Science & Research
Standardization of Computational Fluid Dynamic (CFD) Techniques Used to Evaluate Performance and Blood Damage Safety in Medical Devices: An FDA Critical Path Initiative
The purpose of this project is to determine how computational fluid dynamics can be effectively used to characterize fluid flow and to predict blood damage in medical devices. To address this complex issue, FDA has partnered with academia and industry under the Critical Path Initiative program to advance the application of CFD technology in the development and evaluation of medical devices.
In order to better understand the state-of-the-art of how CFD is currently applied to medical devices, a practical evaluation will be conducted of two different flow models, developed by the project’s Technical Steering Committee. Participants in the project will perform computational simulations of the two models using a set of parameters described in the detailed project plan; this is essentially a Round-robin investigation of virtual models. For comparison and validation of the computational simulations, three select laboratories will perform quantitative flow visualization measurements on physical models. The simulation results will be analyzed in a blinded fashion, and then compared and presented collectively in the literature and on this website. Hence, the two models will also serve as benchmarks for future CFD evaluations.
Participants will also be asked to predict the amount of blood damage which might occur for flow through the models under different conditions. For comparison to the computational predictions, in vitro blood damage experiments will be performed at three different laboratories. The comparison will help in extrapolating the engineering results provided by CFD to the actual expected biological response.
The goal of the project is to develop an FDA Guidance Document (and provide information for standards organizations) on the valid use of CFD in the evaluation of medical devices. This will be based on the techniques utilized, the results obtained, and the suggestions made by researchers in the performance of the CFD, flow visualization, and blood damage experiments for this project. This project is open to anyone who wishes to participate.
Round-robin #1: - Conduct the first round robin test on the nozzle design (Sep. 2008 – Dec. 15, 2008).
Round-robin #2: - Conduct the second round robin test on the simple ventricular assist design (starting summer 2009).
We value your comments, suggestions, and collaboration, so please contact us at firstname.lastname@example.org to receive further information!
Rich Malinauskas, PhD
Sandy Stewart, PhD
U.S. Food & Drug Administration
Center for Devices & Radiological Health
Office of Science & Engineering Laboratories
The following video presentations are available which describe the project and how CFD simulations contribute to the regulatory process (from "Computer Methods for Cardiovascular Devices: A Workshop Sponsored by FDA/NHLBI/NSF" held in Bethesda, MD on March 18, 2008):
- R. Malinauskas, "Validation of Computational Fluid Dynamic (CFD) Analysis of Medical Devices: An FDA Critical Path Study"
- S. Stewart, "Computer Methods in Cardiovascular Device Design & Evaluation: Overview of Regulatory Best Practices"
Computational fluid dynamics (CFD) simulations are increasingly being used to determine flow patterns and fluid forces in the development and evaluation of blood-contacting medical devices, such as prosthetic heart valves, blood pumps, and oxygenators. CFD also has the potential for calculating the values of physical parameters (such as shear stress and dwell time) that may affect the level of blood damage the device may cause, including red blood cell damage (hemolysis), platelet activation, and thrombosis. While computer simulations are efficient in creating device designs that can decrease the need for expensive prototyping and laboratory testing, there are no standardized methods available for using CFD techniques to assess safety of the final medical device designs. Validation with experimental data from bench testing is necessary to show that the applied CFD methods are appropriate and accurate. The lack of reliable and standardized test methods in this complex area creates a burden on product developers and the Food and Drug Administration (FDA) in assessing the safety of new devices. This project will determine the limits of the applicability of CFD techniques by comparing computational simulations against suitable experimental models in a round-robin study. The primary goal is to create a Guidance Document for industry-wide use on proper validation and use of CFD models in the assessment of medical device safety.
The primary goal of this project is to accelerate the safety assessment of medical devices in the preclinical stage, with particular attention paid to blood damage, and to standardize CFD techniques for such use. The project will focus on (1) establishing a collaboration among academia, industry, and the FDA to address this goal, (2) developing well-defined computational and physical models to be evaluated in round-robin testing, (3) creating a Guidance Document on using validated computational fluid dynamics in device applications to the FDA, (4) evaluating predictive computational techniques used to estimate blood damage, and (5) disseminating the results of the project through an FDA website, internal and external reports, training, publications, and presentations.
This project fulfills the main goals of the Critical Path Initiative, and FDA’s mission, in the following ways: 1) it establishes a vital collaboration between academia, product developers, and the FDA to address a burden which delays efficient design, implementation, and regulatory review of medical devices, 2) through predictive computational modeling, it targets scientific advances to modernize techniques used to evaluate device safety, 3) it provides needed training for academic programs to help cultivate product development professionals, 4) it is applicable to an at-risk under-served population, as ventricular assist device (VAD) blood pumps are being designed for pediatric use through the National Institutes of Health (NIH) Pediatric Circulatory Support initiative,† 5) it will develop a Guidance Document with techniques for evaluating devices based on computational fluid dynamics which can be used industry-wide, 6) it will accelerate further developments in both computational modeling and in new medical products, 7) it has the potential to reduce the risk and cost of performing animal and human testing, and 8) it will provide technical data for a new standard to be developed by the Association for the Advancement of Medical Instrumentation, on Mechanical Circulatory Support Devices Evaluation, which will include recommendations for CFD and flow visualization. Lastly, the enthusiastic support which this project has received from academia and the medical device industry reinforces its importance to the FDA.
The Fluid Mechanics Laboratory at Center for Devices and Radiological Health (CDRH) will sponsor a computational "round-robin," whereby specially designed computer aided design (CAD) models representing blood contacting medical devices will be supplied to participants in industry and academia. Two archetypes will be developed which cover the critical flow regimes seen in medical devices: a simple nozzle model and a more complex ventricular assist device (VAD) blood pump model. The simple nozzle model is representative of flows occurring in many medical devices due to sudden changes in the geometry of the flow path that includes both laminar and turbulent shear stresses. The VAD model is relevant to the thirty or more blood pump designs currently being developed for adults, and pediatric VAD pumps being developed as part of the NIH Pediatric Circulatory Support initiative.† The exact geometries, flow conditions, and fluid characteristics of the computational models will be supplied to all of the participants. The results will be collated, analyzed, and compared by the FDA in a blinded approach. To validate the computational flow simulations, the results will be compared to detailed laboratory flow measurements made by FDA and the primary collaborators in transparent geometric replicas (using quantitative velocimetry) using the same range of flow parameters. In essence, results from actual laboratory flow models will provide benchmark data for the computational models. To compare how blood safety is assessed, participants will also be invited to use their computational flow results to predict a level of red blood cell damage (hemolysis). This will provide an understanding of the state-of-the-art use of CFD models for predicting blood damage.
Year 1 (Aug. 2007 – Aug. 2008) – Establish a Technical Working Group to design and implement the round robin CFD investigations. Recruit collaborators for the project.
Year 2 (Aug. 2008 – Aug. 2009)
Conduct the first round robin test on the nozzle design (Sep. 2008 – Nov. 2008).
Conduct the second round robin test on the VAD design (start summer 2009)
Year 3 (Aug. 2009 – August 2010)
Complete analysis of round robin test results.
Conduct in vitro blood damage experiments (at select labs) for comparison to CFD predictions.
1. Guidance Document for Computational Studies in Device Review: This will be a guidance for industry and FDA staff in the use of computational fluid dynamics for providing supporting data in device applications, whether as new devices or for changes to existing designs. This will be prepared in collaboration with CDRH’s Office of Device Evaluation. Timeline: A draft document will be prepared by the end of the first year of the project (Aug. 2008), and refinements will be made in FY08 and FY09 as the experimental results are analyzed.
2. National/International Standards: The results of this study will provide needed input into standards development for the use of computational fluid dynamic methods in medical device design and validation. For example, the current version of the ISO standard on heart valve prostheses (ISO 5840, Cardiovascular Implants – Cardiac Valve Prostheses) suggests using computational studies but does not provide any details on how this is to be performed or validated. Also, the laboratory experiments using the transparent flow models will provide publicly available benchmark data which could be used to assess improvements in CFD simulations in the future. Timeline: A draft document will be prepared by the end of the third year of the project (FY09). This study will also provide input into a new standard to be developed by the Association for the Advancement of Medical Instrumentation, on Mechanical Circulatory Support Devices Evaluation, which will include recommendations for CFD and flow visualization.
3. Website/Papers/ Presentations/ Training: The progress and results of this study will be disseminated through an FDA website, presentations at national meetings, training at the FDA, and publications in peer reviewed journals. Timeline: These goals will be reached in years two (FY08) and three (FY09) of the project.
4. Benchmark Flow and Hemolysis Models: The results from this study will be publicly available and will serve as benchmarks for future research and development in this area.
Updated September 24, 2008