Computers in Biology and Medicine · 2021 · 19 citations · 54 references
EngineeringNeurosciencePhysiologyBiomedical Signal ProcessingNeurophysiologyTime Series AnalysisBlood Pressure Signal AnalysisCerebral Blood FlowCerebral PhysiologyElectroencephalogram AnalysisIntracranial Pressure MonitoringTime-frequency AnalysisCausalityCerebral Blood Flow RegulationEmpirical Mode DecompositionGranger Causality AnalysisIncreased Intracranial PressureIntracranial HemodynamicsIntracranial Pressure DynamicsNeural Time Series Analysis
PURPOSE: Slow vasogenic waves in arterial blood pressure (ABP), intracranial pressure (ICP) and cerebral blood flow velocity (FV) carry information on multiple brain homeostatic control mechanisms. This work presents an approach to evaluate causal relation between oscillatory modes of these signals as an alternative to time or frequency domain Granger analysis. METHODS: Forty-five patients with simultaneous recordings of ICP, ABP and FV during CSF infusion studies were examined retrospectively. Each time series was decomposed into ten intrinsic mode functions (IMFs) via Ensemble Empirical Mode Decomposition (EEMD) and, afterwards, Granger causality (GC) was computed. RESULTS: of each time series, covering a frequency range between 0.013 and 0.155 Hz. Most significant connections were from FV to ICP, being stronger during elevation of mean ICP during infusion study. No G-causality was found between any of the IMFs during the baseline phase. CONCLUSION: Nonlinearity and nonstationarity of the cerebral and systemic signals can be addressed using EEMD decomposition There is a causal influence of slow waves of FV on slow waves on ICP during the plateau phase of the infusion study for a frequency band between 0.095 and 0.155 Hz. This relationship is magnified during mild intracranial hypertension.
54
Norden E. Huang, Steven Long, Man‐Li C. Wu et al. · Proceedings of the Royal Society A Mathematical Physical and Engineering Sciences · 1998 · 23.3K citations · Full text
Computing, Computational Engineering, Biomedical Signal Processing +16
TESTING FOR SERIAL CORRELATION IN LEAST SQUARES REGRESSION. II
J. Durbin, G. S. Watson · Biometrika · 1951 · 3.6K citations