Meetings 2015

Date

Speaker(s)

Affiliation

Subject

30 December

Christmas holidays

16 December

Marly van Assen

Master student Technical Medicine, AMC Amsterdam

Quantitative analysis of EEG reactivity in comatose patients

2 December

Jan Groeneveld and Jules Heuberger

Centre for Human Drug Research, Leiden

Research at CHDR, Leiden: biomarker and method development to aid proof of pharmacology studies in healthy subjects and patients

18 November

Gerco Hassink

Biomedical Signals and Systems

The effect of exogenous human α-synuclein on neuronal network functionality

4 November

Sergei Fedorovich

Institute of Biophysics and Cell Engineering NASB, Minsk

Neuronal presynaptic endings in health and disease

21 October

Lisanne Jansen Holleboom

Master student Technical Medicine, SEIN Heemstede

Automatic seizure detection on the EMU with help of EEG/ECG

23 September

Gert Kwakkel

Department of Neurorehabilitation, VU University Medical Centre

Understanding motor recovery of the upper paretic limb post stroke: Lessons from the EXPLICIT-stroke program

22 July

Laura van de Weerd

Master student Science Education and Communication

Gap junction coupling in a neural field model

July & August

Summer holidays

17 June

Else Tolner

Department of Human Genetics, Leiden University Medical Centre

Migraine and epilepsy: two sides of the same coin?

27 May

Irina Stoyanova

Clinical Neurophysiology

Acylated ghrelin improves neuronal recovery in hypoxia-induced synaptic failure in rat dissociated cortical neurons

13 May

Cecile de Vos

Medisch Spectrum Twente

Brain correlates of pain processing studied with MEG

29 April

Huan Yang

Applied Analysis

Modeling and identification of the nociceptive pathway using psychophysical measurements

1 April

Liesbeth Wijers

Master student Biomedical Engineering

Cation catastrophe: choices and consequences in modeling single neuron depolarization

18 March

Annika de Goede

Clinical Neurophysiology

Paired pulse transcranial magnetic stimulation (TMS) in healthy participants and epilepsy patients

4 February

Anniek Heerschop

Master student Technical Medicine

Continuous electroencephalo­graphy in traumatic brain injury patients at the intensive care; an explorative, observational study

7 January

Marjolein Admiraal

Master student Technical Medicine

Computerized EEG analysis for prognosis after cardiac arrest - Revision of the cerebral recovery index

Marly van Assen, Master student Technical Medicine, AMC Amsterdam

Quantitative analysis of EEG reactivity in comatose patients

EEG reactivity has been reported as a predictor of outcome in comatose patients with post-anoxic encephalopathy. However, visual analysis of EEG reactivity has a high inter-rater variability and is time consuming. Therefore, analysis of EEG reactivity may benefit from a quantitative approach enabling automatic interpretation.

We performed a prospective cohort study in comatose ICU patients in two Dutch ICUs. Reactivity to auditory, visual and sensory stimuli was tested following a strict protocol. Visual analysis was regarded as the gold standard. Reactivity was defined as a change in amplitude or frequency in the EEG after stimulation, excluding muscle activity. Stimulus responses were scored by three independent experts (JH, MvP, MT-C), responses on which two experts agreed were used. Data was separated in an independent training and validation set. Spectral characteristics of the EEG before and after each stimulus were compared. These were determined by both parametric and non-parametric methods.

Jan Groeneveld & Jules Heuberger, Centre for Human Drug Research, Leiden

Research at CHDR, Leiden: biomarker and method development to aid proof of pharmacology studies in healthy subjects and patients

At the Centre for Human Drug Research, a clinical pharmacology research institute in Leiden, methods and biomarkers are developed with the purpose to measure pharmacological effects in humans. These methods are used in early phase clinical drug studies aimed at proof-of-pharmacology of new drugs that are being developed by the pharmaceutical and biotechnology industry. Two examples of such studies will be given: one of a recently finished study and one of an ongoing study.

The study that was recently finished was aimed at finding a measure of the turnover rate of myelin in the central nervous system of humans. Six healthy subjects were administered deuterated water (D2O) for a period of 10 weeks. Over a period of 1.5-2 years they underwent 5 lumbar punctures at different time points. One of the breakdown products of myelin, beta-galactosylceramide, was extracted from the CSF and the increase in heavier isotopomers of this molecule was followed over time using mass spectrometry. A modeling approach yielded an expected turnover of this molecule of 413 days, which is also assumed for myelin. This method may be used in future studies to determine whether drugs that are being developed for multiple sclerosis can positively influence (re-)myelination.

The study that is still ongoing is aimed at determining whether peripheral excitability studies can be used to measure acute pharmacodynamic effects of riluzole and retigabine in patients with amyotrophic lateral sclerosis (ALS). Recent findings suggest that hyperexcitability (as a pathophysiological insult) may precede the structural damage to the motor neuronal system in ALS. Interestingly riluzole, the only registered drug for the treatment of ALS, has shown to partially normalize some hyperexcitability parameters (both peripheral and cortical) in patients with ALS. Therefore, if this phenotype can be reliably measured in patients with ALS, modulation of hyperexcitability could serve as a proof-of-pharmacology biomarker to track the effect of therapeutic interventions.

Gerco Hassink, Biomedical Signals and Systems

The effect of exogenous human α-synuclein on neuronal network functionality

The formation of protein aggregates called Lewy Bodies (LB) and Lewy Neurites (LN) is the pathological hallmark of Parkinson’s disease and other synucleinopathies. The main component of these protein aggregates is α-synuclein (αS). In synucleinopathies, elevated levels of αS found in the cerebrospinal fluid are associated with the disease onset. Little is known about the mechanisms by which increased levels of αS and the appearance of αS inclusions contribute to development of synucleinopathies. Here we study the effect of exogenously added αS on in vitro cultured neuronal networks.

Sergei Fedorovich, Institute of Biophysics and Cell Engineering NASB, Minsk

Neuronal presynaptic endings in health and disease

It is shown that cognitive defect in case of some neurodegenerative disease is correlated with synapse elimination but not with neuronal loss. Several factors suggests that synapses can be ‘’weak link’ in case of several environmental and metabolic stresses. There are big ratio surface/volume, ‘’second-hand’ mitochondria, calcium influx due neurotransmitter receptors activation. However few data is available on adaptation of synapses to environmental stresses.

We use in our investigation isolated neuronal presynaptic endings termed synaptosomes. Synaptosomes can be prepared from whole brain or different parts of brain from adult animals by homogenization with following differential centrifugation. Synaptosomes remain main properties of intact neuronal presynaptic endings. They able to neurotransmitter release and synaptic vesicle recycling. We use set of fluorescent dyes which permit us to image synaptic vesicle recycling, mitochondria and plasma membrane potential, intracellular sodium and chlorine concentration, free radical formation.

Brain ischemia leads to decrease of extracellular and intracellular pH down to 5.5. We compare influence of extracellular acidification and intracellular acidification induced by amiloride on main function of synaptosomes. It was shown that both decrease pHo and pHi leads to plasma membrane depolarization. Lowering of pHo but not pHi induces oxidative stress in synaptosomes. Further, it was shown that primary source for free radical in this case is intrasynaptosomal mitochondria. Oxidative stress in presynaptic terminals in vivo in case of extracellular acidification can be reason for irreversible synaptic failure in stroke.

Ketogenic diet is used for treatment of epilepsy and Alzheimer disease on early stages. Principle of this method consist in replacing of carbohydrates by fats in everyday food. It leads to increase of ketone bodies level (for instance, beta-hydroxybutirate) in blood. Despite clinical using molecular mechanisms of ketogenic diet are still not clear. We have shown that beta-hydroxybutirate able to fuel synaptic vesicle recycling. However replacing of glucose on beta-hydroxybutirate can inhibits endocytosis. We suggest that inhibiting of endocytosis can partially explain antiseizures properties of ketogenic diet.

Lisanne Jansen Holleboom, Master student Technical Medicine, SEIN Heemstede

Automatic seizure detection on the EMU with help of EEG/ECG

On the epilepsy monitoring unit (EMU) at Stichting Epilepsie Instellingen Nederland (SEIN) Heemstede, epilepsy patients are admitted for EEG and video registration. These patients are observed 24 hours a day by nurses who staff the observation room. In some cases seizures are missed or only recognized in a later stadium of the seizure. A tool is needed to help the nurses detect seizures more accurately and more rapidly.

Many articles are published about algorithms for seizure detection, yet still no method has been implemented for online use in clinical practice. Most algorithms are focussed on an on/off alarm, where false alarms might still be a problem. We believe that an additional trend display of some kind would be of great value for the EMU setting. Therefore, our focus will be on a trend display of certain features that could help the nurses to react more rapidly to seizures.

We performed a study to look into the performance of the ‘manual’ seizure detection in the current situation to objectify the needs and to be able to look at the improvement that could be made with automated detection of seizures. We are currently starting the trend display study and have completed the current ‘manual’ performance study.

Gert Kwakkel, Department of Neurorehabilitation, VU University Medical Centre

Understanding motor recovery of the upper paretic limb post stroke: Lessons from the EXPLICIT-stroke program

The time course of stroke recovery as well as the impact of exercise therapy on the pattern of (motor) recovery is poorly understood. This lecture presents the current knowledge about the predictability of the time course of body functions and activities following a stroke.

There is growing evidence that the natural logarithmic pattern of functional recovery can be modified by early started, intensive task-oriented practice. However, the impact of practice on learning-dependent and intrinsic, spontaneous mechanisms of neurological recovery is not well understood. Several, probably interrelated mechanisms, have been identified that affect recovery after stroke. These mechanisms underlying recovery are highly interactive and operate within different, sometimes limited time-windows after stroke. In this lecture, a hypothetical phenomenological model for understanding skill reacquisition after stroke will be presented. Subsequently, the need for elucidating the longitudinal association between neurological recovery and regaining meaningful activities will be discussed in order to understand what and how patients learn when they show functional improvement post stroke. This statement will be discussed in light of measuring serial outcomes post stroke at different levels of ICF. There is growing evidence that the effects of neurorehabilitation are rather adaptive (substitution) rather than based on mechanisms of ‘true neurological repair’ (restitution). Restitution seems to be restricted to a time window of the first 8 weeks post stroke within which spontaneous neurological recovery occurs. After this critical time window, upper limb recovery is mainly the result of the most efficient optimization of the intact end-effectors to accomplish meaningful tasks.

Future studies should focus on understanding stroke recovery by studying the distinction of restitution of neurological impairments from adaptive mechanisms of motor recovery. To do that, designs are needed based on intensive serial measurements in which the longitudinal relationship between kinematics and neuroplasticity is investigated early post stroke.

Laura van de Weerd, Master student Science Education and Communication

Gap junction coupling in a neural field model

The brain activity has been studied extensively during the last years. Research has shown that many physiological phenomena are associated with particular oscillations. In case of pathologies abnormal oscillatory power comes along. For example, beta oscillations are highly increased in the motor cortex and basal ganglia of patients with Parkinson’s disease. For better understanding of these phenomena, this study uses a realistic neural field model to study the effect of the gap junction distribution on the occurrence of oscillations. A continuum neural field model, described by a system of differential equations, was implemented in MATLAB. This model was used for simulations to study the effect of the boundary conditions and the spatial distribution of the diffusion on the system. The results showed that the oscillations occur at those points were the first derivative of the diffusion is approximately 0 and the amount of diffusion rises above the diffusion threshold. The frequency of the oscillations in the realistic model does not comply with the expectation of a realistic model. Therefore, further research is necessary for more accurate and reliable results.

Else Tolner, Department of Human Genetics, Leiden University Medical Centre

Migraine and epilepsy: two sides of the same coin?

Migraine often co-occurs with epilepsy, displaying a bi-directional association. Although the neurological symptoms associated with attacks of migraine are quite distinct from those of attacks of epilepsy, in both cases hyperexcitability appears to underlie disease susceptibility. Also, in both cases attacks may be provoked by trigger factors such as bright light, hyperventilation or stress. It is intriguing that genes associated with migraine are also related to epilepsy, whereby the pathogenic mutations are often related to dysfunction of ion transport. Familial hemiplegic migraine type 1 (FHM1) mutant mice that carry a missense mutation in the Cacna1a gene, which had been previously identified in patients, display enhanced glutamatergic neuronal excitability and increased susceptibility to cortical spreading depression, the most likely mechanism of the migraine aura.

In my talk, I will show findings from functional studies on CSD and seizure activity in FHM1 migraine mice that underscore the overlap in mechanisms between migraine and epilepsy. Longitudinal in vivo DC-EEG recordings and visual evoked potential studies in freely behaving and anesthetized mice, paralleled by EEG studies migraine patients, are used to identify predictive biomarkers for attacks. Recent implementation of optogenetic technology allowed us to induce CSD, and investigate effects of cortical excitability changes in relation to migraine and epilepsy, in a non-invasive manner in freely behaving mice. Additional development of biosensors for monitoring brain K+ and pH changes in relation to CSD and seizures help to provide insight in the role of metabolic changes related to neuronal excitability and attacks.

Irina Stoyanova, Clinical Neurophysiology

Acylated ghrelin improves neuronal recovery in hypoxia-induced synaptic failure in rat dissociated cortical neurons

In the core of a brain infarct, loss of neuronal function is followed by neuronal death within minutes. In the penumbra, surrounding the infarct core, damage is not yet irreversible due to some blood supply from surrounding arteries. It is assumed that in the penumbra neurons are inactive, but structurally intact and viable. They may eventually recover, or proceed to cell death, depending on remaining levels of perfusion and timely reperfusion. Therapies to prevent collateral damage of penumbral brain tissue have a large potential to improve neurological outcome of patients with ischemic stroke, but are lacking. Therefore, we developed an in vitro model using dissociated cortical neurons to investigate neuronal dynamics during and after partial hypoxia. Briefly, we applied hypoxia (PO2 lowered from 150 to 21 mmHg) during 6h in 104 cultures. Three hours after restoration of normoxia, half of the cultures were treated with ghrelin for 24h. Cultures with and without ghrelin treatment were processed immunocytochemically for detection of the synaptic marker synaptophysin. Additionally, the fraction of the inactive synapses was evaluated by staining for the marker synapsin-1. We observed that hypoxia led to drastic decline of the number of synapses, followed by some recovery after return to normoxia, but still below the pre-hypoxic level. Ghrelin treatment significantly increased the synaptic density, as compared with controls or with the pre-hypoxic period. Furthermore, ghrelin treated cultures had much less inactive synapses than the non-treated group, and the ratio of active/inactive synapses was close to the pre-hypoxic level. In conclusion, ghrelin may improve synaptic recovery after partial hypoxia, such as present in the penumbra.

Cecile de Vos, Medisch Spectrum Twente

Brain correlates of pain processing studied with MEG

At the MEG lab of the Montreal Neurological Institute a study on the analgesic effects of distraction and attention has been conducted. Healthy subjects received painful electrical stimuli at their right thumb while they were paying attention to the stimuli or were distracted from the stimuli. With MEG (magneto encephalography) the cortical processing of the stimuli has been recorded and I am analysing the data. I will shortly introduce the MEG measurements, explain the data analysis I have done so far and present some preliminary results. It is still work in progress, so questions and suggestions are welcome.

Huang Yang, Applied Analysis

Modeling and identification of the nociceptive pathway using psychophysical measurements

Abnormalities in the nociceptive pathway can contribute to the development of chronic pain. Sensitization is an example of malfunctioning of either peripheral or central nociceptive subsystems. Using quantitative sensory testing, one can only infer sensitization, but not determine the defective subsystem. The states of the subsystems may be characterized using computational modeling together with experimental data. Here, we consider human-subject experiments with the nociceptive detection of electrocutaneous stimulation. We developed a neurophysiologically plausible model replicating experimental observations. We retained six lumped model parameters characterizing peripheral and central mechanisms. To diagnose states of nociceptive subsystems, we have addressed the problem whether and how one can obtain reliable estimates of parameters using a limited amount of data from a 10-minute experiment.

Liesbeth Wijers, Master student Biomedical Engineering

Cation catastrophe: choices and consequences in modeling single neuron depolarization

Spreading depression (SD) is a phenomenon associated with neuronal depolarization, massive re-distribution of ions and cytotoxic cell swelling. Waves of SD put enormous metabolic strain on brain tissue and are thus associated with delayed cell death and function loss. One approach to find a therapeutic target for SD prevention is to capture the necessary and sufficient processes that can mimic the observed behavior in a mathematical model.

Several ion-based single cell models with Hodgkin-Huxley dynamics have been published that can exhibit a transient depolarization, thought to be the cellular correlate of SD. However, some constraints and assumptions in these models are in our view not (bio)physically admissible. We aimed to develop such a model that is simple, has volume dynamics and also obeys the (bio)physical constraint of bulk electroneutrality. In silico experiments were conducted to assess if such a single cell can produce a transient depolarization and if diffusion of K+, Na+ and Cl− is sufficient to spread depolarization from cell to cell.

Our conclusion is that the current model is insufficient to explain cell-to-cell progression of spreading depolarization. Furthermore, once the sodium content of the system is allowed to vary, a second stable but depolarized equilibrium emerges at physiological pumping strengths. These results indicate that additional dynamics or limitations are necessary to explain transient depolarization. In addition, they signal a need for further experimental validation of widely used modeling assumptions.

Annika de Goede, Clinical Neurophysiology

Paired pulse transcranial magnetic stimulation (TMS) in healthy participants and epilepsy patients

Epilepsy is characterized by an enduring predisposition to generate epileptic seizures, due to an increased cortical excitability (CE). Although the routine electroencephalogram (EEG) can assist in identifying increased CE, its sensitivity is limited (30-55%). This creates the need for a better diagnostic tool. Once the diagnosis of epilepsy is confirmed, in most patients anti-epileptic drugs (AEDs) are prescribed. However, there is actually no method to evaluate the efficacy of medication. As transcranial magnetic stimulation (TMS) provides the opportunity to assess cortical excitability, we will investigate whether this technique can be used to improve the diagnostic process in epilepsy and to evaluate the effectiveness of prescribed AEDs. I will shortly describe the design and set-up of both clinical studies. Furthermore, I will shows results of paired pulse TMS measurements in seven healthy participants, where TMS is combined with 64-channel full band EEG and electromyography (EMG) of the abductor pollicis brevis.

Anniek Heerschop, Master student Technical Medicine

Continuous electroencephalo-graphy in traumatic brain injury patients at the intensive care; an explorative, observational study

Rationale and objective: Traumatic brain injury (TBI) is one of the leading causes of death and invalidity in young people, making it a large problem for the public health. Patients with TBI admitted to the intensive care unit (ICU) are often sedated and ventilated, which makes clinical evaluation unreliable. In addition, after the initial event, various processes may result in secondary brain injury. Secondary brain damage is detrimental but could (possibly) be prevented, though only in a very short time window. Therefore early detection of clinical deterioration and early prognostication are needed. Early prognosis could also help in determining the further treatment and in informing the family about the prospects of the patient. TBI is very heterogenic in both cause and pathology making it hard to find prognostic parameters present in all patients. During this study the use of continuous electroencephalogram (cEEG) in early prognostication and early detection of deterioration and/or improvement of clinical status in patients with moderate or severe TBI is investigated. For clinical application, quantitative analysis is necessary to reduce time-consuming visual analysis of cEEG measurements and to allow evaluation by non-trained personnel.

Methods: Within 12 hours after admission to the intensive care unit (ICU) cEEG measurements were started in 21 patients with TBI. From these cEEG measurements quantitative EEG (qEEG) were calculated. The qEEG features determined were; power, relative powers, amplitude, alpha delta ratio (ADR), power ratio index (PRI), center of gravity (COG), brain symmetry index (BSI), spectral edge frequency (SEF) and infra slow activity (ISA). These features and intracranial pressure (ICP), the Glasgow coma score (GCS), the FOUR-score and admission data including age and gender were correlated with outcome. Outcome was defined as poor for non-survivors and good for survivors.

Results: The median amplitude of the first 24 hours was significantly different between non-survivors and survivors (p<0.05). A threshold of 10µV had a sensitivity of 18% and a specificity of 100% for predicting poor outcome, a threshold of 18µV had a sensitivity of 70% and a specificity of 82% for predicting good outcome. ICP measures were negatively correlated with power but did not discriminate between poor and good outcome.

Conclusions: We showed that measuring cEEG in TBI patient within 12 hours after the occurrence of the trauma is feasible. The median of the amplitude over 24 hours was significantly different between survivors and non-survivors. Median amplitudes lower than 10µV were only found in non-survivors. Mean values for other qEEG features did not differ between non-survivors and survivors but trends within patients did give information about clinical deterioration of a patient. Single features should be combined in order to be more predictive.

Marjolein Admiraal, Master student Technical Medicine

Computerized EEG analysis for prognosis after cardiac arrest - Revision of the cerebral recovery index

Objective: Early prognostication after cardiac arrest is limited since the introduction of targeted temperature management treatment, but continuous EEG (cEEG) monitoring seems a promising tool. We aimed to improve an existing automatic outcome prediction algorithm, the cerebral recovery index (CRI) by adding a newly developed automatic detection algorithm for burst suppression with and without identical bursts and evaluating the relevance of the features in the current CRI.

Methods: For development of the automatic detection algorithm for burst suppression with and without identical bursts, three methods for burst onset detection (based on non linear energy operator and based on recurrence rate with an without embedding) were compared to manual burst onset detection in a training set of 46 patients with burst suppression after cardiac arrest. The optimal performing method was evaluated in a test set of 19 additional EEGs and implemented in the CRI. Weight factors of the other features in the CRI were optimized in a different training set of 88 patients and the optimal performing CRI was evaluated in a test set of another 88 patients.

Results: Recurrence rate without embedding resulted in 100% specificity and 67% sensitivity for detection of burst suppression with identical bursts in the test set. The CRI was optimised by adding the burst identicity score, removing Shannon entropy and alpha/delta ratio and changing weight factors for regularity and coherence in the delta band. This resulted in higher sensitivities for good (30% at 12-24h post resuscitation) but especially for poor outcome (33% at 12-24h, 36% at 24-36h and 63% at 36-48h) compared to the old CRI.

Conclusion: The revised CRI predicts especially poor outcome with a high sensitivity and specificity and can be implemented in clinical practice to aid prognostication in patients after cardiac arrest.