Tobias Kaufmann1, Andrea Kübler. 1. Department of Psychology I, University of Würzburg, Marcusstr. 9-11, 97070 Würzburg, Germany. Institute of Clinical Medicine, University of Oslo, Kirkeveien 166, 0450 Oslo, Norway.
Abstract
OBJECTIVE: The speed of brain-computer interfaces (BCI), based on event-related potentials (ERP), is inherently limited by the commonly used one-stimulus paradigm. In this paper, we introduce a novel paradigm that can increase the spelling speed by a factor of 2, thereby extending the one-stimulus paradigm to a two-stimulus paradigm. Two different stimuli (a face and a symbol) are presented at the same time, superimposed on different characters and ERPs are classified using a multi-class classifier. Here, we present the proof-of-principle that is achieved with healthy participants. APPROACH: Eight participants were confronted with the novel two-stimulus paradigm and, for comparison, with two one-stimulus paradigms that used either one of the stimuli. Classification accuracies (percentage of correctly predicted letters) and elicited ERPs from the three paradigms were compared in a comprehensive offline analysis. MAIN RESULTS: The accuracies slightly decreased with the novel system compared to the established one-stimulus face paradigm. However, the use of two stimuli allowed for spelling at twice the maximum speed of the one-stimulus paradigms, and participants still achieved an average accuracy of 81.25%. This study introduced an alternative way of increasing the spelling speed in ERP-BCIs and illustrated that ERP-BCIs may not yet have reached their speed limit. Future research is needed in order to improve the reliability of the novel approach, as some participants displayed reduced accuracies. Furthermore, a comparison to the most recent BCI systems with individually adjusted, rapid stimulus timing is needed to draw conclusions about the practical relevance of the proposed paradigm. SIGNIFICANCE: We introduced a novel two-stimulus paradigm that might be of high value for users who have reached the speed limit with the current one-stimulus ERP-BCI systems.
OBJECTIVE: The speed of brain-computer interfaces (BCI), based on event-related potentials (ERP), is inherently limited by the commonly used one-stimulus paradigm. In this paper, we introduce a novel paradigm that can increase the spelling speed by a factor of 2, thereby extending the one-stimulus paradigm to a two-stimulus paradigm. Two different stimuli (a face and a symbol) are presented at the same time, superimposed on different characters and ERPs are classified using a multi-class classifier. Here, we present the proof-of-principle that is achieved with healthy participants. APPROACH: Eight participants were confronted with the novel two-stimulus paradigm and, for comparison, with two one-stimulus paradigms that used either one of the stimuli. Classification accuracies (percentage of correctly predicted letters) and elicited ERPs from the three paradigms were compared in a comprehensive offline analysis. MAIN RESULTS: The accuracies slightly decreased with the novel system compared to the established one-stimulus face paradigm. However, the use of two stimuli allowed for spelling at twice the maximum speed of the one-stimulus paradigms, and participants still achieved an average accuracy of 81.25%. This study introduced an alternative way of increasing the spelling speed in ERP-BCIs and illustrated that ERP-BCIs may not yet have reached their speed limit. Future research is needed in order to improve the reliability of the novel approach, as some participants displayed reduced accuracies. Furthermore, a comparison to the most recent BCI systems with individually adjusted, rapid stimulus timing is needed to draw conclusions about the practical relevance of the proposed paradigm. SIGNIFICANCE: We introduced a novel two-stimulus paradigm that might be of high value for users who have reached the speed limit with the current one-stimulus ERP-BCI systems.
Authors: Álvaro Fernández-Rodríguez; María Teresa Medina-Juliá; Francisco Velasco-Álvarez; Ricardo Ron-Angevin Journal: Med Biol Eng Comput Date: 2021-03-20 Impact factor: 2.602
Authors: Nadine Simon; Ivo Käthner; Carolin A Ruf; Emanuele Pasqualotto; Andrea Kübler; Sebastian Halder Journal: Front Hum Neurosci Date: 2015-01-09 Impact factor: 3.169