Pattern Recognition with Clinical Electroencephalograms.
Abstract
The digital computer has been an invaluable aid when analyzing several types of physiological and biomedical data. However, almost all attempts at analyzing electroencephalograms (EEG) with a digital computer have been thwarted by the peculiar nature and uncertain origins of this signal. No general procedures have yet been developed that handle on-going EEG data automatically and with little human intervention. This report proposes a framework that both categorizes all known computer-aided analyses of clinical EEG and suggests novel approaches to EEG analysis. This framework is based on pattern recognition methodology and leads to algorithms that select features, study data structure, and make decisions about the condition of an EEG. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 30, 1970
- Accession Number
- AD0720813
Entities
People
- Richard C. Dubes
Organizations
- Michigan State University