Computational Model Credibility

Question 1

computational models

Answer 1

• provide information that evaluates the technical performance, safety, and/or effectiveness of medical devices
• used throughout the total product life cycle: validating initial concept/design/development, supporting nonclinical and clinical activities, and providing post-market surveillance
• device evaluations that may not be possible experimentally, and simulating in vitro/in vivo evaluations

Question 2

ASME V&V 40

Answer 2

• international standard on assessing credibility of CM through verification and validation
• models need to appropriately represent reality
• underlying focus is model risk: possibility that the CM leads to an incorrect decision resulting in an adverse outcome

Question 3

Context of use

Answer 3

• defines the specific role and scope of the computational model
• includes a detailed statement of what will be modelled and how the model outputs will be used
• also includes description of other supporting evidence and how much the CM contributes

Question 4

Model influence

Answer 4

• contribution of the computational model relative to other contributing evidence in making a decision
• one portion of model risk

Question 5

Decision consequence

Answer 5

• significance of an adverse outcome resulting from an incorrect decision (usually in the context of potential patient harm)
• second portion of model risk

Question 6

Model credibility

Answer 6

• trust in the predictive capability of a model for the context of use
• trust can be established by collecting evidence from credibility activities

Question 7

Code verification

Answer 7

• goals: identify and remove errors in the source code and numerical algorithms of the computational software
• software quality assurance (software functions correctly and produces repeatable results)
• numerical code verification (correct implementation and functioning of numerical algorithms)

Question 8

Model verification

Answer 8

• calculation verification: estimating numerical error in the quantities of interest (involves estimating discretization, numerical solver, and user errors)

Question 9

Model validation

Answer 9

• assessing the degree to which the computational model is an appropriate representation of reality
• computation model validation assesses model form and inputs (governing equations, system configuration, conditions, etc.)
• comparators provide data against which results are evaluated (bench testing, experiments, etc.)
• perform assessment of accuracy of the simulation output after the V&V activity outputs are compared

Question 10

aneurysm

Answer 10

• degradation and bulging of the wall of a blood vessel that supplies blood to the brain