therapyplan
Planning
Diagnosis/assessment
9
Levels of Automated Support(Van Bemmel and
Musen, 1997)
10
Medical Decision-Support Systems

• Task
• Diagnosis/interpretation
• Therapy/management
• Scope
• Broad (e.g., Internist-I/QMR internal medicine
  Dx DxPlain Iliad EON for guideline-based
  therapy)
• Narrow (e.g., de Dombals system for diagnosis of
  acute abdominal pain MYCIN infectious diseases
  Dx ECG interpretation systems ONCOCIN support
  of application of oncology protocols)

11
Types of Clinical Decision-Support Systems

• Control level
• Human-initiated consultation (e.g., MYCIN, QMR)
• Data-driven reminder (e.g., MLMs)
• Closed loop systems (e.g., ICU ventilator
  control)
• Interaction style
• Prescriptive (e.g., ONCOCIN)
• Critiquing (e.g., VT Attending)

12
Diagnostic/Prognostic Methods

• Flow charts/clinical algorithms
• Statistical and other supervised and
  nonsupervised classification methods
• Neural networks, ID3, C4.5, CART, clustering
• Bayesian/probabilistic classification
• Naïve Bayes, belief networks, influence diagrams
• Rule-based systems (MYCIN)
• Ad hoc heuristic systems (DxPlain)
• Cognitive-studies inspired systems (Internist I)

13
Example de Dombals System (1972)

• Domain Acute abdominal pain (7 possible
  diagnoses)
• Input Signs and symptoms of patient
• Output Probability distribution of diagnoses
• Method Naïve Bayesian classification
• Evaluation an eight-center study involving 250
  physicians and 16,737 patients
• Results
• Diagnostic accuracy rose from 46 to 65
• The negative laparotomy rate fell by almost half
• Perforation rate among patients with appendicitis
  fell by half
• Mortality rate fell by 22
• Results using survey data consistently better
  than the clinicians opinions and even the
  results using human probability estimates!

14
Temporal Reasoning and Planning in Medicine

• Almost all medical data are time stamped or time
  oriented (e.g., patient measurements, therapy
  interventions)
• It is virtually impossible to plan therapy, apply
  the therapy plan, monitor its execution, and
  assess the quality of the application or its
  results without the concept of time

15
Time in Natural Language
From Mr. Jones was alive after Dr. Smith
operated on him Does it follow that Dr.
Smith operated on Mr. Jones before Mr. Jones was
alive? Is Before the inverse of After?
16
Understanding a Narrative

• List all, find at least one, or prove the
  impossibility of a legal scenario for the
  following statements
• John had a headache after the treatment
• While receiving treatment, John read a paper
• before the headache, John experienced a visual
  aura
• One legitimate scenario (among many) is
• John read the paper from the very beginning of
  the treatment until some point before its end
  after reading the paper, he experienced a visual
  aura that started during treatment and ended
  after it then he had a headache.

Paper
Aura
Treatment
Headache
17
Monitoring

• Determine if an oncology patients record
  indicates a second episode that has been lasting
  for more than 3 weeks, of Grade II bone-marrow
  toxicity (as derived from the results of several
  different types of blood tests), due to a
  specific chemotherapy drug.

18
Planning and Execution

• If the patient develops sever anemia for more
  than 2 weeks, reduce the chemotherapy dose by 25
  for the next 3 weeks and in parallel monitor the
  hemoglobin level every day.

19
Display and Exploration of Time-Oriented Data
20
A Temporal-Reasoning TaskTemporal Abstraction
Input time-stamped clinical data and relevant
events Output interval-based abstractions
Identifies past and present trends and states
Supports decisions based on temporal patterns
modify therapy if the patient has a second
episode of Grade II bone-marrow toxicity
lasting more than 3 weeks Focuses on
interpretation, rather than on forecasting
21
Temporal AbstractionA Bone-Marrow
Transplantation Example
PAZ protocol
BMT
Expected CGVHD
.
M0
M1
M2
M3
M1
M0
Granu-
Platelet
locyte
D
D
counts
D
D
D
counts
D
D
D
(
)
D
D
D
D
D
D
D

D



D

(
)








D
D
150K

2000



100K


1000
400
0
200
100
50
Time (days)
22
Uses of Temporal AbstractionsIn Medical Domains
Planning therapy and monitoring patients over
time Creating high-level summaries of
time-oriented patient records Supporting
explanation in medical decision-support
systems Representing the intentions of
therapy guidelines Visualization and
exploration of time-oriented medical data
23
Temporal Reasoning Versus Temporal Maintenance

• Temporal reasoning supports inference tasks
  involving time-oriented data often connected
  with artificial-intelligence methods
• Temporal data maintenance deals with storage and
  retrieval of data that has multiple temporal
  dimensions often connected with database systems
• Both require temporal data modeling

24
Medical Image Processing

• Input X-Ray, CT-scan, MRI, PET, etc.
• Tasks
• Correction of multiple artifacts
• RegistrationSuperimposition to enhance
  visualization
• Segmentation Decomposition into semantically
  meaningful regions

25
Summary Medical Informatics

• Multidisciplinary research, development, and
  application
• inspired by and benefits underlying core
  scientific/engineering areas
• Medical Decision support systems
• Tasks Diagnosis, therapy
• Mode Human initiated, data driven, closed loop
• Interaction style Prescriptive, critiquing
• Multiple diagnostic/therapeutic methodologies