Time series methods in the monitoring of intracranial pressure. Part 1: Problems, suggestions for a monitoring scheme and review of appropriate techniques

Abstract

Intracranial hypertension is a dangerous condition and poses a constant threat to patients suffering, for example, a severe head injury. Anticipation of this condition can, however, be difficult. This paper introduces problems that are created in patient management by the pathophysiology of intracranial hypertension and describes methods that have been devised to anticipate the condition, indicating their limitations. A scheme incorporating trend detection and prediction is then suggested for monitoring intracranial pressure (ICP). Several techniques from the fields of time series analysis and control engineering were selected as candidates for the basis of the scheme. These are described and their strengths and weaknesses discussed in the context of the ICP monitoring scheme. The application of time series methods to patient monitoring is reviewed in parallel with these descriptions to indicate the momentum of developments in this area of work. This theoretical background has enabled the direction of further work to be established. In particular it is suggested that adpptive methods are necessary to provide a satisfactory response to significant changes in the measured variable, to avoid repeated alarms and to allow useful short-term predictions of the variable to be made.