Efficiency of independent component analysis for equivalent current dipole localization

YAMANOİ T., Tanaka Y., Toyoshima H., Ohnishi S., Yamazaki T.

5th International Workshop on Advanced Computational Intelligence and Intelligent Informatics, IWACIII 2017, Beijing, China, 2 - 05 November 2017, (Full Text) identifier

  • Nəşrin Növü: Conference Paper / Full Text
  • Çap olunduğu şəhər: Beijing
  • Ölkə: China
  • Açar sözlər: Brain Activity, Electroencephalogram, Equivalent Current Dipole Localization Method, Independent Component Analysis
  • Açıq Arxiv Kolleksiyası: Konfrans Materialı
  • Adres: Bəli

Qısa məlumat

Recently, independent component analysis (ICA) has been used especially for EEG analysis. The ICA is a technical method for solving the so-called "cocktail party problem". The present authors applied ICA to single-trial EEGs for preprocessing to verify its efficiency for the EEG analysis thereafter. The authors measured electroencephalograms (EEGs) from subjects when recalling several types of images. Each image presented consisted of four types of line drawings of body parts. The electrodes were fixed on the scalp of subjects during the experiment. However, recorded EEGs had multiple components, including muscle and brain potentials. We thought the ICA is effective to remove these noises. In this paper, the authors attempted to estimate spatiotemporal brain activity using equivalent current dipole source localization (ECDL). First, the authors applied the ICA as preprocessing to be measured EEGs. Then we applied the ECDL method to processed EEGs to estimate spatiotemporal brain activity using the ECDL and three-dipole model. Last, we compared with these results and our previous results using ECDL analysis on Event Related Potentials (ERPs) without using ICA.