FDA Benchmark Medical Device Flow Models for CFD Validation.

Area of Science:

• Biomedical Engineering
• Medical Device Development
• Computational Fluid Dynamics

Background:

• Computational fluid dynamics (CFD) is crucial for designing blood-contacting medical devices.
• Lack of standardized validation methods for CFD simulations and blood damage predictions hinders device safety evaluation.
• The U.S. Food and Drug Administration (FDA) initiated a study to address these validation gaps.

Purpose of the Study:

• To summarize the FDA initiative on CFD validation for medical devices.
• To report findings from benchmark studies using nozzle and centrifugal blood pump models.
• To assess current CFD techniques and their accuracy in predicting blood flow characteristics.

Main Methods:

• Experimental testing of benchmark models (nozzle, blood pump) to gather velocity, pressure, and hemolysis data.
• Independent computational simulations by over 20 groups to evaluate CFD methods.
• Comparison of experimental data with CFD predictions, focusing on flow separation regions.

Main Results:

• Discrepancies between CFD velocities and experimental measurements were most common in flow separation areas.
• 57% of CFD pressure head predictions for the blood pump model were within one standard deviation of experimental data.
• Only 37% of CFD submissions included predictions for blood damage (hemolysis).

Conclusions:

• Standardized validation is essential for reliable CFD use in medical device development.
• CFD simulations require careful consideration of flow separation for accurate predictions.
• The FDA initiative contributed to developing guidance for computational studies in regulatory submissions.