Identifying potential training factors in a vibrotactile P300-BCI

Abstract

Brain–computer interfaces (BCI) often rely on visual stimulation and feedback. Potential end-users with impaired vision, however, cannot use these BCIs efficiently and require a non-visual alternative. Both auditory and tactile paradigms have been developed but are often not sufficiently fast or accurate. Thus, it is particularly relevant to investigate if and how users can train and improve performance. We report data from 29 healthy participants who trained with a 4-choice tactile P300-BCI during five sessions. To identify potential training factors, we pre-post assessed the robustness of the BCI performance against increased workload in a dual task condition and determined the participants’ somatosensory sensitivity thresholds with a forced-choice intensity discrimination task. Accuracy (M = 79.2% to 92.0%) and tactually evoked P300 amplitudes increased significantly, confirming successful training. Pre-post somatosensory sensitivity increased, and workload decreased significantly, but results of the dual task condition remained inconclusive. The present study confirmed the previously reported feasibility and trainability of our tactile BCI paradigm within a multi-session design. Importantly, we provide first evidence of improvement in the somatosensory system as a potential mediator for the observed training effects.

Recording the tactile P300 with the cEEGrid for potential use in a brain-computer interface

“…As in the cEEGrid data, ERP polarity of P37 was inversed. Eidel et al 10.3389/fnhum.2024.1371631 Frontiers in Human Neuroscience 10 frontiersin.org study (74.9% in session one, Eidel and Kübler, 2022). In contrast to the results of the prior study, however, we found no significant correlation between the somatosensory sensitivities and (offline) accuracies.…”

“…Table 2 provides an overview of the determined sensitivity thresholds per participant. As in our last study (Eidel and Kübler, 2022), some participants quickly revealed a distinct threshold, whereas the responses of other participants were less consistent, necessitating more trials (on average, 37 ± 8.7 trials were applied). The equal/unequal responses from P20 were still too inconsistent to calculate a threshold, despite performing 53 trials, and were excluded from the analysis.…”

“…Although it is hard to compare these different paradigms, the relatively high above-chance rate seems encouraging, as it shows that a meaningful information transfer may be possible for most users. Accuracy achieved with the tactile P300 BCI has been shown to improve with training (Eidel and Kübler, 2022). We hypothesize that such improvement would also occur when using the cEEGrid, though this must be tested in further studies with healthy and physically impaired participants alike.…”

“…Although the performances of tactile P300 BCIs are usually lower than those of visual paradigms (see Eidel and Kübler, 2020 for an overview), we have repeatedly found that training led to significant performance increases among healthy participants ( Herweg et al, 2016 ; Eidel and Kübler, 2020 ; Ziebell et al, 2020 ). Latest evidence further suggested that somatosensory sensitivity correlated with accuracy and with performance increases across multiple sessions ( Eidel and Kübler, 2022 ).…”

Modulation of Human Frontal Midline Theta by Neurofeedback: A Systematic Review and Quantitative Meta-Analysis

“…However, in contrast to protocols with continuous feedback, ERP-based protocols would require intermittent feedback on the basis of specific trials, which overall implies that less frequent feedback is provided. Nevertheless, a learning curve for increasing ERP amplitudes after training has been demonstrated in the field of brain-computer interfacing (BCI, Eidel and Kübler, 2022; Ziebell et al, 2020) and, in general, observations from BCI research may have great potential to be transferred to the domain of FMT-neurofeedback, as already suggested by Jeunet et al (2018). Specifically, they propose the use of adaptive thresholds, machine learning tools for customizing neurofeedback features, applying performance predictors for selecting suitable neurofeedback protocols, integrating tactile and computerized social and emotional feedback to boost motivation, and the identification of optimal mental strategies through algorithms initially developed for BCI to enhance individual training outcomes.…”