Investigation of pre-processing method to detect selective attention to auditory oddball sequences

Daisuke Mizutani, Shinichiro Kanoh

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Event-related potentials elicited by two individual oddball tone sequences presented to both right and left ears can be used to realize auditory BCI to detect selected attention. In this study, event-related potentials elicited by oddball tone sequences were examined and the methods to pre-process the data (spatial filter, frequency filter and down sampling) were investigated to improve pattern classification accuracies. As a result, it was shown that the down sampling (100 Hz) of the measured data was prominent for improving the classification accuracies (over 85%) in all of six subjects.

Original languageEnglish
Title of host publicationBMEiCON 2017 - 10th Biomedical Engineering International Conference
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1-5
Number of pages5
Volume2017-January
ISBN (Electronic)9781538608821
DOIs
Publication statusPublished - 2017 Dec 19
Event10th Biomedical Engineering International Conference, BMEiCON 2017 - Hokkaido, Japan
Duration: 2017 Aug 312017 Sep 2

Other

Other10th Biomedical Engineering International Conference, BMEiCON 2017
CountryJapan
CityHokkaido
Period17/8/3117/9/2

    Fingerprint

Keywords

  • auditory oddball paradigm
  • brain-computer interface (BCI)
  • event-related potential (ERP)
  • linear discriminant analysis (LDA)
  • pattern classification
  • selective attention

ASJC Scopus subject areas

  • Health Informatics
  • Instrumentation
  • Biomedical Engineering
  • Medicine (miscellaneous)
  • Health(social science)

Cite this

Mizutani, D., & Kanoh, S. (2017). Investigation of pre-processing method to detect selective attention to auditory oddball sequences. In BMEiCON 2017 - 10th Biomedical Engineering International Conference (Vol. 2017-January, pp. 1-5). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/BMEiCON.2017.8229142