Determination of 1 DOF Elbow Angles by using EEG Signals

LÖK Ş. İ., ÖZKURT A.

Innovations in Intelligent Systems and Applications Conference (ASYU), Adana, Türkiye, 4 - 06 Ekim 2018, ss.19-23 identifier identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası:
  • Doi Numarası: 10.1109/asyu.2018.8554015
  • Basıldığı Şehir: Adana
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.19-23
  • Anahtar Kelimeler: EEG, Elbow angle determination, artificial neural networks, SINGLE-TRIAL EEG, CLASSIFICATION
  • Dokuz Eylül Üniversitesi Adresli: Evet

Özet

In this study, it was aimed to determine elbow movements and elbow angles with changes in the EEG signal by thinking voluntarily or involuntaryly. In this study, EEG signals generated during 4 different arm movements of the subjects were taken and the EEG data of these 4 cases were discriminated from each other. For this purpose, a common spatial pattern (CSP) and power spectral density (PSD) methods are used to extract the feature, and artificial neural networks (ANN) are used for classification. The average success rate was 85.65% in the first person and 79.93% in the second person under test. In future studies, real-time control of robotic arms or similar prosthetic devices with various arm movements will be possible to implement by using this method.