Toward a fully integrated wireless wearable EEG-NIRS bimodal acquisition system.

OBJECTIVE: Interactions between neuronal electrical activity and regional changes in microcirculation are assumed to play a major role in physiological brain activity and the development of pathological disorders, but have been poorly elucidated to date. There is a need for advanced diagnostic tools to investigate the relationships between these two physiological processes. APPROACH: To meet these needs, a wireless wearable system has been developed, which combines a near infrared spectroscopy (NIRS) system using light emitting diodes (LEDs) as a light source and silicon photodiodes as a detector with an integrated electroencephalography (EEG) system. MAIN RESULTS: The main advantages over currently available devices are miniaturization and integration of a real-time electrical and hemodynamic activity monitor into one wearable device. For patient distributed monitoring and creating a body-area network, up to seven same devices can be connected to a single base station (PC) synchronously. Each node presents enhanced portability due to the wireless communication and highly integrated components resulting in a small, lightweight signal acquisition device. Further progress includes the individual control of LEDs output to automatically or interactively adjust emitted light to the actual local situation online, the use of silicon photodiodes with a safe low-voltage power supply, and an integrated three dimensional accelerometer for movement detection for the identification of motion artifacts. SIGNIFICANCE: The device was tested and validated using our enhanced EEG-NIRS tissue mimicking fluid phantom for sensitivity mapping. Typical somatotopic electrical evoked potential experiments were performed to verify clinical applicability.

Spatiotemporal source analysis in scalp EEG vs. intracerebral EEG and SPECT: a case study in a 2-year-old child.

OBJECTIVE: This case study aims to demonstrate that spatiotemporal spike discrimination and source analysis are effective to monitor the development of sources of epileptic activity in time and space. Therefore, they can provide clinically useful information allowing a better understanding of the pathophysiology of individual seizures with time- and space-resolved characteristics of successive epileptic states, including interictal, preictal, postictal, and ictal states. METHODS: High spatial resolution scalp EEGs (HR-EEG) were acquired from a 2-year-old girl with refractory central epilepsy and single-focus seizures as confirmed by intracerebral EEG recordings and ictal single-photon emission computed tomography (SPECT). Evaluation of HR-EEG consists of the following three global steps: (1) creation of the initial head model, (2) automatic spike and seizure detection, and finally (3) source localization. During the source localization phase, epileptic states are determined to allow state-based spike detection and localization of underlying sources for each spike. In a final cluster analysis, localization results are integrated to determine the possible sources of epileptic activity. The results were compared with the cerebral locations identified by intracerebral EEG recordings and SPECT. RESULTS: The results obtained with this approach were concordant with those of MRI, SPECT and distribution of intracerebral potentials. Dipole cluster centres found for spikes in interictal, preictal, ictal and postictal states were situated an average of 6.3mm from the intracerebral contacts with the highest voltage. Both amplitude and shape of spikes change between states. Dispersion of the dipoles was higher in the preictal state than in the postictal state. Two clusters of spikes were identified. The centres of these clusters changed position periodically during the various epileptic states. CONCLUSION: High-resolution surface EEG evaluated by an advanced algorithmic approach can be used to investigate the spatiotemporal characteristics of sources located in the epileptic focus. The results were validated by standard methods, ensuring good spatial resolution by MRI and SPECT and optimal temporal resolution by intracerebral EEG. Surface EEG can be used to identify different spike clusters and sources of the successive epileptic states. The method that was used in this study will provide physicians with a better understanding of the pathophysiological characteristics of epileptic activities. In particular, this method may be useful for more effective positioning of implantable intracerebral electrodes.

Usefulness of simultaneous EEG-NIRS recording in language studies.

One of the most challenging tasks in neuroscience in language studies, is investigation of the brain's ability to integrate and process information. This task can only be successfully addressed by applying various assessment techniques integrated into a multimodal approach. Each of these techniques has its advantages and disadvantages, but help to elucidate certain aspects of the capacity of neural networks to process information. These methods provide information about changes in electrical, hemodynamic and metabolic activities. Ideally, they should be noninvasive in order to facilitate their use particularly in children. In the present review, we describe the advantages of simultaneous electroencephalographic (EEG) acquisition with near infrared spectroscopy (NIRS) to provide a better understanding of the mechanisms involved in cerebral activation. This coregistration is also useful to avoid misleading interpretation of NIRS, notably during the various phases of sleep. Development and implementation of the various tools required and assessment strategies are also discussed.

Development of an autonomic portable single-board computer based high resolution NIRS device for microcirculation analysis.

Near Infrared Spectroscopy (NIRS) is a wellestablished non-invasive technique for measuring metabolic changes in biological tissue. In this paper we describe the design and development of an autonomic portable single board computer based high resolution NIRS device, which allows quantification of these changes. The sensor-patch consisting of 8LEDs and 2photo-detectorsprovides8 channels for each detector, offering increased depth resolution for monitoring microcirculatory activity..NIRS data is acquired with a sampling rate of about 2Hz per channel using the data acquisition board which consists of a 16 bit ADC, a LED driver and programmable gain amplifiers. The components on the data acquisition board are controlled via the Advantech's PCM-3355L SBC based on Windows XP platform. The software was created using Visual Basic 6.0 and Microsoft Visual C++ 6.0. It offers optionally a real time 'monitoring' and a static data (offline) visualization mode. The most unique feature of the system is its ability to auto-calibrate itself i.e. Adopt the intensity of the LEDs output light to different experimental conditions, e.g. local melanin content, density of the tissue, and emitter-detector distances. To validate the device various experiments have been carried out such as measurements on resting and working gastrocnemius and biceps muscle in ambulatory situations. The achieved results confirmed adequate performance and reliability of the device.

EEG-NIRS in epilepsy in children and neonates.

Coregistration of EEG-near infrared spectroscopy (NIRS) is a recent technique used to analyse changes in both electrical and local hemodynamic activities. Here, we describe some technical aspects of simultaneous EEG-NIRS signal acquisition focusing on recent EEG-NIRS sensors, notably the Electroptode(®)™. Advantages and disadvantages of simultaneous EEG-NIRS acquisition are discussed in comparison to other common techniques in epilepsy. Most important recent results are presented and discussed, notably those providing new insights into the mechanisms propelling neurons to synchronize, resulting in inter-critical spikes and different types of seizures.

Development of the human choriocapillaris.

Vasculogenesis and/or angiogenesis are thought to be the major mechanisms for new vessel formation during development. A third mechanism, haemo-vasculogenesis, has been described in which blood vessel and blood cells (haematopoiesis (expression of CD34(+)) and erythropoiesis (presence of epsilon chain of haemoglobin or Hb-epsilon(+))) differentiate from a common precursor, the haemangioblast. This review describes the mechanism(s) for development of human choroidal vascular from 6 until 22 weeks gestation (WG). Endothelial cell or EC (CD31, CD34, CD39, VEGFR-2) and angioblast (CD39, VEGFR-2) markers were present in choriocapillaris (CC) by 7 WG through 22 WG. From 6 to 8 WG, many erythroblasts (nucleated Hb-epsilon(+) RBCs) were observed in the CC layer. Erythroblasts (Hb-epsilon(+)) were also positive for CD34, CD31, and/or VEGFR-2. Proliferation of vascular cells (Ki67+), suggesting angiogenesis, was not observed until 12 WG. TEM analysis demonstrated that CC was structurally immature even at 11 WG: no basement membrane, absence of pericytes, and poorly formed lumens that were filled with filopodia. Contiguous fenestrations and significant PV-1 (protein in diaphragms of fenestrations) were not observed until 21-22 WG. Smooth muscle actin was prominent at 20 WG and the maturation of pericytes was confirmed by TEM. Therefore, the embryonic CC appears to form initially by haemo-vasculogenesis (Hb-epsilon(+)/CD31(+) cells), whereas angiogenesis (CD34(+)/Ki67(+)) appears to be the mode of intermediate and large choroidal vessel development later in the foetus. Contiguous fenestrations, mature pericytes, and EC basal lamina occur late in development, around 22 WG, which coincides with photoreceptors developing inner segments.

Local haemodynamic changes preceding interictal spikes: a simultaneous electrocorticography (ECoG) and near-infrared spectroscopy (NIRS) analysis in rats.

The mechanisms that drive neurons to synchronize in epileptic spikes are still subject to debate. In the present study, we used a combination of electrocorticography and near-infrared spectroscopy (ECoG/NIRS) to evaluate haemodynamic changes before, during and after epileptic spikes induced by administration of bicuculline methiodide (BM) onto the sensorimotor cortex in 8 adult Sprague-Dawley rats. Simultaneous ECoG/NIRS signals were recorded during an initial reference period (to measure spontaneous bioelectrical/metabolic activities) and then again 60 min after BM administration. Spikes in the ECoG were detected by an in-house program based on MatLab 7.0. The appearance times of the P1 peaks were used to determine corresponding time periods in the NIRS for further analysis. We observed a pronounced pre-spike modification in the haemodynamics, which became visible latest 5 s before the spike, achieving after some oscillations its minimum at round about the P1 appearance time. The post-spike period was characterized by an initial increase in oxyhaemoglobin (HbO) and total haemoglobin (HbT) to a maximum at about 2 s after the spike followed by a phase of declining oscillations disappearing after 10 to 15 s after the spike. We discuss the mechanisms underlying the haemodynamic and electrical changes that occur before, during and after epileptiform spikes. The haemodynamic changes observed with NIRS and occurring before the spikes constitute a haemodynamic predictor of electrical synchronization of spikes.

Haemodynamic changes during seizure-like activity in a neonate: a simultaneous AC EEG-SPIR and high-resolution DC EEG recording.

OBJECTIVE: We sought to define the interaction between neonatal epileptic discharges and the haemodynamic activities in a control situation (i.e. in the absence of cardiorespiratory perturbation or any interaction with normal, ongoing, synchronized neuronal activity). METHOD: Alternating-current electroencephalography (AC EEG), near-infrared spectroscopy (NIRS), and high-resolution direct-current (HR DC) EEG were performed in a curarized, ventilated neonate with a flat interictal EEG. The seizure-like discharges (SLD) first spike was used as a trigger for further averaging of NIRS, AC and DC EEG. Source localization was performed on the averaged spike and the averaged, negative DC shift. RESULTS: SLD were of maximal amplitude in centroparietal areas and induced a change in local haemodynamic parameters characterized by a first increase in [HHb] followed by an increase in [HbO(2)] and [HbT]. [HHb] returned to baseline at the end of the seizure and decreased thereafter. The negative DC shift started before the first spike and the increase in haemodynamic parameters. It then became positive and returned to baseline at the end of the seizure. Source localization revealed different positions for the first spike and the negative DC shift. DISCUSSION: Pure SLD in neonates might induce a negative blood oxygen level-dependent (BOLD) effect on the cortex, which occurs after the negative DC shift and which has a closer temporal relationship with the neuronal discharge than a positive BOLD effect.

Segmentation of scalp and skull in neonatal MR images using probabilistic atlas and level set method.

In this paper, we present a novel automatic algorithm for scalp and skull segmentation in T1-weighted neonatal head MR images. First, the probabilistic scalp and skull atlases are constructed. Second, the scalp outer surface is extracted based on an active mesh method. Third, maximum number of boundary points corresponding to the scalp inner surface is extracted using the constructed scalp probabilistic atlas and a set of knowledge based rules. In the next step, the skull inner surface and maximum number of boundary points of the outer surface are extracted using a priori information of the head anatomy and the constructed skull probabilistic atlas. Finally, the fast sweeping, tagging and level set methods are applied to reconstruct surfaces from the detected points in three-dimensional space. The results of the new segmentation algorithm on MRI data acquired from nine newborns (including three atlas and six test subjects) were compared with manual segmented data provided by an expert radiologist. The average similarity indices for the scalp and skull segmented regions were equal to 89% and 71% for the atlas and 84% and 63% for the test data, respectively.

Neonatal probabilistic models for brain, CSF and skull using T1-MRI data: preliminary results.

Inappropriate results may be produced if one uses adult or pediatric atlases for evaluation of neonatal cerebral images for morphological studies. This is mainly due to anatomical particularities typical for this early stage of development. In this paper, we describe the construction of a digital neonatal brain atlas from a set of images of neonates aged between 39 and 42 weeks. It consists of probabilistic models for brain, cerebrospinal fluid (CSF) and skull. In the first step, the selected images are segmented automatically followed by manual correction. In the second step, the images are normalized to a stereotaxic space defined by the neonatal brain atlas template GRAMFC_T(39-42) using a popular normalization algorithm implemented in Statistical Parametric Mapping (SPM). The normalization parameters of individual subjects are then used to resample the corresponding brain, CSF and skull. Finally, to construct the probabilistic models, the average is computed for each voxel location. The atlas might be used for different applications such as source localization or neonatal structural image analysis.

Design of a digital phantom of the neonatal brain.

In this paper, we present the design and implementation of a 3D digital phantom of the neonatal brain. Commonly used digital brain phantoms (e.g. BrainWeb) are based on adults' brains. With the increasing interest in computer aided analysis of neonatal Magnetic Resonance (MR) images, it becomes necessary to create a special digital phantom for neonatal brains. This is because of the pronounced differences not only in size but more important in geometrical proportions of different brain tissues in adults and neonates and the additional need to subdivide the white matter of neonatal brains into two different types. Thus, the here created neonatal brain phantom consists of 6 different tissue types: scalp, skull, gray matter, myelinated and non-myelinated white matter and cerebrospinal fluid. Every voxel has a vector consisting of 6 probabilities of being part of one of these six tissues. The digital brain phantom will be used for simulation of tomographic images of the newborns' head and may serve as well as an evaluation data set for comparison of analysis methods for neonatal MR images, e.g. segmentation/registration algorithms, providing the possibility of controlled degradation of image data.

Coupled oxygenation oscillation measured by NIRS and intermittent cerebral activation on EEG in premature infants.

Electroencephalography of premature neonates shows a physiological discontinuity of electrical activity during quiet sleep. Near infrared spectroscopy (NIRS) shows spontaneous oscillations of hemoglobin oxygenation and volume. Similar oscillations are visible in term neonates and adults, with NIRS and other functional imaging techniques (fMRI, Doppler, etc.), but are generally thought to result from vasomotion and to be a physiological artifact of limited interest. The origin and possible relationship to neuronal activity of the baseline changes in the NIRS signal have not been established. We carried out simultaneous EEG-NIRS recordings on six healthy premature neonates and four premature neonates presenting neurological distress, to determine whether changes in the concentration of cerebral oxy- and deoxy- and total hemoglobin were related to the occurrence of spontaneous bursts of cerebral electric activity. Bursts of electroencephalographic activity in neonates during quiet sleep were found to be coupled to a transient stereotyped hemodynamic response involving a decrease in oxy-hemoglobin concentration, sometimes beginning a few seconds before the onset of electroencephalographic activity, followed by an increase, and then a return to baseline. This pattern could be either part of the baseline oscillations or superimposed changes to this baseline, influencing its shape and phase. The temporal patterns of NIRS parameters present an unique configuration, and tend to be different between our healthy and pathological subjects. Studies of physiological activities and of the effects of intrinsic regulation on the NIRS signal should increase our understanding of these patterns and EEG-NIRS studies should facilitate the integration of NIRS into the set of clinical tools used in neurology.

Automated neonatal seizure detection: a multistage classification system through feature selection based on relevance and redundancy analysis.

OBJECTIVE: Automatic seizure detection obtains valuable information concerning duration and timing of seizures. Commonly used methods for EEG seizure detection in adults are inadequate for the same task in neonates because they lack the specific age-dependant characteristics of normal and pathological EEG. This paper presents an automatic seizure detection system for newborn with focus on feature selection via relevance and redundancy analysis. METHODS: Two linear correlation-based feature selection methods and the ReliefF method were applied to parameterized EEG data acquired from six neonates aged between 39 and 42 weeks. To evaluate the effectiveness of these methods, features extracted from seizure and non-seizure segments were ranked by these methods. The optimized ranked feature subsets were fed into a backpropagation neural network for classifying. Its performance was used as indicator for the feature selection effectiveness. RESULTS: Results showed an average seizure detection rate of 91%, an average non-seizure detection rate of 95%, an average false rejection rate of 95% and an overall average detection rate of 93% with a false seizure detection rate of 1.17/h. CONCLUSIONS: This good performance in detecting newborn ictal activities has been achieved based on an optimized subset of 30 features determined by the ReliefF-based detector, which corresponds to a reduction of the number of features of up to 75%. SIGNIFICANCE: The presented approach takes into account specific characteristics of normal and pathological EEG. Thus, it can improve the accuracy of conventional seizure detection systems in newborn.

A new method to model change in cutaneous blood flow due to mechanical skin irritation part I: comparison between experimental and numerical data.

Mechanical skin irritation creates vasodilation. Vasodilation of vascular networks induces increase in blood volume and blood velocity. Both can be measured by Laser Doppler Velocimetry. We propose in this article a method permitting comparison between experimental and numerical results. Experimental data was obtained by Laser Doppler Velocimetry. Numerical results were obtained by a continuous model of the vascular network. The model consists of three layers. First and last layer are described by anisotropic and heterogeneous porous media. They represent the irrigation and the drainage of the vascular system. The intermediate layer is described by a lumped parameter model that does not permit horizontal fluxes. All vessels are compliant. The permeabilities depend on the volumes of the specific layer. Skin irritation is modeled by a change in compliance of small arterial blood vessels. The comparison between experimental and numerical data is based on the model proposed by Bonner and Nossal [1981. Model for laser Doppler measurements of blood flow. Appl. Opt. 20, 2097-2107]. The model describes the Doppler frequency spectrum S(omega) as a function of the optical phenomena creating the frequency shift. The comparison is based on the model of the first moment M1 approximately integralomegaS(omega)domega. The variables of the first moment can be determined by results of the numerical model. We have shown, that it exists a linear relation between the change in compliance and the following increase in first moment. Using this linear relation experimental and numerical data can be compared.

A new method to model change in cutaneous blood flow due to mechanical skin irritation part II: parameter identification procedure.

Mechanical skin irritation, for example a light scratch with a needle, induces histamine and neuropeptide release on the line of stroke and in the surrounding tissue. Both histamine and neuropeptides are vasodilators. They cause vasodilation by changing the contraction state of the vascular smooth muscles and hence vessel compliance. Smooth muscle contraction state is very difficult to measure in vivo. For that reason we propose in this article an identification procedure to establish an irritation law. The law gives change in vessel compliance as a function of space, time and the intensity of the stroke. We have showed that vessel compliance increases immediately after the stroke not only on the line of stroke, but also in the surrounding tissue. Then, after a short delay, vessel compliance starts decreasing in the surrounding tissue, whereas vessel compliance on the line of stroke keeps increasing. Hence, blood is transported from the surrounding tissue to the line of stroke. In this way, higher blood volume on the line of stroke can be obtained than by only changing vessel compliance locally.

Relationship between ultrastructure and biomechanical properties of the knee meniscus.

The purpose of this study was to determine the biomechanical properties of the knee meniscus and to relate them to its ultrastructure. The knee joint menisci are semicircular, fibrocartilaginous structures interposed between the femoral and tibial condyles. For a long time, they were considered to be embryologic vestiges. This study describes the response of the knee joint meniscus to circumferential, radial and axial compressive forces. The results show an anisotropic response of the knee joint meniscus to unconfined compression. The Young's modulus increased approximately twofold between vertical and circumferential or radial directions with a 10 mm/min-compression rate. This response is probably a direct consequence of the orientation of collagen fibres.

First phase microcirculatory reaction to mechanical skin irritation: a three layer model of a compliant vascular tree.

Mechanical skin irritation creates vasodilatation in the line of a stroke and in the surrounding tissue. To obtain further insight on underlying physiological mechanisms we developed a model of the vascular network comprised of three layers, where the first and the last one have a tree structure. They represent the arterial and the venous system, respectively. Both are connected via an intermediate zone representing the core of the microcirculation, which is described by means of a compliant compartment model. Irritation induces change in compliance of vessels situated at the entrance of the intermediate zone. Thus the model describes flow and pressure behavior due to mechanical skin irritation.

A three-layer continuous model of porous media to describe the first phase of skin irritation.

Mechanical skin irritation induces vasodilation on the line of scratch and in the neighboring zone. In order to model the effect of an irritation on the microcirculation, the vascular network has been described using a three-layer model. The first and last layer, considered as horizontal two-dimensional porous media, describe irrigation and drainage of the system, respectively. The intermediate layer, described by means of a lumped parameter method, does not permit horizontal fluxes. Hierarchical fluxes are directed from the first to the second layer and then towards the drainage layer in order to take into account physiological flow direction. Irritation is modeled by changing compliance of vessels situated at the entrance of the micro-circulation. The model permits to investigate the influence of change in compliance on flow and pressure behavior at microscopic and macroscopic level.

Coordination dynamics of circulatory and respiratory rhythms during psychomotor drive reduction.

A 0.15-Hz rhythm band in cutaneous blood oscillations in awake human subjects was studied in cardiovascular-respiratory time series of five subjects relaxing naïvely or practicing hypnoid relaxation (autogenic training, or AT). Time series analysis used nonlinear algorithms, time-frequency distribution (TFD), postevent scan (PES) method, and linear fast Fourier transform (FFT) algorithm. This 0.15-Hz rhythm band caused phase synchronization with respiration at 1:2, 1:1, and 2:1 integer number (n/m) ratios for extended periods. During wave epochs, the 0.15-Hz rhythm band was amplified, causing the 0.15-Hz rhythm band to also appear in interbeat intervals and arterial blood pressure fluctuations. If phase synchronization of the 0.15-Hz rhythm band with respiration was established at a 1:1 integer number ratio, it was maintained and resulted in consensualization of all cardiovascular-respiratory oscillations at this frequency. Simultaneous cardiovascular and respiratory oscillations at about 0.1 Hz did not affect the appearance of the 0.15-Hz rhythm band in the photoplethysmography (PPG) signal. Recent evidence suggests the emergence of the 0.15-Hz rhythm band and n/m phase synchronization to result from nonequilibrium phase transitions operational in the network of lower brainstem neurons and associated parasympathetic neuronal effectors. These findings corroborate our notion of the 0.15-Hz rhythm band as a marker of the trophotropic mode of operation.

Influence of age and osteoporosis on calcaneus trabecular bone structure: a preliminary in vivo MRI study by quantitative texture analysis.

Recent developments in high-resolution MR imaging techniques have opened up new perspectives for structural characterization of trabecular bone by non-invasive methods. In this study, 3-D MR imaging was performed on 17 healthy volunteers and 6 osteoporotic patients. Two different MR sequences were used to evaluate the impact on MR acquisition on texture analysis results. Images were analyzed with four automated methods of texture analysis (grey level histogram, cooccurrence, runlength and gradient matrices) enabling quantitative analysis of grey level intensity and distribution within three different regions of interest (ROI). Texture analysis is not very frequently used since the interpretation of the large number of calculated parameters is difficult. We applied multiparametric data analyses such as principal component analysis (CFA) and hierarchical ascending classification (HAC) to determine the relevant parameters to differentiate between three sets of images (healthy young volunteers, healthy postmenopaused and osteoporotic patients). The results suggest that relevant texture information (depending on the ROI localization in the calcaneus) can be extracted from calcaneus MR images to evaluate osteoporosis and age effects on trabecular bone structure if strictly the same acquisition sequences are used for all patients' examination.