Scholarlink

<p>The steady-state visual evoked potential (SSVEP) is a neural response with excellent information transmission capability, which makes it an attractive input source for brain-computer interfaces (BCI). The SSVEP is obtained from an electroencephalogram (EEG), which is typically measured on the scalp. Ear-EEG is more convenient and comfortable for the user because the ear is hairless, but the ear-EEG signal suffers from more attenuation than the conventional EEG signal. Designing an ear-EEG SSVEP-BCI and evaluating its performance is a long and extensive process owing to the need for long-term tests with multiple subjects. We previously developed a method for predicting the performance of an ear-EEG SSVEP-BCI using a small number of trials but did not sufficiently explore its versatility and applicability. In this study, we developed an interactive BCI design and evaluation process using multiple methods and datasets. The results showed that the K-nearest neighbors and curve-fitting methods could both be used to successfully predict the BCI accuracy. In particular, we demonstrated that the BCI accuracy could be predicted even when the training and test data were independent of each other.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Interactive design and evaluation process for ear-EEG SSVEP-BCI

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月8日

查看全文
  • Sodai Kondo,
  • Hisaya Tanaka

摘要

The steady-state visual evoked potential (SSVEP) is a neural response with excellent information transmission capability, which makes it an attractive input source for brain-computer interfaces (BCI). The SSVEP is obtained from an electroencephalogram (EEG), which is typically measured on the scalp. Ear-EEG is more convenient and comfortable for the user because the ear is hairless, but the ear-EEG signal suffers from more attenuation than the conventional EEG signal. Designing an ear-EEG SSVEP-BCI and evaluating its performance is a long and extensive process owing to the need for long-term tests with multiple subjects. We previously developed a method for predicting the performance of an ear-EEG SSVEP-BCI using a small number of trials but did not sufficiently explore its versatility and applicability. In this study, we developed an interactive BCI design and evaluation process using multiple methods and datasets. The results showed that the K-nearest neighbors and curve-fitting methods could both be used to successfully predict the BCI accuracy. In particular, we demonstrated that the BCI accuracy could be predicted even when the training and test data were independent of each other.