| Zugriffsnummer | 20801 |
| Dokumenttyp | Konferenzartikel in Zeitschrift |
| Peer Review | unbekannt |
| Sprache | Englisch |
| Titel | Identification of the cardiac artifact applying ICA in simultaneously measured MEG and EEG data |
| Autor(in); Institution |
Zavala-Fernandez, H.; TU Berlin, Fachgebiet EMSP, Berlin, GERMANY
Sander-Thömmes, Tilmann; 8.2, Biosignale, PTB-Berlin
Burghoff, Martin; 8.2, Biosignale, PTB-Berlin
Orglmeister, R.; TU Berlin, Fachgebiet EMSP, Berlin, GERMANY
Trahms, Lutz; 8.2, Biosignale, PTB-Berlin
|
| Quelle/Jahr | Biomedizinische Technik: 52 (2007) |
| Availability | [CD-ROM] ; file name: 1569048481.pdf |
| ISSN | 0939-4990 |
| Verlag | Berlin: de Gruyter |
| Konferenzangaben | BMT 2007, 41. Jahrestagung der DGBMT - Deutschen Gesellschaft für Biomedizinische Technik im VDE, Aachen, 26-29, September, 2007, Germany |
| Zusammenfassung | Brain activity measurements can be significantly contaminated by an artifact due to the electrical currents in the heart muscle. This artifact caused by the QRS complex is easily visible in MEG data, in contrast to that it is difficult to identify in EEG data. Multi-modal Independent Component Analysis is defined and applied to simultaneously measured MEG and EEG data to identify the artifact. Using a procedure called single modality demixing to classify the result of the independent component analysis the QRS complex is found in EEG data and compared with a signal obtained by averaging. It is conjectured that the weak QRS signal in the EEG data is identified in the multi-modal approach due to the strength of the QRS signal in the MEG data. |
