The aging of the population poses an increasing burden on the healthcare system, which can be alleviated by artificial intelligence: it can provide health advice to patients tailored to their personal needs and current context, and can support treatment decisions of doctors based on all available data. There an ever-growing amount of such data, much of it coming from increasingly ubiquitous mobile and other sensing devices. This is the field of research of the Ambient Intelligence Group, of which the young researcher will become a member.
The PhD topic will be defined according to the interests of the young researcher, but will concern artificial intelligence in relation to health, well-being, and possibly sensor data. Furthermore, we will strive to include methods of generative artificial intelligence, including large language and multimodal models.
A potential topic is the development of generative models for the creation of synthetic health data, which circumvent the privacy issues while still retaining all the important relations of real data. Such models can be used for simulating various interventions and predicting health outcomes and risks. Another example is the conversion of data about people – sensor, tabular, and other – into text and reasoning about it with the help of large language models. A third example is the use and adaptation of large language models for interaction with patients in natural language, and the generation of advice and interventions such as cognitive behavioural therapy. If the young researcher is keen on an interesting topic from the field of artificial intelligence that does not concern health or generative models, that is also a possibility.