Life Prediction Method for Superelastic Nitinol in Safety Critical Medical Devices

Fatigue life analysis of superelastic nitinol in medical device applications is challenging due to uncertainties involved in the determination of boundary conditions and in the stress analysis associated with complex design geometries. This is further compounded by the lack of accurate statistical information on the fatigue life distribution of the material.  Conventional life prediction has been based on deterministic approach using median survival data. However, the specific statistical distribution of fatigue life for superelastic nitinol has not been rigorous described in the literature. Reliability analysis often makes use of normal or log-normal distribution to derive a reliability-based life prediction. For implantable medical device applications where the reliability requirement is high, the median survival data may not provide the necessary level of confidence for patient safety. The present experimental program consists of fatigue testing of various nitinol specimens in a simulated physiological environment for a minimum survival of 400 million cycles, or 10 equivalent years of patient life. The results will be analyzed using various statistical distribution functions. In particular, the focus is on the examination of the most suitable life prediction model. The overall goal is to present a statistical framework for predicting the durability performance of structure-critical medical devices with a confidence level meeting the requirements of international standards, regulatory guidance and clinical practice for patient safety.