Computational Modeling Is Transforming How Medical Devices Are MadeComputational Modeling Is Transforming How Medical Devices Are Made

Computational modeling is becoming increasingly important in the development and testing of medical devices, explained Matthew Myers, PhD, a research physicist at FDA in a presentation at BIOMEDevice Boston. The agency is now working with collaborators to create computer models incorporating radiological imaging data of healthy and diseased anatomy for a variety of diseases. FDA plans on ultimately integrating these models with genomic and physiological data to create complete physiological models and simulations for testing. 

Ultimately, the technology could transform the path to market for device markers, and at present, it is most commonly used in the pre-submission stage where it is mainly used for device optimization and has proven useful in comparing different product designs.

For instance, computational modeling can be used in early product development of reusable medical devices to simulate the amount of debris from blood, soft tissue, and bone fragments that can cling to the surface of devices such as arthroscopic shavers, endoscopes, and surgical instruments. At present, the technology can be used to locate regions of poor mixing in a design using velocity vectors, particle paths, vorticity plots, and so forth. Additionally, it can be used to compare the degree of debris retention in multiple device designs.

At present, computation modeling, however, cannot predict the absolute levels of retention of a particular debris type such as blood cells or bone fragments. It is also unable to predict debris levels retained after cleaning.

At the post-submission stage, computation modeling is well suited for forensics to detect the cause of a device failure or under-performance. That is, it can help explain why things went wrong.

The use of computational modeling at the submission stage is currently less than either pre- or post-submission. Its importance at the submission stage is growing, Myers explained, The technology could potentially be used to reduce the amount of time and money invested in bench testing, animal studies, and clinical trials. 

While the technology now can't be used to make absolute predictions of, for instance, blood damage, thrombus formation, and aggregation of debris on device surfaces, it is useful for making relative predictions. This fact makes computational modeling well suited for studying the affects of minor design changes and comparing predicate devices. 

 Brian Buntz is the editor-in-chief of MPMN. Follow him on Twitter at @brian_buntz.