Digital analysis of the electroencephalogram, the blood pressure wave, and the electrocardiogram

The electroencephalogram (EEG), the blood pressure wave, and the electrocardiogram (ECG), produce patterns that the eye of the physician has empirically correlated with important aspects of health. Digital computer techniques for the recognition of these patterns are made particularly difficult by the realities of pattern context sensitivity, frequent signal artifact, real-time operation, finite storage limitations, and reasonable cost. Evaluations of these techniques are handicapped by the absence of absolute standards, the wide signal variability associated with pathologic states, and the sheer mechanics of comparison with human analysis. Computer analysis of the EEG has been directed toward monitoring sleep and certain pathologic states, leaving the more difficult problem of diagnosis to the trained neurologist. Automatic pattern recognition of the blood pressure wave has been implemented with straightforward techniques for diagnostic use in the cardiac catheterization laboratory and for monitoring in the intensive care unit. Computer analysis of the ECG has been directed toward morphological and rhythm diagnosis, having great potential utility in clinical heart stations, and toward rhythm monitoring, a most practical application arising in coronary intensive care units. Promising systems are emerging, but years of evaluation and adjustment will be necessary to meet the need for both accuracy and economy.