Reliability Prediction Framework for Implantable Medical Devices
Wednesday, February 24, 2016 - 11:25am - 11:50am
Wei Gan (Medtronic)
Demonstrating high reliability of implantable devices providing life-saving therapies is imperative in the medical device industry. While rigorous and conservative reliability testing occurs before product release, these tests occur for a final design that will be released to market and do not directly guide the design process. Instead, obtaining device reliability predictions in the design process ensures that the most robust design is selected from a family of device architectures. These predictions are obtained from a probabilistic approach that is capable of demonstrating reliability of medical devices subject to in vivo use conditions over a virtual population of differing patients. This process encompasses the use of physics-based model predictions, surrogate model development and calibration to bench-level component data when available, and uncertainty quantification and propagation. The framework of this process is outlined and implementation challenges within industry are highlighted.