Bidirectional Electrical Neural Interfaces

Are you curious about how technology can be used to communicate with neurons?

We perform research on electrical neural interfaces and develop technologies that enable bi-directional communication with neurons.

The most prominent results of our efforts are fully immersible digital neural recording probes. These probes consist of the probe shank only and thus do not have a bulky probe base. They are implemented as a regular and modular architecture using an incremental first-order Delta-Sigma ADC under each of the 144 electrodes. A serial 4-wire data interface allows for simultaneous readout of all channels. Moreover, the probes are robust against illumination artifacts and electromagnetic interference due to a clever design.

Comparison of different probe architectures with our probe highlighted in red (source: own picture).

A further development of the probes using a two-stage incremental Delta-Sigma ADC enabled a reduction of the area and power requirement per recording channel by 20 % from 0.0049 mm² to 0.00378 mm² and by 65 % from 39.14 µW to 13.94 µW, respectively. The recordings of action and local field potentials are performed with an energy requirement of only 429 pJ/conversion.

For numerous applications, stimulation of neurons is essential in addition to recording neural information. Our lightweight, battery-powered neural headstage enables the generation of temporally synchronized stimuli for freely acting animals. Wireless configuration is accomplished using Bluetooth as well as a user interface for visual configuration of stimuli generated using amplitude modulation, time shifting, and mono/biphasic pulses. Stimulation can be monaural or binaural.

Neural recorders with high robustness to spurious signals define another focus of our research. These are needed to derive neural information despite artificial stimulation and are typically implemented as a closed-loop system that enables on-chip stimulation, derivation and processing of neural signals. By cancelling stimulation artifacts based on stochastic signal processing at the analog front-end, our recorder is able to derive neural signals down to 2 mV with stimulation artifacts of up to 100 mV.

If you're interested in studying microelectronics or simply fascinated by the intersection of technology and neuroscience, our research offers a wealth of exciting opportunities to explore. Whether you're a non-expert, student, or expert in the field, we invite you to learn more about our work and the possibilities it holds for the future.