Detection of the Sleep Spindle in the Human Electroencephalogram.
Abstract
The purpose of this study is to develop and evaluate a computer program for detection of spindles in the human EEG. The computer program combines frequency analysis and cross-correlation to detect a sleep spindle, which is defined as a rhythmic burst of EEG activity of 12-14 Hz with a minimum of 0.5 sec duration. The results are presented in detection accuracies as an event to be counted and as projected accuracies for sleep classification when the spindle detection program is applied to separation of spindle stages and non-spindle stages. Over 678 minutes of nocturnal EEG data were analyzed with a resulting overall spindle detection accuracy of 66% (11% miss error and 23% false detection error). When the spindle detection program is used to separate the sleep data into spindle stages (Stage -2) and non-spindle stages (Stages -0, -1, and REM) an impressive 92% overall accuracy is attained. (Modified author abstract)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1973
- Accession Number
- AD0768343
Entities
People
- Dominick Richard Martinelli
Organizations
- Air Force Institute of Technology