Single Trial Estimation of Peak Latency and Amplitude of Multiple Correlated ERP Components
| Autor: | Mojtaba Ranjbar, Mohammad Mikaeili, Anahita Khorrami Banaraki |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
medicine.diagnostic_test
business.industry Research areas Computer science Speech recognition 0206 medical engineering Biomedical Engineering Pattern recognition 02 engineering and technology General Medicine Temporal correlation Electroencephalography 020601 biomedical engineering 03 medical and health sciences 0302 clinical medicine Amplitude Event-related potential medicine Artificial intelligence Single trial Latency (engineering) business 030217 neurology & neurosurgery Communication channel |
| Zdroj: | Journal of Medical and Biological Engineering. 38:161-172 |
| ISSN: | 2199-4757 1609-0985 |
| Popis: | Event related potentials (ERPs) are conventionally extracted by averaging EEG signals over many trials but some characteristics of these signals are lost as a result. Recently single-trial ERP extraction has become one of the main research areas in neuroscience. Spatiotemporal filtering method which uses more than one channel is one of these extraction methods. A modified spatiotemporal filtering method is proposed here for single-trial estimation of correlated ERP component parameters (peak latency and peak amplitude). The error of this method in extracting the peak amplitude and peak latency of ERP components is less than 10% and changing the temporal correlation coefficient of the main ERP components, does not change the results notably. Our proposed method for peak amplitude and peak latency estimation is proved to be better than other reported methods (especially for peak amplitude estimation) and has less computational cost in comparison with other spatiotemporal methods. The ability of our method to generalize to any number of correlated ERP components is one of the key points in our work. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full text from SpringerLink