Neural mechanisms of training an auditory event‐related potential task in a brain–computer interface context

Halder, Sebastian; Leinfelder, Teresa; Schulz, Stefan M.; Kübler, Andrea · 2019 · Crossref

DOI: 10.1002/hbm.24531

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the neural mechanisms underlying training for an auditory event-related potential (ERP) brain–computer interface (BCI). While BCIs offer communication solutions for individuals with severe motor impairments, approximately one-third of users fail to achieve control, highlighting the need to understand how training optimizes performance. The authors aimed to identify specific brain activation changes associated with learning an auditory P300 BCI task and to determine if these mechanisms overlap with those of motor imagery BCIs. The research involved ten healthy volunteers who underwent two functional magnetic resonance imaging (fMRI) sessions: one before (t1) and one after (t5) three sessions of EEG-based auditory BCI training (t2–t4). The BCI task required participants to attend to specific animal sounds presented in a virtual matrix to select letters, ignoring other sounds. During fMRI scans, participants performed the same auditory attention task without online feedback. The study utilized a rapid-presentation event-related fMRI design to measure blood oxygen level-dependent responses. EEG data from the training sessions were used to verify performance improvements, measured by information transfer rate (ITR), while fMRI data were analyzed to contrast brain activation between attended and ignored stimuli and to compare pre- and post-training states. Results demonstrated that training significantly improved BCI performance, with the mean ITR increasing from 2.9 bits/min in the first training session to 4.0 bits/min in the final session. Neural analysis revealed that attending to stimuli strongly activated the putamen, supplementary motor area (SMA), and superior temporal gyrus (STG). Crucially, training induced specific neural changes: activation decreased in the superior frontal gyrus, while the STG and supramarginal gyrus exhibited a stronger hemodynamic rebound. These findings suggest that training enhances stimulus perception while reducing mental workload. Furthermore, a conjunction analysis revealed overlapping activation in the SMA and putamen between this auditory ERP task and a previously studied motor imagery BCI task, indicating shared neural mechanisms across different BCI paradigms. The study concludes that effective training for auditory ERP-BCIs involves improved perceptual processing and reduced cognitive load, reflected in distinct changes in frontal and temporal brain regions. The identification of common neural substrates between auditory and motor imagery BCIs suggests that general principles of BCI learning may apply across different paradigms. These insights can inform the development of optimized training protocols and novel BCI designs, potentially improving success rates for users who initially struggle to control these interfaces.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• neuro workload indices