Automatic identification of intracranial pressure waveform during external ventricular drainage clamping: segmentation via wavelet analysis

Abstract

Objective: The objective of this study is to develop and validate a method for automatically identifying segments of intracranial pressure (ICP) waveform data from external ventricular drainage (EVD) recordings during intermittent drainage and closure.
Methods: The proposed method uses time-frequency analysis through wavelets to distinguish periods of ICP waveform in EVD data. By comparing the frequency compositions of the ICP signals (when the EVD system is clamped) and the artifacts (when the system is open), the algorithm can detect short, uninterrupted segments of ICP waveform from the longer periods of non-measurement data. The method involves applying a wavelet transform, calculating the absolute power in a specific range, using Otsu thresholding to automatically identify a threshold, and performing a morphological operation to remove small segments. Two investigators manually graded the same randomly selected one-hour segments of the resulting processed data. Performance metrics were calculated as a percentage. 
Results: The study analyzed data from 229 patients who had EVD placed following subarachnoid hemorrhage between June 2006 and December 2012. Of these, 155 (67.7%) were female and 62 (27%) developed delayed cerebral ischemia. A total of 45,150 hours of data were segmented. 2044 one-hour segments were randomly selected and evaluated by two investigators (MM and DN). Of those, the evaluators agreed on the classification of 1,556 one-hour segments. The algorithm was able to correctly identify 86% (1338 hours) of ICP waveform data. 8.2%(128 hours) of the time the algorithm either partially or fully failed to segment the ICP waveform. 5.4 % (84 hours) of data, artifacts were mistakenly identified as ICP waveforms (false positives). 
Conclusion: The proposed algorithm automates the identification of valid ICP waveform segments of waveform in EVD data and thus enables the inclusion in real-time data analysis for decision support. It also standardizes and makes research data management more efficient.

Deriving Automated Device Metadata From Intracranial Pressure Waveforms: A Transforming Research and Clinical Knowledge in Traumatic Brain Injury ICU Physiology Cohort Analysis

“…Figure S4 (http://links.lww.com/ CCX/B362) articulates observed differences between EVDs and IPMs in boxplots and kernel density plots for a sample of the selected features, which demonstrate greater complexity and volatility in EVDs versus consistency and stability in IPMs as key differentiators. Features selected were Higuchi fractal dimension short-time-frequency energy band (11)(12)(13)(14)(15)(16), spectral centroid, summation of power spectral density in 20-40 Hz, autocorrelation, coefficients of polynomial fit on ICP beat, Hurst exponent, eigenvalues of phase space (34) (Fig. S4, A and B, http://links.lww.com/ CCX/B362), coefficients of an autoregressive model with order 4 (Fig.…”

“…Machine learning (ML) approaches are powerful tools to further analyze and interpret ICP waveform data, which offer the promise to expand our ability to detect and predict intracranial hypertension or clinical deterioration (8)(9)(10)(11)(12). Quantitative characteristics have been used to enhance ICP signal quality and recognize nonartifactual ICP pulses (7), identify clamping of external ventricular drain (EVD)-derived ICP waveforms (13), and quantify the P1, P2, and P3 peaks within ICP pulse (which reflect the routine cycle of ICP in the brain, containing valuable continuous information on dynamic cerebrospinal pathophysiology, rather than the overall mean value) (14) (Fig. 1).…”