Search

The virtual epileptic patient (VEP) is a large-scale brain modeling method based on virtual brain technology, using stereoelectroencephalography (SEEG), anatomical data (magnetic resonance imaging [MRI] and connectivity), and a computational neuronal model to provide computer simulations of a patient's seizures. VEP has potential interest in the presurgical evaluation of drug-resistant epilepsy by identifying regions most likely to generate seizures. We aimed to assess the performance of the VEP approach in estimating the epileptogenic zone and in predicting surgical outcome.VEP modeling was retrospectively applied in a cohort of 53 patients with pharmacoresistant epilepsy and available SEEG, T1-weighted MRI, and diffusion-weighted MRI. Precision recall was used to compare the regions identified as epileptogenic by VEP (EZVEP showed a precision of 64% and a recall of 44% for EZVEP is the first computational model that estimates the extent and organization of the epileptogenic zone network. It is characterized by good precision in detecting epileptogenic regions as defined by a combination of visual analysis and EI. The potential impact of VEP on improving surgical prognosis remains to be exploited. Analysis of factors limiting the performance of the actual model is crucial for its further development.

Surgical interventions in epileptic patients aimed at the removal of the epileptogenic zone have success rates at only 60-70%. This failure can be partly attributed to the insufficient spatial sampling by the implanted intracranial electrodes during the clinical evaluation, leading to an incomplete picture of spatio-temporal seizure organization in the regions that are not directly observed. Utilizing the partial observations of the seizure spreading through the brain network, complemented by the assumption that the epileptic seizures spread along the structural connections, we infer if and when are the unobserved regions recruited in the seizure. To this end we introduce a data-driven model of seizure recruitment and propagation across a weighted network, which we invert using the Bayesian inference framework. Using a leave-one-out cross-validation scheme on a cohort of 45 patients we demonstrate that the method can improve the predictions of the states of the unobserved regions compared to an empirical estimate that does not use the structural information, yet it is on the same level as the estimate that takes the structure into account. Furthermore, a comparison with the performed surgical resection and the surgery outcome indicates a link between the inferred excitable regions and the actual epileptogenic zone. The results emphasize the importance of the structural connectome in the large-scale spatio-temporal organization of epileptic seizures and introduce a novel way to integrate the patient-specific connectome and intracranial seizure recordings in a whole-brain computational model of seizure spread., Author summary The electrical activity of the brain during an epileptic seizure can be observed with intracranial EEG, that is electrodes implanted in the patient’s brain. However, due to the practical constraints only selected brain regions can be implanted, which brings a risk that the abnormal electrical activity in some non-implanted regions is hidden from the observers. In this work we introduce a method to infer what is happening in the unobserved parts based on the incomplete observations of the epileptic seizure. The method relies on the assumption that the seizure spreads along the white-matter structural connections, and finds the explanation of the whole-brain seizure spread consistent with the data. The structural connectome can be estimated from diffusion-weighted imaging for an individual patient, therefore this way the patient-specific structural connectome is utilized to better analyze the patients’ seizure recordings.

Objective In focal epilepsies, the accurate delineation of the epileptogenic network is a fundamental step before surgery. For years, the relationship between the interictal epileptic spikes (defining the “irritative zone”, IZ) and the sites of seizure initiation (SOZ) has been a matter of debate. Methods Our goal was to investigate from intracerebral recordings (stereoelectroencephalography, SEEG) the distribution of interictal epileptic spikes (based on a spike frequency index, SI) and the topography of the SOZ (based on the Epileptogenicity Index, EI) in patients having focal neocortical epilepsies. Thirty-one patients were studied. A total of 539 brain regions were quantified in term of both spike generation (SI) and seizure initiation (EI). Results We found a 56% (18/32) rate of agreement between maximal EI and maximal SI values. When considering separately patients with focal cortical dysplasia (FCD), the proportion of patients with good concordance was ∼75% (15/20), whereas it was only 33% (4/12) in the non FCD group. Conclusions Our results show that a significant part of patients have some dissociation between regions showing pronounced spiking activity and those showing high epileptogenicity. e is clinically important. Significance For patients with these dissociations, other markers than spiking frequency remain to be investigated. In the FCD group, the good concordance between SI and EI confirms that the mapping of the irritative zone is clinically important.

The brain at rest exhibits a spatio-temporally rich dynamics which adheres to systematic behaviours that persist in task paradigms but appear altered in disease. Despite this hypothesis, many rest state paradigms do not act directly upon the rest state and therefore cannot confirm hypotheses about its mechanisms. To address this challenge, we combined transcranial magnetic stimulation (TMS) and electroencephalography (EEG) to study brain’s relaxation toward rest following a transient perturbation. Specifically, TMS targeted either the medial prefrontal cortex (MPFC), i.e. part of the Default Mode Network (DMN) or the superior parietal lobule (SPL), involved in the Dorsal Attention Network. TMS was triggered by a given brain state, namely an increase in occipital alpha rhythm power. Following the initial TMS-Evoked Potential, TMS at MPFC enhances the induced occipital alpha rhythm, called Event Related Synchronisation, with a longer transient lifetime than TMS at SPL and a higher amplitude. Our findings show a strong coupling between MPFC and the occipital alpha power. Although the rest state is organized around a core of resting state networks, the DMN functionally takes a special role among these resting state networks.

International audience; Ictal MEG recordings constitute rare data. The objective of this study was to evaluate ictal magnetic source localization (MSI), using two algorithms: linearly constrained minimum variance (LCMV), a beamforming technique and equivalent current dipole (ECD). Ictal MSI was studied in six patients. Three of them were undergoing post-operative re-evaluation. For all patients, results were validated by the stereoelectroencephalographic (SEEG) definition of the epileptogenic zone (EZ). EZ was quantified using the epileptogenicity index (EI) method, which accounts for both the propensity of a brain area to generate rapid discharges and the time for this area to become involved in the seizure. EI values range from 0 (no epileptogenicity) to 1 (maximal epileptogenicity). Levels of concordance between ictal MSI and EZ were determined as follows: A: ictal MSI localized the site whose value EI = 1, B: MSI localized a part of the EZ (not corresponding to the maximal value of EI = 1), C: a region could be identified on ictal MSI but not on SEEG, D: a region could be identified on SEEG but not on MSI, E: different regions were localized on MSI and SEEG. Ictal MEG pattern consisted of rhythmic activities between 10 and 20 Hz for all patients. For LCMV (first maxima), levels of concordance were A (two cases), B (two cases) and E (two cases). For ECD fitted on each time point separately (location characterized by the best goodness-of-fit value), levels of concordance were A (one case), B (one case), D (three cases) and E (one case). For ECD calculated for the whole time window, levels of concordance were A (two cases) and D (four cases). Source localization methods performed on rhythmic patterns can localize the EZ as validated by SEEG. In terms of concordance, LCMV was superior to ECD. In some cases, LCMV allows extraction of several maxima that could reflect ictal dynamics. In a medial temporal lobe epilepsy case, ictal MSI indicated an area of delayed propagation and was non-contributory to the presurgical assessment.

TheVirtualBrain (TVB) is a neuroinformatics Python package representing theconvergence of clinical, systems, and theoretical neuroscience in the analysis,visualization and modeling of neural and neuroimaging dynamics. TVB iscomposed of a flexible simulator for neural dynamics measured across scalesfrom local populations to large-scale dynamics measured byelectroencephalography (EEG), magnetoencephalography (MEG) and functionalmagnetic resonance imaging (fMRI), and core analytic and visualizationfunctions, all accessible through a web browser user interface. A datatypesystem modeling neuroscientific data ties together these pieces with persistentdata storage, based on a combination of SQL & HDF5. These datatypes combinewith adapters allowing TVB to integrate other algorithms or computationalsystems. TVB provides infrastructure for multiple projects and multiple users,possibly participating under multiple roles. For example, a clinician mightimport patient data to identify several potential lesion points in thepatient's connectome. A modeler, working on the same project, tests thesepoints for viability through whole brain simulation, based on the patient'sconnectome, and subsequent analysis of dynamical features. TVB also drivesresearch forward: the simulator itself represents the culmination of severalsimulation frameworks in the modeling literature. The availability of thenumerical methods, set of neural mass models and forward solutions allows forthe construction of a wide range of brain-scale simulation scenarios. Thispaper briefly outlines the history and motivation for TVB, describing theframework and simulator, giving usage examples in the web UI and Pythonscripting.

TheVirtualBrain (TVB) is the first integrative neuroinformatics platform for the modeling of full brain network dynamics. TVB simulator, written in Python, enables the systematic model-based inference of neurophysiological mechanisms on different brain scales that underlie the generation of macroscopic commonly used neuroimaging signals (EEG, MEG, fMRI). In the framework of TVB we build full brain network models by incorporating biologically realistic large scale couplings of neural populations. The couplings within the network are captured by the space-time structure of the connectivity matrix, which defines the connection strengths and time delays via signal transmission between all network nodes. Researchers from computational, theoretical and clinical neuroscience can benefit of this tool. The web interface allows users without programming knowledge to access a powerful computing platform as well as visualization and analysis tools. Users with strong programming skills benefit of all the advantages provided by the Python programming through a shell interface.

Focal drug resistant epilepsy is a neurological disorder characterized by seizures caused by abnormal activity originating in one or more regions together called as epileptogenic zone. Treatment for such patients involves surgical resection of affected regions. Epileptogenic zone is typically identified using stereotactic EEG recordings from the electrodes implanted into the patient’s brain. Identifying the epileptogenic zone is a challenging problem due to the spatial sparsity of electrode implantation. We propose a probabilistic hierarchical model of seizure propagation patterns, based on a phenomenological model of seizure dynamics called Epileptor. Using Bayesian inference, the Epileptor model is optimized to build patient specific virtual models that best fit to the log power of intracranial recordings. First, accuracy of the model predictions and identifiability of the model are investigated using synthetic data. Then, model predictions are evaluated against a retrospective patient cohort of 25 patients with varying surgical outcomes. In the patients who are seizure free after surgery, model predictions showed good match with the clinical hypothesis. In patients where surgery failed to achieve seizure freedom model predictions showed a strong mismatch. Our results demonstrate that proposed probabilistic model could be a valuable tool to aid the clinicians in identifying the seizure focus., Vattikonda et al. combine a conceptual mathematical model of seizure generation with personalized information about brain connectivity to tackle the challenge of predictive brain surgery. The validated model was used to successfully determine the outcome of surgery in retrospective patient data for removal of epileptogenic brain tissue.

Despite the importance and frequent use of Bayesian frameworks in brain network modeling for parameter inference and model prediction, the advanced sampling algorithms implemented in probabilistic programming languages to overcome the inference difficulties have received relatively little attention in this context. In this technical note, we propose a probabilistic framework, namely the Bayesian Virtual Epileptic Patient (BVEP), which relies on the fusion of structural data of individuals to infer the spatial map of epileptogenicity in a personalized large-scale brain model of epilepsy spread. To invert the individualized whole-brain model employed in this study, we use the recently developed algorithms known as No-U-Turn Sampler (NUTS) as well as Automatic Differentiation Variational Inference (ADVI). Our results indicate that NUTS and ADVI accurately estimate the degree of epileptogenicity of brain regions, therefore, the hypothetical brain areas responsible for the seizure initiation and propagation, while the convergence diagnostics and posterior behavior analysis validate the reliability of the estimations. Moreover, we illustrate the efficiency of the transformed non-centered parameters in comparison to centered form of parameterization. The Bayesian framework used in this work proposes an appropriate patient-specific strategy for estimating the epileptogenicity of the brain regions to improve outcome after epilepsy surgery.