An open-source automated platform for three-dimensional visualization of subdural electrodes using CT-MRI coregistration

Dichotomous frequency-dependent phase synchrony in the sensorimotor network characterizes simplistic movement

It is hypothesized that disparate brain regions interact via synchronous activity to control behavior. The nature of these interconnected ensembles remains an area of active investigation, and particularly the role of high frequency synchronous activity in simplistic behavior is not well known. Using intracranial electroencephalography, we explored the spectral dynamics and network connectivity of sensorimotor cortical activity during a simple motor task in seven epilepsy patients. Confirming prior work, we see a "spectral tilt" (increased high-frequency (HF, 70-100 Hz) and decreased low-frequency (LF, 3-33 Hz) broadband oscillatory activity) in motor regions during movement compared to rest, as well as an increase in LF synchrony between these regions using time-resolved phase-locking. We then explored this phenomenon in high frequency and found a robust but opposite effect, where time-resolved HF broadband phase-locking significantly decreased during movement. This "connectivity tilt" (increased LF synchrony and decreased HF synchrony) displayed a graded anatomical dependency, with the most robust pattern occurring in primary sensorimotor cortical interactions and less robust pattern occurring in associative cortical interactions. Connectivity in theta (3-7 Hz) and high beta (23-27 Hz) range had the most prominent low frequency contribution during movement, with theta synchrony building gradually while high beta having the most prominent effect immediately following the cue. There was a relatively sharp, opposite transition point in both the spectral and connectivity tilt at approximately 35 Hz. These findings support the hypothesis that task-relevant high-frequency spectral activity is stochastic and that the decrease in high-frequency synchrony may facilitate enhanced low frequency phase coupling and interregional communication. Thus, the "connectivity tilt" may characterize behaviorally meaningful cortical interactions.

Network analysis of preictal iEEG reveals changes in network structure preceding seizure onset

Seizures likely result from aberrant network activity and synchronization. Changes in brain network connectivity may underlie seizure onset. We used a novel method of rapid network model estimation from intracranial electroencephalography (iEEG) data to characterize pre-ictal changes in network structure prior to seizure onset. We analyzed iEEG data from 20 patients from the iEEG.org database. Using 10 s epochs sliding by 1 s intervals, a multiple input, single output (MISO) state space model was estimated for each output channel and time point with all other channels as inputs, generating sequential directed network graphs of channel connectivity. These networks were assessed using degree and betweenness centrality. Both degree and betweenness increased at seizure onset zone (SOZ) channels 37.0 ± 2.8 s before seizure onset. Degree rose in all channels 8.2 ± 2.2 s prior to seizure onset, with increasing connections between the SOZ and surrounding channels. Interictal networks showed low and stable connectivity. A novel MISO model-based network estimation method identified changes in brain network structure just prior to seizure onset. Increased connectivity was initially isolated within the SOZ and spread to non-SOZ channels before electrographic seizure onset. Such models could help confirm localization of SOZ regions.

A framework For brain atlases: Lessons from seizure dynamics

Brain maps, or atlases, are essential tools for studying brain function and organization. The abundance of available atlases used across the neuroscience literature, however, creates an implicit challenge that may alter the hypotheses and predictions we make about neurological function and pathophysiology. Here, we demonstrate how parcellation scale, shape, anatomical coverage, and other atlas features may impact our prediction of the brain's function from its underlying structure. We show how network topology, structure-function correlation (SFC), and the power to test specific hypotheses about epilepsy pathophysiology may change as a result of atlas choice and atlas features. Through the lens of our disease system, we propose a general framework and algorithm for atlas selection. This framework aims to maximize the descriptive, explanatory, and predictive validity of an atlas. Broadly, our framework strives to provide empirical guidance to neuroscience research utilizing the various atlases published over the last century.

LeGUI: A Fast and Accurate Graphical User Interface for Automated Detection and Anatomical Localization of Intracranial Electrodes

Accurate anatomical localization of intracranial electrodes is important for identifying the seizure foci in patients with epilepsy and for interpreting effects from cognitive studies employing intracranial electroencephalography. Localization is typically performed by coregistering postimplant computed tomography (CT) with preoperative magnetic resonance imaging (MRI). Electrodes are then detected in the CT, and the corresponding brain region is identified using the MRI. Many existing software packages for electrode localization chain together separate preexisting programs or rely on command line instructions to perform the various localization steps, making them difficult to install and operate for a typical user. Further, many packages provide solutions for some, but not all, of the steps needed for confident localization. We have developed software, Locate electrodes Graphical User Interface (LeGUI), that consists of a single interface to perform all steps needed to localize both surface and depth/penetrating intracranial electrodes, including coregistration of the CT to MRI, normalization of the MRI to the Montreal Neurological Institute template, automated electrode detection for multiple types of electrodes, electrode spacing correction and projection to the brain surface, electrode labeling, and anatomical targeting. The software is written in MATLAB, core image processing is performed using the Statistical Parametric Mapping toolbox, and standalone executable binaries are available for Windows, Mac, and Linux platforms. LeGUI was tested and validated on 51 datasets from two universities. The total user and computational time required to process a single dataset was approximately 1 h. Automatic electrode detection correctly identified 4362 of 4695 surface and depth electrodes with only 71 false positives. Anatomical targeting was verified by comparing electrode locations from LeGUI to locations that were assigned by an experienced neuroanatomist. LeGUI showed a 94% match with the 482 neuroanatomist-assigned locations. LeGUI combines all the features needed for fast and accurate anatomical localization of intracranial electrodes into a single interface, making it a valuable tool for intracranial electrophysiology research.

Electrocorticography and stereo EEG provide distinct measures of brain connectivity: implications for network models

Brain network models derived from graph theory have the potential to guide functional neurosurgery, and to improve rates of post-operative seizure freedom for patients with epilepsy. A barrier to applying these models clinically is that intracranial EEG electrode implantation strategies vary by centre, region and country, from cortical grid & strip electrodes (Electrocorticography), to purely stereotactic depth electrodes (Stereo EEG), to a mixture of both. To determine whether models derived from one type of study are broadly applicable to others, we investigate the differences in brain networks mapped by electrocorticography and stereo EEG in a cohort of patients who underwent surgery for temporal lobe epilepsy and achieved a favourable outcome. We show that networks derived from electrocorticography and stereo EEG define distinct relationships between resected and spared tissue, which may be driven by sampling bias of temporal depth electrodes in patients with predominantly cortical grids. We propose a method of correcting for the effect of internodal distance that is specific to electrode type and explore how additional methods for spatially correcting for sampling bias affect network models. Ultimately, we find that smaller surgical targets tend to have lower connectivity with respect to the surrounding network, challenging notions that abnormal connectivity in the epileptogenic zone is typically high. Our findings suggest that effectively applying computational models to localize epileptic networks requires accounting for the effects of spatial sampling, particularly when analysing both electrocorticography and stereo EEG recordings in the same cohort, and that future network studies of epilepsy surgery should also account for differences in focality between resection and ablation. We propose that these findings are broadly relevant to intracranial EEG network modelling in epilepsy and an important step in translating them clinically into patient care.

Virtual resection predicts surgical outcome for drug-resistant epilepsy

Patients with drug-resistant epilepsy often require surgery to become seizure-free. While laser ablation and implantable stimulation devices have lowered the morbidity of these procedures, seizure-free rates have not dramatically improved, particularly for patients without focal lesions. This is in part because it is often unclear where to intervene in these cases. To address this clinical need, several research groups have published methods to map epileptic networks but applying them to improve patient care remains a challenge. In this study we advance clinical translation of these methods by: (i) presenting and sharing a robust pipeline to rigorously quantify the boundaries of the resection zone and determining which intracranial EEG electrodes lie within it; (ii) validating a brain network model on a retrospective cohort of 28 patients with drug-resistant epilepsy implanted with intracranial electrodes prior to surgical resection; and (iii) sharing all neuroimaging, annotated electrophysiology, and clinical metadata to facilitate future collaboration. Our network methods accurately forecast whether patients are likely to benefit from surgical intervention based on synchronizability of intracranial EEG (area under the receiver operating characteristic curve of 0.89) and provide novel information that traditional electrographic features do not. We further report that removing synchronizing brain regions is associated with improved clinical outcome, and postulate that sparing desynchronizing regions may further be beneficial. Our findings suggest that data-driven network-based methods can identify patients likely to benefit from resective or ablative therapy, and perhaps prevent invasive interventions in those unlikely to do so.

iEEGview: an open-source multifunction GUI-based Matlab toolbox for localization and visualization of human intracranial electrodes

OBJECTIVE

The precise localization of intracranial electrodes is a fundamental step relevant to the analysis of intracranial electroencephalography (iEEG) recordings in various fields. With the increasing development of iEEG studies in human neuroscience, higher requirements have been posed on the localization process, resulting in urgent demand for more integrated, easy-operation and versatile tools for electrode localization and visualization. With the aim of addressing this need, we develop an easy-to-use and multifunction toolbox called iEEGview, which can be used for the localization and visualization of human intracranial electrodes.

APPROACH

iEEGview is written in Matlab scripts and implemented with a GUI. From the GUI, by taking only pre-implant MRI and post-implant CT images as input, users can directly run the full localization pipeline including brain segmentation, image co-registration, electrode reconstruction, anatomical information identification, activation map generation and electrode projection from native brain space into common brain space for group analysis. Additionally, iEEGview implements methods for brain shift correction, visual location inspection on MRI slices and computation of certainty index in anatomical label assignment.

MAIN RESULTS

All the introduced functions of iEEGview work reliably and successfully, and are tested by images from 28 human subjects implanted with depth and/or subdural electrodes.

SIGNIFICANCE

iEEGview is the first public Matlab GUI-based software for intracranial electrode localization and visualization that holds integrated capabilities together within one pipeline. iEEGview promotes convenience and efficiency for the localization process, provides rich localization information for further analysis and offers solutions for addressing raised technical challenges. Therefore, it can serve as a useful tool in facilitating iEEG studies.

Temporal Pattern of Ripple Events in Temporal Lobe Epilepsy: Towards a Pattern-based Localization of the Seizure Onset Zone

Ripples (80-250 Hz) are brief high-frequency oscillations that are often detected in intracranial EEG (iEEG) and are currently investigated as a potential biomarker to facilitate the Iocalization of the seizure onset zone (SOZ) in patients with drug-resistant epilepsy. While the rate and shape of these oscillations have been positively correlated with the SOZ, the temporal pattern of these oscillations in the epileptic brain still requires investigation. In this study, we investigate the temporal pattern of ripple events in five patients with temporal lobe epilepsy (TLE), which is one of the most common forms of epilepsy. The rate of ripple events is positively correlated with the SOZ in TLE but its diagnostic utility in localizing the SOZ remains unclear, which suggests that additional ripple-related features should be investigated. By combining point process modeling and cluster analysis, we show that a recurrent, non-stationary bursting pattern characterizes the SOZ channels consistently across patients, while the non-SOZ channels have poor between-channel similarity and no consistent pattern over time nor across patients. Furthermore, the degree of separation between SOZ and non-SOZ model parameter vectors is significantly higher (ANOVA test, ${P}$-value $P\lt 0.01$) than the degree of separation between the ripple rates, which suggests that the temporal pattern more than the rate may contribute to the pre- surgical Iocalization of the SOZ.

Integrated analysis of anatomical and electrophysiological human intracranial data

Human intracranial electroencephalography (iEEG) recordings provide data with much greater spatiotemporal precision than is possible from data obtained using scalp EEG, magnetoencephalography (MEG), or functional MRI. Until recently, the fusion of anatomical data (MRI and computed tomography (CT) images) with electrophysiological data and their subsequent analysis have required the use of technologically and conceptually challenging combinations of software. Here, we describe a comprehensive protocol that enables complex raw human iEEG data to be converted into more readily comprehensible illustrative representations. The protocol uses an open-source toolbox for electrophysiological data analysis (FieldTrip). This allows iEEG researchers to build on a continuously growing body of scriptable and reproducible analysis methods that, over the past decade, have been developed and used by a large research community. In this protocol, we describe how to analyze complex iEEG datasets by providing an intuitive and rapid approach that can handle both neuroanatomical information and large electrophysiological datasets. We provide a worked example using an example dataset. We also explain how to automate the protocol and adjust the settings to enable analysis of iEEG datasets with other characteristics. The protocol can be implemented by a graduate student or postdoctoral fellow with minimal MATLAB experience and takes approximately an hour to execute, excluding the automated cortical surface extraction.

Segmenting the Brain Surface From CT Images With Artifacts Using Locally Oriented Appearance and Dictionary Learning

The accurate segmentation of the brain surface in post-surgical computed tomography (CT) images is critical for image-guided neurosurgical procedures in epilepsy patients. Following surgical implantation of intracranial electrodes, surgeons require accurate registration of the post-implantation CT images to the pre-implantation functional and structural magnetic resonance imaging to guide surgical resection of epileptic tissue. One way to perform the registration is via surface matching. The key challenge in this setup is the CT segmentation, where the extraction of the cortical surface is difficult due to the missing parts of the skull and artifacts introduced from the electrodes. In this paper, we present a dictionary learning-based method to segment the brain surface in post-surgical CT images of epilepsy patients following surgical implantation of electrodes. We propose learning a model of locally oriented appearance that captures both the normal tissue and the artifacts found along this brain surface boundary. Utilizing a database of clinical epilepsy imaging data to train and test our approach, we demonstrate that our method using locally oriented image appearance both more accurately extracts the brain surface and better localizes electrodes on the post-operative brain surface compared to standard, non-oriented appearance modeling. In addition, we compare our method to a standard atlas-based segmentation approach and to a U-Net-based deep convolutional neural network segmentation method.

Decision Support System for Seizure Onset Zone Localization Based on Channel Ranking and High-Frequency EEG Activity

Interictal high-frequency oscillations (HFO) are a promising biomarker that can help define the seizure onset zone (SOZ) and predict the surgical outcome after the epilepsy surgery. The utility of HFO in planning the surgery, though, is unclear. Reasons include the variability of the HFO across patients and brain regions and the influence of the sleep-wake cycle, which causes large fluctuations in the ratio between the HFO observed in SOZ and non-SOZ regions. To cope with these limitations, a rank-based solution is proposed to identify the SOZ by using the HFO in multichannel intracranial EEG. A time-varying index of the epileptic susceptibility of the different brain areas is derived from the HFO rate and a support vector machine is applied on this index to identify the SOZ. The solution is trained and tested on separate groups of patients to avoid the use of patient-specific information and provides optimal SOZ prediction using as little as 30 min of recordings per channel (window). Tested on 14 patients with various combinations of seizure type, epilepsy etiology, and SOZ arrangement (172.7 ± 90.1 h/channel per patient and 75.6 ± 23.5 channels/patient, mean ± S.D.), our solution identified the SOZ with 0.92 ± 0.03 accuracy and 0.91 ± 0.03 area under the ROC curve (mean ± S.D.) across patients. For each patient, the window onset time was varied over 72 continuous hours and the prediction of the SOZ remained insensitive to the onset time, thus showing potential for surgery planning.

Investigation of subdural electrode displacement in invasive epilepsy surgery workup using neuronavigation and intraoperative MRI

OBJECTIVES

One of the main obstacles of electrode implantation in epilepsy surgery is the electrode shift between implantation and the day of explantation. We evaluated this possible electrode displacement using intraoperative MRI (iopMRI) data and CT/MRI reconstruction.

METHODS

Thirteen patients (nine female, four male, median age 26 ± 9.4 years) suffering from drug-resistant epilepsy were examined. After implantation, the position of subdural electrodes was evaluated by 3.0 T-MRI and thin-slice CCT for 3D reconstruction. Localization of electrodes was performed with the volume-rendering technique. Post-implantation and pre-explantation 1.5 T-iopMRI scans were coregistered with the 3D reconstructions to determine the extent of electrode dislocation.

RESULTS

Intraoperative MRI at the time of explantation revealed a relevant electrode shift in one patient (8%) of 10 mm. Median electrode displacement was 1.7 ± 2.6 mm with a coregistration error of 1.9 ± 0.7 mm. The median accuracy of the neuronavigation system was 2.2 ± 0.9 mm. Six of twelve patients undergoing resective surgery were seizure free (Engel class 1A, median follow-up 37.5 ± 11.8 months).

CONCLUSION

Comparison of pre-explantation and post-implantation iopMRI scans with CT/MRI data using the volume-rendering technique resulted in an accurate placement of electrodes. In one patient with a considerable electrode dislocation, the surgical approach and extent was changed due to the detected electrode shift.

ABBREVIATIONS

ECoG: electrocorticography; EZ: epileptogenic zone; iEEG: invasive EEG; iopMRI: intraoperative MRI; MEG: magnetoencephalography; PET: positron emission tomography; SPECT: single photon emission computed tomography; 3D: three-dimensional.

A novel HFO-based method for unsupervised localization of the seizure onset zone in drug-resistant epilepsy

High frequency oscillations (HFOs) are potential biomarkers of epileptic areas. In patients with drug-resistant epilepsy, HFO rates tend to be higher in the seizure onset zone (SOZ) than in other brain regions and the resection of HFO-generating areas positively correlates with seizure-free surgery outcome. Nonetheless, the development of robust unsupervised HFO-based tools for SOZ localization remains challenging. Current approaches predict the SOZ by processing small samples of intracranial EEG (iEEG) data and applying patient-specific thresholds on the HFO rate. The HFO rate, though, varies largely over time with the patient's conditions (e.g., sleep versus wakefulness) and across patients. We propose a novel localization method for SOZ that uses a time-varying, HFO-based index to estimate the epileptic susceptibility of the iEEG channels. The method is insensitive to the average HFO rate across channels (which is both patient- and condition-specific), tracks the channel susceptibility over time, and predicts the SOZ based on the temporal evolution of the HFO rate. Tested on a preliminary dataset of continuous multi-day multichannel interictal iEEG recordings from two epileptic patients (117±97.6 h/per patient, mean ± S.D.), the reported SOZ prediction had an average 0.70±0.18 accuracy and 0.67±0.07 area under the ROC curve (mean ± S.D.) across patients.

iElectrodes: A Comprehensive Open-Source Toolbox for Depth and Subdural Grid Electrode Localization

The localization of intracranial electrodes is a fundamental step in the analysis of invasive electroencephalography (EEG) recordings in research and clinical practice. The conclusions reached from the analysis of these recordings rely on the accuracy of electrode localization in relationship to brain anatomy. However, currently available techniques for localizing electrodes from magnetic resonance (MR) and/or computerized tomography (CT) images are time consuming and/or limited to particular electrode types or shapes. Here we present iElectrodes, an open-source toolbox that provides robust and accurate semi-automatic localization of both subdural grids and depth electrodes. Using pre- and post-implantation images, the method takes 2–3 min to localize the coordinates in each electrode array and automatically number the electrodes. The proposed pre-processing pipeline allows one to work in a normalized space and to automatically obtain anatomical labels of the localized electrodes without neuroimaging experts. We validated the method with data from 22 patients implanted with a total of 1,242 electrodes. We show that localization distances were within 0.56 mm of those achieved by experienced manual evaluators. iElectrodes provided additional advantages in terms of robustness (even with severe perioperative cerebral distortions), speed (less than half the operator time compared to expert manual localization), simplicity, utility across multiple electrode types (surface and depth electrodes) and all brain regions.

Glutamate imaging (GluCEST) lateralizes epileptic foci in nonlesional temporal lobe epilepsy

When neuroimaging reveals a brain lesion, drug-resistant epilepsy patients show better outcomes after resective surgery than do the one-third of drug-resistant epilepsy patients who have normal brain magnetic resonance imaging (MRI). We applied a glutamate imaging method, GluCEST (glutamate chemical exchange saturation transfer), to patients with nonlesional temporal lobe epilepsy based on conventional MRI. GluCEST correctly lateralized the temporal lobe seizure focus on visual and quantitative analyses in all patients. MR spectra, available for a subset of patients and controls, corroborated the GluCEST findings. Hippocampal volumes were not significantly different between hemispheres. GluCEST allowed high-resolution functional imaging of brain glutamate and has potential to identify the epileptic focus in patients previously deemed nonlesional. This method may lead to improved clinical outcomes for temporal lobe epilepsy as well as other localization-related epilepsies.

Learning intervention-induced deformations for non-rigid MR-CT registration and electrode localization in epilepsy patients

This paper describes a framework for learning a statistical model of non-rigid deformations induced by interventional procedures. We make use of this learned model to perform constrained non-rigid registration of pre-procedural and post-procedural imaging. We demonstrate results applying this framework to non-rigidly register post-surgical computed tomography (CT) brain images to pre-surgical magnetic resonance images (MRIs) of epilepsy patients who had intra-cranial electroencephalography electrodes surgically implanted. Deformations caused by this surgical procedure, imaging artifacts caused by the electrodes, and the use of multi-modal imaging data make non-rigid registration challenging. Our results show that the use of our proposed framework to constrain the non-rigid registration process results in significantly improved and more robust registration performance compared to using standard rigid and non-rigid registration methods.

Localization of Metal Electrodes in the Intact Rat Brain Using Registration of 3D Microcomputed Tomography Images to a Magnetic Resonance Histology Atlas

Simultaneous neural recordings taken from multiple areas of the rodent brain are garnering growing interest due to the insight they can provide about spatially distributed neural circuitry. The promise of such recordings has inspired great progress in methods for surgically implanting large numbers of metal electrodes into intact rodent brains. However, methods for localizing the precise location of these electrodes have remained severely lacking. Traditional histological techniques that require slicing and staining of physical brain tissue are cumbersome, and become increasingly impractical as the number of implanted electrodes increases. Here we solve these problems by describing a method that registers 3-D computerized tomography (CT) images of intact rat brains implanted with metal electrode bundles to a Magnetic Resonance Imaging Histology (MRH) Atlas. Our method allows accurate visualization of each electrode bundle's trajectory and location without removing the electrodes from the brain or surgically implanting external markers. In addition, unlike physical brain slices, once the 3D images of the electrode bundles and the MRH atlas are registered, it is possible to verify electrode placements from many angles by "re-slicing" the images along different planes of view. Further, our method can be fully automated and easily scaled to applications with large numbers of specimens. Our digital imaging approach to efficiently localizing metal electrodes offers a substantial addition to currently available methods, which, in turn, may help accelerate the rate at which insights are gleaned from rodent network neuroscience.

Data integration: Combined imaging and electrophysiology data in the cloud

There has been an increasing effort to correlate electrophysiology data with imaging in patients with refractory epilepsy over recent years. IEEG.org provides a free-access, rapidly growing archive of imaging data combined with electrophysiology data and patient metadata. It currently contains over 1200 human and animal datasets, with multiple data modalities associated with each dataset (neuroimaging, EEG, EKG, de-identified clinical and experimental data, etc.). The platform is developed around the concept that scientific data sharing requires a flexible platform that allows sharing of data from multiple file formats. IEEG.org provides high- and low-level access to the data in addition to providing an environment in which domain experts can find, visualize, and analyze data in an intuitive manner. Here, we present a summary of the current infrastructure of the platform, available datasets and goals for the near future.

Seizure localization using ictal phase-locked high gamma: A retrospective surgical outcome study

OBJECTIVE

To determine whether resection of areas with evidence of intense, synchronized neural firing during seizures is an accurate indicator of postoperative outcome.

METHODS

Channels meeting phase-locked high gamma (PLHG) criteria were identified retrospectively from intracranial EEG recordings (102 seizures, 46 implantations, 45 patients). Extent of removal of both the seizure onset zone (SOZ) and PLHG was correlated with seizure outcome, classified as good (Engel class I or II, n = 32) or poor (Engel class III or IV, n = 13).

RESULTS

Patients with good outcomes had significantly greater proportions of both SOZ and the first 4 (early) PLHG sites resected. Improved outcome classification was noted with early PLHG, as measured by the area under the receiver operating characteristic curves (PLHG 0.79, SOZ 0.68) and by odds ratios for resections including at least 75% of sites identified by each measure (PLHG 9.7 [95% CI: 2.3-41.5], SOZ 5.3 [95% CI: 1.2-23.3]). Among patients with resection of at least 75% of the SOZ, 78% (n = 30) had good outcomes, increasing to 91% when the resection also included at least 75% of early PLHG sites (n = 22).

CONCLUSIONS

This study demonstrates the localizing value of early PLHG, which is comparable to that provided by the SOZ. Incorporation of PLHG into the clinical evaluation may improve surgical efficacy and help to focus resections on the most critical areas.

Segmenting the Brain Surface from CT Images with Artifacts Using Dictionary Learning for Non-rigid MR-CT Registration

This paper presents a dictionary learning-based method to segment the brain surface in post-surgical CT images of epilepsy patients following surgical implantation of electrodes. Using the electrodes identified in the post-implantation CT, surgeons require accurate registration with pre-implantation functional and structural MR imaging to guide surgical resection of epileptic tissue. In this work, we use a surface-based registration method to align the MR and CT brain surfaces. The key challenge here is not the registration, but rather the extraction of the cortical surface from the CT image, which includes missing parts of the skull and artifacts introduced by the electrodes. To segment the brain from these images, we propose learning a model of appearance that captures both the normal tissue and the artifacts found along this brain surface boundary. Using clinical data, we demonstrate that our method both accurately extracts the brain surface and better localizes electrodes than intensity-based rigid and non-rigid registration methods.