Hands-on training Artificial Intelligence (AI) in Medical Devices

Artificial intelligence (AI)-based medical devices are subject to European Regulations 2017/745 (MDR) or 2017/746 (IVDR) and in Germany to the Medical Devices Implementation Act (MPDG). The EU Artificial Intelligence Act (AIA) results in even higher hurdles for market access. Furthermore, many AI-specific standards and guidelines are currently being published (e.g., by the IG-NB Interest Group of Notified Bodies for Medical Devices in Germany) and best practice methods for the development of AI models are constantly being updated.  

The first challenge for manufacturers is to implement the development and evaluation of the AI model as processes in their quality management system in accordance with the respective requirements. Taking AI-specific features into account, other regulatory processes, such as those relating to the software life cycle and risk management as well as clinical evaluation and usability must also be adapted.  

In our hands-on training, we first get to know a fictitious company and its AI-based medical device software. We look at the training and testing of a simple artificial neural network as an AI model, in so far as this is necessary for the regulatory treatment of the product. For this purpose, we use a real and publicly available AI model with a medical purpose as an example. We apply various regulatory templates (in English) step-by-step, which later can be transferred to your product. 

Dr. Thorsten Prinz (VDE, Offenbach/Main)