Towards Universal Human-Machine Interfaces: A real-life Brain-Computer Interface Case Study

Modern-day human-machine assistive devices are often individually customized such that they suite the specific needs for a particular subject. While this approach yields an optimal solution for the patient, it is time-costly and expensive. Moreover, such solution is not applicable for every patient, for example patients suffering from a disease of neurological origin. Most human-machine interfaces rely on a persons working efferent pathways, in order to operate the assistive device and thus is not universally applicable by definition. More recently, Brain-Computer Interfaces (BCIs) have attracted attention because they allow people with disabilities of neurological origin to control external applications. Therefore, they seem to be a promising alternative as universal human-machine interface. This work extends the feasibility report for Brain-Computer Interface development in [1] with approaches from [2] and [3] applied in the Brain-Computer Interface Race of the Cybathlon championship for athletes with disabilities. The author of [1] carefully weigh up and listed the advantages and disadvantages for BCI development summarized under the keywords of universality, equipment, sensitivity, and timing issues that hamper the development and usage of a BCI. New findings in the field of neurofeedback proposed in [2] enable to learn operating a BCI fast and reliable, thus guarantee less sensitivity, smaller timing windows, and more task enjoyment for the subject. Moreover, the combined methods from [2] and [3] allow for universal access of the BCI as a device for human-machine interfacing. These hypotheses are supported by the experimental results with five subjects in [2] and ten subjects in [3] and their validity was shown by statistical tests. Finally, while the BCI adapts to the patient [3], using a BCI should be possible for anyone after training [2]. References [1] Felzer, Torsten. "On the possibility of developing a brain-computer interface (bci)." Technical University of Darmstadt, Department of Computer Science, Darmstadt (2001). [2] Faber, Natalie, et al. "Neurofeedback for State of the Art Paradigms in Brain-Computer Interfacing." Cybathlon Symposium, Zürich (2016). [3] Friess, Tamara, et al. "Personalized Brain-Computer Interfaces for Non-Laboratory Environments." Cybathlon Symposium, Zürich (2016). * Presenting author Attending author(s)