Professional Documents
Culture Documents
Gartner Inc.'s Hype Cycle showing the fluctuating support of a new technology
SOME AI ACHIEVEMENTS
machine translation
data mining
industrial robotics,
speech recognition
banking software
medical diagnosis,
Google's search engine
fuzzy logic controllers
heuristic search
data analytics
WHAT IS GOING ON NOW?
Some Important Current AI Projects
Personal assistants (e.g. Microsoft)
Speech recognition (e.g. Siri)
“Watson” for question-answering (IBM)
Robo-docs (e.g. X Prize $10 Million For AI Diagnostician)
NELL: Never-Ending Language Learning machine that
“reads” (Carnegie Mellon)
DARPA Programs e.g Grand Challenges, Semantic Web
EU-FP7 Support
HOW DID AI INTERACT WITH
MEDICINE?
Analogous “hype” cycle over the decades
Early emphasis on “doc in the box” e.g. diagnosis programs
Heavy use of rules-based expert systems
Some early focused programs
MYCIN: rules-based system, developed at Stanford originally
for the diagnosis and treatment of bacterial infections of the
blood and later extended to handle other infectious diseases
as well.
Present Illness Program (PIP), for taking the history of the
present illness of a patient with renal (kidney) disease
ABEL, a program for the diagnosis (and eventually
treatment) of acid/base and electrolyte disturbances.
Some General Diagnosis Programs
Iliad: Bayesian general diagnosis system that grew out of an
early electronic record, the HELP system
INTERNIST-I a computer-assisted diagnostic tool designed
to capture the expertise of just one man, Jack D. Myers, MD
Dxplain uses modified Bayes to produce a ranked list of
diagnoses, provide justification for each disease, suggest
further steps
Some Early Lab Programs
PUFF Expert System for Interpretation of Pulmonary Function
Data) was probably the first AI system to be used in clinical
practice.
APACHE was one of the first medical decision support systems to
be commercialized - Provides a numerical indicator of severity of
illness, risk of mortality, risk of active treatment (an indication of
appropriate resource utilization) to predict clinical results in the
ICU
GermWatcher checks for hospital-acquired (nosocomial)
infections
PEIRS (Pathology Expert Interpretative Reporting System) PEIRS
interpreted about 80-100 reports a day with a diagnostic accuracy
of about 95%
WERE AI PROGRAMS WELCOMED BY
HEALTHCARE?
Not by clinicians
Not by educators
Some for clinical surveillance
Commonly in laboratories
Significantly in signal analysis (ECG, EEG, EMG)
*WHY ARE AI PROGRAMS REJECTED?*
Product not wanted or puerile
Onerous, time-consuming data input
Clumsy interface
Program intrusive (esp for reminders)
Plethora of single-issue programs
Even small error rate not acceptable
Inventor makes no effort to disseminate
Not interfaced with EHR
Not invented here
Technophobia
WHAT ABOUT AI AND HEALTHCARE TODAY?
Dep’t of Extravagant Predictions
Detecting ‘wrong blood in tube’ errors: Evaluation of a Bayesian network approach
Decision support in heart failure through processing of electro- and echocardiograms
Missing data imputation using statistical and machine learning methods in a real breast cancer
problem
Evolving fuzzy medical diagnosis of Pima Indians diabetes and of dermatological diseases
Quantitative prediction of MHC-II binding affinity using particle swarm optimization
THE TWO WORLDS OF AI
Grand Challenges Feasible Miniature Projects
Personal Assistants Wrong blood in tube
Watson Dx diabetes Pima Indians
Nell Missing data in breast ca pt.
USUALLY: USUALLY:
Private sector funding Federal funding
Profit potential Promotion potential
Dissemination likely Dissemination rare
DOES NIH CARE ABOUT HEALTH-CARE AI?
NLM funds healthcare informatics, including AI
(mostly feasible miniatures through standard NIH grant
programs)
NIBIB might be receptive to image , bio-engineering, or
robotics projects
Other Institutes not focused on health-care delivery but
Could be interested in AI for clinical research
Could be interested in secondary use of EHR data
Are interested in computational solutions to genome-related
problems
Why Is NLM Interested in AI?
NLM MISSION IS ABOUT BIOMEDICAL INFORMATION