Zpracování dlouhodobých EEG záznamů

Abstract

This paper presents results of development of long-term EEG records analysis software and its application to real world coma signals. Due to the nature of coma signals, presented methods are also suitable for other types of signals (sleep and neonatal EEG etc.). Digital form of EEG signals makes it possible to use methods of quantitative EEG processing, which are not available in paper form. Quantitative EEG processing techniques include simple statistical methods, filtration, segmentation, classification, coherence analysis, principal and independent component analysis, effective visualization of data, brain mapping etc. All of those tools are developed to ease the work of clinicians.