NeuXus: A Biosignal Processing and Classification Pipeline for Real-Time Brain-Computer Interaction

In the last few years, Brain-Computer Interfaces (BCI’s) have progressed as an emerging research area in the domain of neurotechnology. This is primarily due to the introduction of low-cost electroencephalographic (EEG) systems that render BCI technology accessible for non-medical research, but also due to the advancements of signal processing and machine learning methods. Consequently, BCI’s can now provide a wide new range of possibilities in the way users interact with a computer system (e.g., neuroadaptive interfaces). However, major challenges must still be addressed for BCI systems to mature into an established communication medium for effective brain-computer interaction. One of the major challenges involves the easy integration of real-time processing pipelines with portable EEG systems for out-of-the-lab use. To date, despite the amount of options current open-source tools provide, most toolboxes focus mainly in extending the processing and classification methods but lack on the ability to provide an easy-to-design yet extensible architecture for ubiquitous use. Here, we present NeuXus, a modular Python toolbox for real-time biosignal processing and pipeline design. NeuXus is open-source and platform independent, providing high-level implementation of processing pipelines for easy BCI design and deployment, including for simultaneous EEG-fMRI acquisitions through real-time gradient artifact correction.