| pubmed-article:8982975 | pubmed:abstractText | This paper describes a complete real-time system for EEG signal analysis. Specific software and hardware have been designed to provide biologists with an efficient tool, which allows a complete study of the different states of vigilance as well as the paroxysmal activities. The analysis method which is based on the wavelet transform is first presented and compared to the standard spectral approach. The dedicated digital signal processor card, based on the Motorola 96002 processor chip, that has been designed to support real-time acquisition and real-time processing of EEG signals is then presented. We finally illustrate the proposed method by processing real EEG signals of rats, and show that it opens up new prospects in the domain of EEG-based diagnosis. We propose a new representation, called globalization, that provides a global view and better detection of paroxysmal activities. | lld:pubmed |
