Rupert Faltermeier

Comparison of Near-Infrared Spectroscopy and Tissue Po2 Time Series in Patients After Severe Head Injury and Aneurysmal Subarachnoid Hemorrhage

Monitoring of local oxygen pressure in brain white matter (tipo2) and of local hemoglobin oxygen saturation (rSo2) with near-infrared spectroscopy (NIRS) are increasingly employed techniques in neurosurgical intensive care units. Using frequency-based mathematical methods, the authors sought to ascertain whether both techniques contained similar information. Twelve patients treated in the intensive care unit were included (subarachnoid hemorrhage, n = 3; traumatic brain injury, n = 9). A tipo2 probe and an NIRS sensor were positioned over the frontal lobe with the most pathologic changes on initial computed tomography scan. The authors calculated coherence of tipo2 and rSo2, its overall density distribution, its distribution per data set, and its time evolution. The authors identified a band of significantly correlated frequencies (from 0 to 1.3 × 103 Hz) in more than 90% of the data sets for coherence and overall density distribution. Time evolution showed slow but marked changes of significant coherence. By means of spectral analysis the authors show that tipo2 and rSo2 signals contain similar information, albeit using completely different registration methodologies.

EMDLAB: A toolbox for analysis of single-trial {EEG} dynamics using empirical mode decomposition

BACKGROUND Empirical mode decomposition (EMD) is an empirical data decomposition technique. Recently there is growing interest in applying EMD in the biomedical field. NEW METHOD EMDLAB is an extensible plug-in for the EEGLAB toolbox, which is an open software environment for electrophysiological data analysis. RESULTS EMDLAB can be used to perform, easily and effectively, four common types of EMD: plain EMD, ensemble EMD (EEMD), weighted sliding EMD (wSEMD) and multivariate EMD (MEMD) on EEG data. In addition, EMDLAB is a user-friendly toolbox and closely implemented in the EEGLAB toolbox. COMPARISON WITH EXISTING METHODS EMDLAB gains an advantage over other open-source toolboxes by exploiting the advantageous visualization capabilities of EEGLAB for extracted intrinsic mode functions (IMFs) and Event-Related Modes (ERMs) of the signal. CONCLUSIONS EMDLAB is a reliable, efficient, and automated solution for extracting and visualizing the extracted IMFs and ERMs by EMD algorithms in EEG study.

Weighted Sliding Empirical Mode Decomposition for Online Analysis of Biomedical Time Series

Biomedical signals are in general non-linear and non-stationary. empirical mode decomposition in conjunction with a Hilbert-Huang Transform provides a fully adaptive and data-driven technique to extract intrinsic mode functions. The latter represent a complete set of locally orthogonal basis functions to represent non-linear and non-stationary time series. Large scale biomedical time series necessitate an online analysis, which is presented in this contribution. It shortly reviews the technique of EMD and related algorithms, discusses the recently proposed weighted sliding EMD algorithm (wSEMD) and, additionally, proposes a more sophisticated implementation of the weighting process. As an application to biomedical signals we will show that wSEMD in combination with mutual information could be used to detect temporal correlations of arterial blood pressure and intracranial pressure monitored at a neurosurgical intensive care unit. We will demonstrate that the wSEMD technique renders itself much more flexible than the Fourier based method used in Faltermeier et al. (Acta Neurochir Suppl, 114, 35–38, 2012).

Lack of correlation between Xenon133 and near infrared spectroscopy/indocyanine green rCBF measurements

Abstract In recent literature there are some reports describing cerebral blood flow measurements by a near infrared spectroscopy-based technique with indocyanine-green as an absorbant. To our knowledge there is no systematical study which evaluates this technique in comparison to absolute cerebral blood flow measurements. Ten patients suffering from head injury (n = 9) or subarachnoid hemorrhage (n = 1) were included. Twenty measurements of cerebral blood flow were performed, employing a Xenon133 clearance technique. Near-infrared spectroscopy measurements were performed with the Somanetics 4100 System. Indocyanine-green was given at a total dose of 0.2 mg kg-1 bodyweight intravenously. The indocyanine-green curve was compared to cerebral blood flow measurements according to rising time and area under the curve as suggested in the literature. No correlation between the indocyanine-green clearance curve and the Xenon133 cerebral blood flow measurements could be found. Neither the area under the curve (p = 0.93) nor the rising time (p = 0.75) showed a statistically significant correlation. The near-infrared spectroscopy based indocyanine-green clearance curve measurement method of cerebral blood flow seems not to give reliable results using simple mathematical models (area under the curve and rising time). In view of our findings, we have serious reservations in the potential of this technique.

Dynamic correlation between tissue PO2 and near infrared spectroscopy.

Multimodal O2 monitoring including tissue pO2 measurements and near infrared spectroscopy (NIRS) are techniques increasingly employed for monitoring patients on neurosurgical intensive care units. NIRS measures a mixed venous arterial oxygen saturation, whereas tissue pO2 evaluates the oxygen pressure in the white matter. In contrast to the tissue pO2 measurements, the NIRS at the moment has not been completely established in clinical practice. We wanted to evaluate whether both techniques are monitoring different dynamic changes. Thirteen patients were included (SAH n = 3, TBI n = 10), 12 patients were male and 1 was female. Mean age was 34 years with a range from 16-76 years. Tissue pO2 probes (Licox, GMS, Germany) were implanted in the frontal lobe showing most pathological changes on the initial CT scan. A near infrared spectroscopy sensor (Invos, Somanetics, USA) was placed simultaneously at the patients forehead. Due to the drift of the tissue pO2 probe, only data sets were taken into further account in which a tissue pO2 value above 15 mmHg was measured. 66 data sets were analyzed by calculating the spectral coherence with multi taper methods. The coherence of two independent white noise signals were defined as an observation by chance. The significance level for correlated frequencies was 90%. In the spectral long time regime (frequency > or = 0.02), more than 80% of the data sets showed a higher percentage of correlated frequencies as compared to the observation by chance. The assumption that tissue pO2 and near infrared spectroscopy probes are measuring different dynamic changes in neurosurgical intensive care patients could not be supported by our data.

Sliding Empirical Mode Decomposition

Biomedical signals are in general non-linear and non-stationary which renders them difficult to analyze with classical time series analysis techniques. Empirical Mode Decomposition (EMD) in conjunction with a Hilbert spectral transform, together called Hilbert-Huang Transform, is ideally suited to extract informative components which are characteristic of underlying biological or physiological processes. The method is fully adaptive and generates a complete set of orthogonal basis functions, called Intrinsic Mode Functions (IMFs), in a purely data-driven manner. Amplitude and frequency of IMFs may vary over time which renders them different from conventional basis systems and ideally suited to study non-linear and non-stationary time series. However, biomedical time series are often recorded over long time periods. This generates the need for efficient EMD algorithms which can analyze the data in real time. No such algorithms yet exist which are robust, efficient and easy to implement. The contribution shortly reviews the technique of EMD and related algorithms and develops an on-line variant, called Sliding Empirical Mode Decomposition (SEMD), which is shown to perform well on large scale time series.

Empirical Mode Decomposition - an introduction

Due to external stimuli, biomedical signals are in general non-linear and non-stationary. Empirical Mode Decomposition in conjunction with a Hilbert spectral transform, together called Hilbert-Huang Transform, is ideally suited to extract essential components which are characteristic of the underlying biological or physiological processes. The method is fully adaptive and generates the basis to represent the data solely from these data and based on them. The basis functions, called Intrinsic Mode Functions (IMFs) represent a complete set of locally orthogonal basis functions whose amplitude and frequency may vary over time. The contribution reviews the technique of EMD and related algorithms and discusses illustrative applications.

WEIGHTED SLIDING EMPIRICAL MODE DECOMPOSITION

The analysis of nonlinear and nonstationary time series is still a challenge, as most classical time series analysis techniques are restricted to data that is, at least, stationary. Empirical mode decomposition (EMD) in combination with a Hilbert spectral transform, together called Hilbert-Huang transform (HHT), alleviates this problem in a purely data-driven manner. EMD adaptively and locally decomposes such time series into a sum of oscillatory modes, called Intrinsic mode functions (IMF) and a nonstationary component called residuum. In this contribution, we propose an EMD-based method, called Sliding empirical mode decomposition (SEMD), which, with a reasonable computational effort, extends the application area of EMD to a true on-line analysis of time series comprising a huge amount of data if recorded with a high sampling rate. Using nonlinear and nonstationary toy data, we demonstrate the good performance of the proposed algorithm. We also show that the new method extracts component signals that fulfill all criteria of an IMF very well and that it exhibits excellent reconstruction quality. The method itself will be refined further by a weighted version, called weighted sliding empirical mode decomposition (wSEMD), which reduces the computational effort even more while preserving the reconstruction quality.

Brain status data analysis by sliding EMD

Biomedical signals are in general non-linear and non-stationary which renders them difficult to analyze with classical time series analysis techniques. Empirical Mode Decomposition (EMD) in conjunction with a Hilbert spectral transform, together called Hilbert-Huang Transform, is ideally suited to extract informative components which are characteristic of underlying biological or physiological processes. The method is fully adaptive and generates a complete set of orthogonal basis functions, called Intrinsic Mode Functions (IMFs), in a purely data-driven manner. Amplitude and frequency of IMFs may vary over time which renders them different from conventional basis systems and ideally suited to study non-linear and non-stationary time series. However, biomedical time series are often recorded over long time periods. This generates the need for efficient EMD algorithms which can analyze the data in real time. No such algorithms yet exist which are robust, efficient and easy to implement. The contribution shortly reviews the technique of EMD and related algorithms and develops an on-line variant, called slidingEMD, which is shown to perform well on large scale biomedical time series recorded during neuromonitoring.

Computerized Data Analysis of Neuromonitoring Parameters Identifies Patients with Reduced Cerebral Compliance as Seen on CT

OBJECTIVE Computer-assisted analysis of neuromonitoring parameters may provide important decision-making support to the neurointensivist. A recently developed mathematical model for the simulation of cerebral autoregulation and brain swelling showed that in the case of an intact autoregulation but diminished cerebral compliance, a negative correlation between arterial blood pressure (ABP) and intracranial pressure (ICP) occurs. The goal of our study was to verify these simulation results in an appropriate patient cohort. METHODS Simultaneously measured data (ABP, ICP) of 6 patients (1 female; 5 male) with severe head trauma (n = 5) and stroke (n = 1) were used to calculate time resolved multitaper cross coherence. Further, we calculated the Hilbert phases of both signals, defining a negative correlation in case of a mean Hilbert phase difference greater than 130°. To validate the results, CT scans performed during the critical phases identified were analyzed. RESULTS In five out of six datasets we found long lasting events of negative correlation between ABP and ICP. In all patients, corresponding CT scans demonstrated changes in the intracranial compartment characterized by diminished cerebral compliance. CONCLUSIONS Our data indicate that complex multidimensional data analysis of neuromonitoring parameters can identify complication-specific data patterns with a high degree of accuracy.