INTRODUCING OPTICALLY PUMPED MAGNETOMETERS (OPM) FOR MEASURING THE MEG
Prof. Tonio Ball
Jun.-Prof. Joschka Boedecker
Prof. Frank Hutter
Prof. Wolfram Burgard
The general objective of the IMBIT is to investigate the fundamentals of long-term neuroimplants for the bidirectional interaction between the brain and novel autonomous robotic systems as well as to develop practical applications. One of the three reserach areas in the research program is concerned with novel concepts for neuronally controlled robot systems. We have applied for the MBI large-scale facility for this research area in the IMBIT building; it is the basis for the research area "Neurotechnological Human-Robot Interaction" and was recently approved by the DfG. In this research area, the principles of interaction between the brain and novel autonomous robotic systems will be investigated and developed up to practical applications. In particular, robots controlled by brain-machine interfaces will perform service tasks for their paralyzed users. The MBI large-scale device required for this purpose integrates autonomous robot platforms that move in a handling environment relevant to everyday life and can perform manipulation tasks in it (ServiceBots). On the other hand, this experimental environment must be connected to high-quality signal sources of the relevant types of brain activity in humans.
A major obstacle to the application in everyday life is the preparation of the EEG measurement, which was used in our previous Neurobots setup. With gel-free electrodes, time can be saved considerably, but in our experience the signal quality and wearing comfort are poor. In contrast, optically pumped magnetometers (OPM) for measuring the MEG offer a highly interesting alternative because they are contactless and immediately ready for use and have better physical properties than non-invasive EEG and Superconducting Quantum Interference Device (SQUID).
In the present project we have developed novel BCI paradigms, both for the interaction with real robots and in virtual reality, which we will employ to study the potential of OPM-MEG-based neurobotic applications.