Frost J D
Int J Biomed Comput. 1979 Sep;10(5):357-73. doi: 10.1016/0020-7101(79)90051-5.
A microprocessor-based system for the detection and quantification of sharp EEG waveforms is described. The hierarchical approach utilises initial transient detection based on computation of a second-derivative measure of curvature, followed by pattern-recognition and artifact-rejection routines based on consideration of specific waveform parameters. Initials results demonstrate an ability to approximate human-analysis results, while providing precise measures of amplitude, duration, and sharpness. The decreasing cost of microprocessors makes multichannel configurations economically feasible.
本文描述了一种基于微处理器的系统，用于检测和量化尖锐的脑电图波形。该分层方法利用基于曲率二阶导数测量的初始瞬态检测，随后基于特定波形参数的考虑进行模式识别和伪迹去除程序。初步结果表明，该系统能够近似人类分析结果，同时提供幅度、持续时间和尖锐度的精确测量。微处理器成本的降低使得多通道配置在经济上可行。
