Article
Stationarity in neuro-monitoring data
Search Medline for
Authors
Published: | September 16, 2010 |
---|
Outline
Text
Objective: Neuro-monitoring plays an eminent role in the treatment of raised intracranial pressure. However, signals reflecting the metabolic and circulatory status of impairment of the injured central nervous system are normally corrupted systematically. This is because the patient is part of a therapeutic control loop and the signals acquired are rather determined by the quality of control (stationarity of signals) than by the underlying pathological process.
Methods: To verify control loop hypothesis neuro-monitoring data from twelve randomly selected SHI patients (initial GCS<8, 7 male, 5 female) were analysed for circulatory (blood pressure, intracranial pressure, cerebral perfusion pressure) and metabolic (arterial blood gases, jugular bulb oxygenation, brain tissue oxygen partial pressure) variables (n=10). A total of 120 time series of generally not equidistant sample intervals were assessed for stationarity by Wallis&Moore's runs-test. In the individual case when time-series were almost regularly spaced inspection of serial independence (autocorrelation) was also carried out.
Results: In 23 time-series non-stationarity could be proven, i.e. control loop hypothesis was violated. Trends were mainly found in cerebral perfusion pressure (CPP, n=5) and intracranial pressure (ICP, n=4). The remaining cases spread out on all but one (temperature) signal. Nine patients showed at least one time-series with trend. One patient has had clear trends in 5 of 10 variables that focussed on arterial and venous oxygenation saturation, brain tissue oxygen partial pressure, ICP, and CPP.
Conclusions: The absence of stationarity in about 20% of all times-series analysed speaks in favour of the therapeutic control loop. For analytic purposes, however, the benefit seems to be overestimated, especially when using low data sampling rates. Consequently, neuro-monitoring is the analysis of short-term disturbances that are intentionally compensated by short response time. Information content is thus reduced even if the number of sensory devices would increase.