Together with partners such as RWTH Aachen University, Aachen University Hospital and Forschungszentrum Jülich, we are working on the next generation of retinal implants as part of the Research Training Group GRK 2610 “InnoRetVision”. 

Our research includes

• The design of 3D needle electrode arrays that enable more precise stimulation of retinal target cells. A scalable CMOS-compatible process for the production of needle electrodes using galvanic electrodeposition has already been developed, which will be used in the retina and to clarify neuroscientific questions in the further course of the project. 
• The development of adaptive closed-loop stimulators that read out neuronal signals in real time and automatically adapt the stimulation strategy. Several application-specific integrated circuits (ASIC) have already been designed and industrially manufactured as the first stimulators. Novel stimulation protocols such as high-frequency sinusoidal stimulation will be made possible and researched in the course of the project. Furthermore, the ASICs enable adaptive stimulation, which measures the neural signal during stimulation and can therefore optimize the stimulation strategy in real time.

The NEON project, initiated by research teams from the University of Duisburg-Essen, Ilmenau University of Technology and Essen University Hospital, is developing a mobile, contactless measuring device for recording vital parameters such as pulse, respiratory rate, body temperature and oxygen saturation in patients. The aim is to reduce the risk of hospital-acquired infections by eliminating direct physical contact. The device should also automatically assess the patient's state of health and be applicable to other diseases using artificial intelligence methods. The focus is on both acute monitoring and long-term observation of the patient's state of health.

The Sp:AI:ke-funded MERCUR research project aims to develop hardware solutions of an end-to-end neural processing chain for so-called end-to-end Brain Computer Interfaces (BCI) systems that incorporate online algorithms for spike sorting and neural decoding with analog and digital embedded hardware. This pipeline is a fundamental building block for the development of the next generation of neural implants. The system relies on state-of-the-art machine learning algorithms to decode movement intentions from brain activity. Recording and interpreting brain signals requires a solution that is accurate, fast, energy efficient, can be implemented on a small chip footprint and has low heat dissipation. Existing solutions cannot meet these requirements.