The program consists of eight modules relating to themes of health, law, and technology, and there will be two sessions for every module. The modules are the following:
AI is presenting both new challenges and opportunities in medicine. In terms of the former these include new possibilities not only to detect and treat illness but even potentially to prevent or reduce the likelihood of it occurring in the first place. It will also hopefully permit a reorganisation of healthcare systems allowing for a more efficient use of available resources. In terms of the latter, it will bring both new ethical and regulatory challenges. These must be addressed in order to ensure that new technological developments occur in the most patient centric way possible.
This module will introduce artificial intelligence, as applicable for a non-mathematical audience, including all approaches relevant to medical devices (with the exception of generative and large language approaches which are covered in the Module 3). The features of AI that make it ‘special’ from a regulatory perspective will be described. The developing EU regulations for AI will be described, including how ‘vertical’ medical device regulation and guidance related to the sector, and how it will be affected by ‘horizontal’ legislation (the AI Act) and standards. Challenges and gaps in AI regualion will be considered.
Although a subset of deep learning AI, large language models (LLMs), and other transformer-based generative AI approaches bring with them larger promise and large challenge associated with their use in medicine and medical devices. These potentials and challenges will be described in detail, along with an overview of the current state of development of technologies on-market, and the challenge for regulators in deciding how to apply and enforce regulation in this area.
The sharing of health data is critical to both performing good medicine and to research associated with that aim. Despite advances in the digitisation of patient dossiers (e.g. the EHR) and improved online connectivity great challenges remain. These are not only technical, but also regulatory and even in many cases cultural. Amongst initiatives to improve the sharing of data in this area is the EU’s proposal for a European Health Data Space. The aim behind it is to increase the sharing of electronic health data both for primary (i.e. the treatment of patients) and for secondary (i.e. scientific research, innovation and other) purposes.
Medical device, pharmaceutical regulation and specific aspects of current and developing health data law will be described, as they relate to this diverse, complex and increasingly pervasive family of technologies. Groupings of these concepts in new modes of healthcare delivery will be described, alongside the unique potential and risks these advances bring with them.
The product design process for regulated medical devices will be explored in detail, and we will examine to what degree the most creative and reactive forms of ‘design thinking; can be applied in software-based medical devices developed in quality management systems. To what degree can experimentation with design be carried out within ‘Design Control’, and how might this already complex field be further challenged by flexible aspects of design that may be brough by generative AI influence interfaces and products?
Modern healthcare is generating enormous amounts of data. This trend will intensify further with the increasing use of AI in healthcare. AI processes not only generate enormous amounts of data, but also require such data, including during the training process. Ownership and the right to use the data generated by and required for such processes will be an important issue. The applicability of Intellectual Property framework in this area may therefore be influential to how such processes are able to evolve.
At a time when new technological paradigms are offering the potential to radically reorganise how healthcare is conducted and organised, it is important not to lose sight of the healthcare provider perspective. Medical professionals often have a unique oversight that spans knowledge of patient needs and perceptions, health care organisation and best practices in terms of treatment. It is therefore essential to consider the healthcare provider perspective in order to ensure that vital requirements are not missed. This includes in the design, training, and evaluation of AI driven processes. It is also important to discern when and how they should fit into AI assisted decision making processes, something that is considered an indispensable AI requirement.
