BMBF Bundesministerium für Bildung und Forschung

Research-Campus MODAL


Large-scale data analysis for next generation health care

Health is becoming increasingly important in our daily lives and in society as a whole. As a result, we are investing more than ever before in physical and mental fitness and an overall healthier lifestyle. Not only as a result of demographic change, there is a growing general urge to optimize ourselves in various areas through an individual health management – and thus – in the long run – achieve a better and longer life. This also changes the perception of one’s own role in the health system – towards the so-called informed patient, who has a detailed overview of his or her various health data and can comprehensively analyze them with the help of intelligent systems. Extensive knowledge, e.g. about one’s own genome, enables the targeted use of drugs – in case of some disease – and optimal planning of therapies.

Two important components in this system are on the one hand the required data on the various body signals, such as pulse, heart rhythm (ECG) or the concentration of various minerals, vitamins or other important proteins. These can come from clinical systems, specialized laboratories or even wearables, with which a wide variety of body signals (e.g. ECGs) can already be recorded professionally and almost continuously. The second important component are the involved methods (or algorithms) needed for data analysis. These must not only be safe to use (think of data protection) but also provide comprehensible analysis results. In view of the constantly growing amount of digitally available information, it should also be possible to permanently compare one’s own data with the latest medical findings – for example, to identify possible risks early on or to receive suggestions for an even better lifestyle adapted to the individual needs.

In the MedLab, we are working on the necessary new algorithms that will make this vision a reality. A particular focus is on the development of new methods in the field of interpretable artificial intelligence. This will enable new types of intelligent services for the digital healthcare system of the future. The basis for this is the expansion of the established so called point-wise clinical data analysis to include the continuous analysis of very large data streams from mobile health applications. This opens up new possibilities in the field of early diagnosis, disease management and therapy monitoring. Furthermore, it will allow creation of new medical instruments based on real-time monitoring of high-risk patients and enable better therapy optimization.