curATime selected as a research cluster by the Federal Ministry of Education and Research

Future cluster curATime with the Signal Processing Group's subproject curAIsig approved.

2022/08/05

The curATime research project with the participation of Darmstadt Technical University has been selected by the German Federal Ministry of Education and Research (BMBF) as a new future cluster. For an initial, three-year implementation phase, curATime will receive funding totaling 15 million euros. The aim of the cluster, which is based in the Rhine-Main region, is to develop customized treatment methods and prevention concepts for cardiovascular diseases. The Department of Electrical Engineering and Information Technology at the TU is involved with the sub-project curAIsig, which is being funded with around 412,000 euros over 36 months.

The curATime network combines novel immunological methods with AI-supported biomarker research that identify and combat inflammatory changes in blood vessel walls at an early stage. The cluster is coordinated by TRON gGmbH, an independent, nonprofit translational research organization based in Mainz. Also involved are Mainz University Medical Center and the German Research Center for Artificial Intelligence (DFKI), as well as other partners from the Rhine-Main-Palatinate region.

Subproject curAIsig

The Darmstadt subproject curAIsig provides innovative technologies in system-oriented biomedical research. The interdisciplinary team around project coordinator Prof. Abdelhak M. Zoubir, Dr.-Ing. Michael Muma and Prof. Philipp Wild, head of Clinical Epidemiology and Systems Medicine at the University Medicine Mainz develops and evaluates robust signal processing and statistical learning methods for high-dimensional biodatabases for the discovery of novel and reproducible biomarker signatures. curAIsig addresses problems that cannot be satisfactorily solved by existing artificial intelligence (AI) methods. Clinically interpretable, reproducible, statistically demonstrably robust models cannot be obtained with classical AI (neural networks, deep learning). However, non-robust models often lead to false-positive or false-negative results.

Methodologically, curAIsig builds on the research and expertise in robust signal processing of Zoubir's Signal Processing Group, as well as in parts on methods developed by Muma under his ERC grant “ScReeningData”. They enable computational learning to detect relevant clinical information in highly complex biomedical data. In addition, the team for curAIsig leverages interdisciplinary expertise in already established collaborations. The researchers can access large resources of existing biodatabases. This enables efficient and goal-oriented technology development within the curATime cluster.

clusters4Future

curATime is one of a total of seven future clusters in the second round of the open-topic “Clusters4Future” initiative, which are now entering the implementation phase. The second round was opened in November 2020 with a total of 117 applicants. As one of 15 finalists nationwide, curATime had started the six-month conception phase funded by the BMBF in October 2021.

With the “Clusters4Future” initiative, the BMBF supports regional innovation networks that make a special contribution to knowledge and technology transfer. The clusters of the future are directly linked to cutting-edge research and are intended to ensure that technological and social innovations reach people's everyday lives more quickly. At the same time, the clusters are intended to help advance Germany as a location for research and development both nationally and internationally.

The original article (in German) first appeared here.