A Comparison of Blink Removal Techniques in EEG Signals

• Fernando Moncada ^[1] ; González, Víctor M. ^[1] ; Beatriz García ^[2] ; Víctor Álvarez ^[1] ; Villar, José R. ^[1]
  1. [1] Universidad de Oviedo

     Universidad de Oviedo

     Oviedo, España

     
  2. [2] Complejo Asistencial Universitario de Burgos

     Complejo Asistencial Universitario de Burgos

     Burgos, España

     
• Localización: Hybrid Artificial Intelligent Systems: 16th International Conference, HAIS 2021. Bilbao, Spain. September 22–24, 2021. Proceedings / coord. por Hugo Sanjurjo González, Iker Pastor López, Pablo García Bringas, Héctor Quintián Pardo, Emilio Santiago Corchado Rodríguez, 2021, ISBN 978-3-030-86271-8, págs. 355-366
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • Blink detection and removal is a very challenging task that needs to be solved in order to perform several EEG signal analyses, especially when an online analysis is required. This study is focused on the comparison of three different techniques for blink detection and three more for blink removal; one of the techniques has been enhanced in this study by determining the dynamic threshold for each participant instead of having a common value. The experimentation first compares the blink detection and then, the best method is used for the blink removal comparison. A real data set has been gathered with healthy participants and a controlled protocol, so the eye blinks can be easily labelled. Results show that some methods performed surprisingly poor with the real data set. In terms of blink detection, the participant-tuned dynamic threshold was found better than the others in terms of Accuracy and Specificity, while comparable with the Correlation-based method. In terms of blink removal, the combined CCA+EEMD algorithm removes better the blink artifacts, but the DWT one is considerably faster than the others.