Events

Meetings 2014

Date

Speaker(s)

Affiliation

Subject

24 December

Christmas holidays

 

 

10 December

Barry Ruijter

Clinical Neurophysiology

Quantitative EEG for predicting outcome in postanoxic status epilepticus

26 November

Hil Meijer

Applied Analysis

Using single pulse electrical stimulation to assess network structure in a phenomenological model of epilepsy

12 November

Ciska Heida

Biomedical Signals and Systems

Rest and action tremor in Parkinson’s disease patients treated with DBS - related phenomena?

29 October

Nobel Lectures

Studium Generale

 

15 October

Michel van Putten & Liesbeth Wijers

Clinical Neurophysiology and Master student Biomedical Engineering

Infraslow oscillations in postanoxic encephalopathy

1 October

Stephan van Gils

Applied Analysis

Wilson and Cowan revisited after 42 years

10 September

Martine Breteler

Master student Technical Medicine

Continuous EEG monitoring for delirium detection in the ICU

July & August

Summer holidays

 

 

25 June

Willem Verwey

Cognitive Psychology and Ergonomics

TMS effects on the production of familiar keying sequences

14 May

Joost le Feber

Clinical Neurophysiology

Memory in cultured cortical networks: experiment and modeling

30 April

Kees van Dijk

Biomedical signals and systems

The high density deep brain stimulation lead: a modeling study

2 April

Jeannette Hofmeijer

Clinical Neurophysiology

Neuronal damage during mild hypoxia: synaptic failure?

Abstracts 2014

Barry Ruijter, Clinical Neurophysiology

Quantitative EEG for predicting outcome in postanoxic status epilepticus

Introduction: Electrographic status epilepticus is observed in 10-35% of patients with postanoxic encephalopathy after cardiac arrest, and is associated with poor neurological outcome. We aim to identify quantitative electrographic features that indicate either a good or poor outcome, with respect to seizure patterns and their temporal evolution.

Methods: For 46 patients with electrographic status epilepticus, we assessed continuous EEG recordings during the first 72 hours after cardiopulmonary resuscitation. For every hour, we selected five minute epochs that were assessed visually and categorized into one of seven categories. Temporal evolution of the signal was further assessed using a quantitative continuity parameter. From epochs that were classified as epileptiform, additional quantitative EEG features were extracted. These features were selected for best mimicking a human observer reading EEG. The administration of all sedative medication and antiepileptic drugs was recorded. Outcomes after three months were categorized according to the Cerebral Performance Category, dichotomized into ‘good’ (CPC 1 to 2 = no or moderate neurological disability) and ‘poor’ (CPC 3 to 5 = severe disability, coma, or death).

Results: A poor neurological outcome was observed in 36 patients. All 18 patients who developed status epilepticus from a non-continuous background pattern or during sedative treatment had a poor outcome. Patients with poor outcome had been treated more often with antiepileptic drugs, while their seizures were less likely to disappear within the first 72 hours. In terms of quantitative features, their EEGs showed epileptic discharges with a higher spatial generalization, a more regular inter-discharge interval, a higher morphologic correlation of subsequent discharges, a more suppressed background and their discharges made a higher contribution to the total signal power.

Discussion: We demonstrate that quantitative EEG features can be used to predict poor outcome in patients with postanoxic status epilepticus. Our findings suggest that EEG patterns leading to poor outcome merely reflect severe ischemic brain damage rather than a purely epileptic phenomenon.

Hil Meijer, Applied Analysis

Using single pulse electrical stimulation to assess network structure in a phenomenological model of epilepsy

We study the problem of delineating the epileptogenic zone in epilepsy surgery. Prior to surgery ECoG-data from a subdural grid is monitored for several days in order to observe an actual seizure and see how this spreads. During this monitoring time also active probing using Single Pulse Electrical Stimulation (SPES) is performed. The responses upon stimulation have been shown to relate to epileptogenic brain areas. We want to use these responses to construct a patient-specific computational model. Such a model allows computer experiments to see how resection of a node would affect the seizure rate.

Here we start with a simple phenomenological model of a 4-node network with noisy, bistable dynamics. We determine all networks that have show ictal dynamics. Next we apply stimulation to assess the network structure of these "seizure"-networks. For the reconstructed networks we show that removing a node decreases the seizure frequency for many 4-node networks. Finally, we discuss two larger networks (~20 nodes) based on SPES-data from patients from UMC Utrecht. We compare our modelling results with the clinical determined epileptogenic tissue.

Ciska Heida, Biomedical Signals and Systems

Rest and action tremor in Parkinson’s disease patients treated with DBS - related phenomena?

Rest tremor is the most common and easily recognized symptom of Parkinson’s disease (PD), but still many questions exist with regard to its origin and the neuronal pathways involved. Less well recognized in PD, but often more disabling is the occurrence of action tremor, which is any tremor that is produced during voluntary muscle contraction. In general, deep brain stimulation in the subthalamic nucleus (STN-DBS) operates with different magnitudes of clinical efficacy based on the specific motor deficit, but its effect may also be task-specific. These differential effects of DBS on PD motor symptoms are hardly explained in literature. We have explored if rest and action tremor react in a differential way to clinically effective and less effective DBS using quantitative tremor analysis methods.

Michel van Putten & Liesbeth Wijers, Clinical Neurophysiology & Master student Biomedical Engineering

Infraslow oscillations in postanoxic encephalopathy

Infraslow activity, defined as voltage fluctuations with frequencies below 0.1 Hz, represents an important component of physiological and pathological brain function.

Michel van Putten will present recent data showing that infraslow activity is preserved in ICU patients with a postanoxic encephalopathy, irrespective of neurological outcome. In ~50% of patients with "burst-suppression with identical bursts", bursts appeared in clusters and clusters of bursts were phase-locked to the infraslow oscillations. In two patients, infraslow activity was present while the EEG showed no rhythmic activity above 0.5 Hz.  The presence of infraslow activity in the absence of rhythms > 0.5 Hz lends support to a thalamic driver of these oscillations. Phase-locking of identical bursts likely reflects modulations of cortical excitability by infraslow activity.

Liesbeth Wijers will discuss simulations of burst suppression with identical bursts in a minimal computational model in the absence or presence of infraslow oscillations. The model used is a self-excitatory network that exhibits burst suppression due to short-term and long-term synaptic depression (Tabak et al., “Modeling of spontaneous activity in developing spinal cord using activity-dependent depression in an excitatory network.,” The Journal of neuroscience, vol. 20, Apr. 2000). A slow oscillation was added as external input to the population. The resulting simulations nicely mimic the phase locking of the bursts to the infraslow activity, as observed in our ICU patients.  

Stephan van Gils, Applied Analysis

Wilson and Cowan revisited after 42 years

In their famous 1972 paper, Wilson and Cowan describe the dynamics of a population of neurons. One of the key assumptions is that the mean membrane potential and the firing rate have a sigmoidal dependence. Recent experiments show that the dependence can also take the form of a Gaussian. In this lecture we highlight some of the consequences. This is joint work of the lab of Wim van Drongelen (Chicago) and the Applied Analysis group  (UT).

Martine Breteler, Master student Technical Medicine

Continuous EEG monitoring for delirium detection in the ICU

Background: Recognition and diagnosis of delirium in ICU patients is difficult due to the fluctuating character and the subjective and discontinuous nature of clinical assessment such as CAM-ICU screening. Hence, there is clear need for an objective, continuous measurement method. It is known that delirium is associated with changes in the electroencephalogram (EEG), reflected as an increase in theta and delta and reduction in alpha power. This makes EEG a candidate tool for delirium detection and monitoring in the ICU.

Objective: To determine if changes in EEG activity are indicative for delirium and whether these signals can be used for continuous delirium detection at the ICU.

Methods: In a case-control observational study, continuous EEG recording was applied (maximum 5 days). Exclusion criteria were RASS ≤-3 and neurological disorders. Patients were screened using CAM-ICU 3xday. Around each CAM-ICU score a window of 1 hour was defined within 5 minutes of artifact-free EEG was selected and divided in 30 epochs of 10 seconds. Relative power was computed for the four frequency bands delta, theta, alpha and beta. Furthermore, a subset of five EEG features, including alpha to delta (AD) ratio, theta to alpha (TA) ratio, center of gravity (COG), approximate entropy (AE) and spectral variability (SV) were calculated for each EEG epoch and combined into a learning K-Nearest Neighbor Classifier (KNNC).

Results: 21 patients, mean age of 67±7 years, were included of which 10 delirious patients and 11 non-delirious controls. Spectral analysis of the regular frequency did not show differences between the delirium and control group, however, four of the five EEG features (COG; p=0.01, SV; p=0.01, TA-ratio; p=0.02 and AD-ratio; p=0.01) used for classification revealed significant differences. KNNC achieved an accuracy of 67% and a sensitivity and specificity of 47% and 89% respectively.

Conclusions: As opposed to other studies, spectral analysis of the frequency bands revealed no significant differences between delirious and control patients. This could be due to sedatives, the heterogeneous ICU population or severity of illness for which no correction was made. Although the classifier only reached a moderate sensitivity and specificity, significant differences are seen in four EEG features to distinguish delirious and control patients. Future research is necessary to increase the performance of EEG classification; however this method could have potential for continuous delirium detection.

Willem Verwey, Cognitive Psychology and Ergonomics

TMS effects on the production of familiar keying sequences

Executing discrete movement sequences typically involves a shift with practice from a relatively slow, stimulus-based mode to a fast mode in which performance is based on retrieving and executing entire motor chunks. In the presented two experiments, we explored the involvement of the pre-supplementary motor area (pre-SMA) in discrete sequence skill by applying inhibitory 20min 1-Hz off-line repetitive transcranial magnetic stimulation (rTMS). Based on previous work, we predicted pre-SMA involvement in the selection/initiation of motor chunks, and this was confirmed by our results.

Joost le Feber, Clinical Neurophysiology

Memory in cultured cortical networks: experiment and modeling

Dissociated cortical neurons cultured on multi electrode arrays have received increasing attention to study network aspects of neuronal tissue. In the first week of culturing networks are formed. After ~1 week networks become spontaneously active and reach a mature state after ~3 weeks, with relatively stable activity patterns. Activity patterns result from a certain connectivity, and conversely, certain patterns also affect connectivity through plasticity mechanisms like e.g. spike timing dependent plasticity (STDP). Beyond three weeks, networks appear to develop an activity-connectivity balance, wherein occurring activity patterns support current connectivity. Responses to electrical stimulation usually differ from spontaneously occurring patterns and therefore disturb the activity connectivity balance.

In 23 cultures, we showed that: 1) without external input, functional connectivity was stable at time scales of multiple hours, 2) functional connectivity changed significantly after 10 minutes of electrical (tetanic) stimulation, 3) repeated application of the same stimulus induced much smaller or even negligible connectivity changes, and 4) stimulation at another electrode yielded large changes upon first stimulation and also smaller or no changes after succeeding stimuli. 5) Returning to the first stimulus did not induce connectivity changes larger than spontaneous fluctuations.

A model of 100 neurons1 (80% excitatory, 20% inhibitory) coupled by synapses with short term depression2 and spike timing dependent plasticity3, robustly reproduced the findings that networks develop an activity-connectivity balance and that a first external stimulus induced large connectivity changes, but subsequent stimuli did not. This also applied for a second (different) stimulus, provided that synaptic strengths did not reach extreme values (maximum strength or zero). Return to the first stimulus did not induce changes larger than spontaneous fluctuations.

We concluded that cortical networks memorize inputs. Probably, an external input drives the network out of the existing balance. A new balance develops, now including the response pattern to that stimulus. Consequently, following inputs on the same electrode had no effect on connectivity. A similar pattern occurred upon stimulation at another electrode. Returning to a previously applied stimulus did not change network connectivity, indicating that memory traces exist in parallel. Computational modeling suggests STDP as a crucial factor for this type of memory.

Kees van Dijk, Biomedical Signals and Systems

The high density deep brain stimulation lead: a modeling study

The positive effects of deep brain stimulation of the subthalamic nucleus (STN-DBS) as a treatment for Parkinson’s disease are highly dependent on the location of the DBS lead within the STN. A new high density (HD) lead design, in the process of development by Sapiens Steering Brain Stimulation B.V., provides extra freedom in steering the direction of the stimulating electric field. The objective of this study is to compare the performances of this novel HD lead and the conventional Medtronic lead (model 3389). A computational model, consisting of a finite element method electric field model combined with a multi-compartment neuron models of three clinically relevant neural populations in the subthalamic region, is used to evaluate on the one hand ring-mode and steering-mode stimulation with the HD lead and on the other hand monopolar stimulation with the Medtronic lead.