Summary

A custom brain-computer interface

Inspired by the exciting progress of the world-leading neurotechnology company Neuralink, we decided to try building a brain-computer interface ourselves. A brain-computer interface (or BCI for short) is a system which senses brain activity, interprets it and then controls a computer using this information. This allows the user to control devices using just their thought, with applications ranging from making computer interactions easier to helping people afflicted by CLIS.

While we didn’t expect to be able to reach the level of Neuralink with our limited ressources and knowledge, we hoped to learn more about this technology and make it more accessible for others. Instead of using an expensive, medical-grade EEG for reading brain activity, we decided to go for the lower cost, open-source Ganglion from OpenBCI and later started construction on our own EEG. We developed our own modular software library to read, process and then interpret the EEG signals using AI. This software library is open-source, in the hopes of making it easier for others to do the same.

Details about our project can be primarily found in our GitHub respository Alexander-Reimer/Interpreting-EEG-with-AI (most up to date) and in the German paper below.

References

Alexander Reimer and Matteo Friedrich. “Steuerung per Lidschluss, Erkennung ereigniskorrelierter Potenziale eines EEGs durch ein neuronales Netz [Recognition of ERPs in EEGs using ANNs]”. In: Physikalisch-Technische Bundesanstalt (PTB). Junge Wissenschaft (2023). DOI: 10.7795/320.202308.