Title: Confidence in neural networks: methodological issues arising from a review of safety-related applications
1Confidence in neural networksmethodological
issues arising from a review of safety-related
applications
- P.J.G. Lisboa
- p.j.lisboa_at_livjm.ac.uk
Computing and Mathematical SciencesLiverpool
John Moores University
2Outline
- Developments in commercial safety-related systems
comprising artificial neural networks - Increasing demand for decision support e.g.
healthcare - Where is the evidence of healthcare benefit from
ANNs? - Framework for assuring confidence in neural
networks - Design assurance
- Risk analysis
- Evidence of effectiveness
- Methodological issues arising from the review
3Fire alarm for office blocks
- SiemensFP11
- FirePrintTechnology
- Very high specificity
4Commercial safety-related systems
- Automotive industry
- Tow-by-wireNACT (http//www.mech.gla.ac.uk/nact
/nact.html) - Fuel injectionFAMIMO (http//iridia.ulb.ac.be/fa
mimo/) - Electronic ABSH2C (http//www.control.lth.se/H2C
/)
Lisboa, P.J.G. Industrial use of safety-related
artificial neural networks HSE CR, 2001
http//www.hse.gov.uk/crr_pdf/crr01327.pdf
5Papnet
- Cytology screening
- FDA approvedfor secondary screening
- Proven sensitivity
- Specificityleft to user
- Cost-effective
6Epidemiology of medical error
- In the US 44,000 - 98,000 preventable deaths
attributed to medical errors (Weingart et al, BMJ
2000) - Exceeds combined toll from
- motor crashes
- suicides
- falls
- poisonings
- drowning
7Epidemiology of medical error
- Managing error
- The just-world hypothesis
- Systemic approaches (Reason, BMJ 2000)
8Randomised Controlled Trials
Lisboa, P.J.G. A review of evidence of health
benefit from artificial neural networks in
medical intervention, Neural Networks, 15, 1,
9-37,2002.
9Clinical Trials
10(No Transcript)
11(No Transcript)
12Is there evidence of clinical benefit ?
- Clinician performance ? patient outcome
- Primary to secondary care referrals of patients
with third molars Sens. Spec. Acc. - 1) Control group 0.97 0.22 0.83
- 2) Paper-based clinical algorithm 0.56 0.93 0.73
- 3) MLP-based recommendation 0.56 0.79 0.67
- Which is the best performing system ?
13Is there evidence of clinical benefit ?
- Clinician performance ? patient outcome
- Primary to secondary care referrals of patients
with third molars Sens. Spec. Acc. - 1) Control group 0.97 0.22 0.83
- 2) Paper-based clinical algorithm 0.56 0.93 0.73
- 3) MLP-based recommendation 0.56 0.79 0.67
- 1) 1.2 2) 8.0 3) 2.7
14Performance estimation
- ROC framework
- Boost factorPPV True
positives/Predicted positives
15Predictive models
Deficiencies in standard modelling
methods (Altman Royston, Stat. Med. 2000) 1.
Overoptimistic assessment of predictive
performance 2. Multiple regression using stepwise
variable selection 3. EPV lt 10
(samples/parameters) 4. Case-mix (cohort
variations) 5. External evaluation (protocol
changes)
- Retrospective vs. Temporal
- Prospective vs. External
16Continuum of inference models
numeric to numeric
symbolic to symbolic
numeric to symbolic
unsupervised
supervised
data driven
knowledge driven
statistical methods
signal processing
neural networks
k-means clustering
kernel methods inc. SVM
FFT
production rules
SOM/GTM
CART
logistic regression
multi-layer perceptron
control
ART
rule induction
axiomatic
wavelets
radial basis functions
independent components analysis
reinforcement learning
fuzzy logic
rule extraction
17Software life-cycle
18Risk analysis
Extract from the FDA guidelines
19The continuum of evidence(Drug development)
20The continuum of evidence(Campbell et al, 2000)
21The continuum of evidence
- Ph I Theory
- Regularisation framework
- Ph II Performance optimisation
- Complexity control
- Ph III Generalisation
- HAZOP/FMEA
- RCT case-control study
- Clinical effectiveness
22Medical Devices Directives
The continuum of evidence
- Ph I Theory
- Regularisation framework
- Ph II Performance optimisation
- Complexity control
- Ph III Generalisation
- HAZOP/FMEA
- RCT case-control study
- Clinical effectiveness
- Essential requirements
-
- Performance validation
- Doctrine of Substantially Equiv. Products
- Model evaluation
- Requirement for Learned Intermediaries.
- Risk assessment
- Procedure for post-marketing surveillance
- H S requirements
23Methodological issues arising
- Confidence
- Data
- Regularisation
- Calibration
- Transparency
- Rule-extraction
- Linear-in-the-parameters statistical inference
- Fuzzy or rule-based supervisory models
- Effectiveness
- Performance estimation
- Reliability
- Novelty-detection
24Generalisation
25Performance estimation
- Power calculations
- Bootstrap Ci
26Rule extraction
- Axis parallel boxes network pruning
27Rule extraction
- Axis parallel boxes network pruning
Lisboa, P.J.G., Etchells, T.A and Pountney, D.C.
Minimal MLPs do not model the XOR logic
Neurocomputing, Rapid communication, 48, 1-4,
1033-1037, 2002.
28Good practice
- Embodying a safety-culture
- Specification
- Statistically significant vs. clinically useful
- Transparency
- Verify against clinical prior knowledge
-
- HAZOP FMEA
- Reliability - novelty detection ?
- Maintainability - incremental learning ?
29Summary
- Assuring confidence in complex is not a specific
issue for neural networks but applies to all
inference systems - A framework can be constructed based on a
life-cycle model of safety-related software - Good practice in the design data-based models
- Need to switch between evidence-based and
knowledge-based cultures for v v.