Search

Presurgical evaluation that can precisely delineate the epileptogenic zone (EZ) is one important step for successful surgical resection treatment of refractory epilepsy patients. The noninvasive EEG-fMRI recording technique combined with general linear model (GLM) analysis is considered an important tool for estimating the EZ. However, the manual marking of interictal epileptic discharges (IEDs) needed in this analysis is challenging and time-consuming because the quality of the EEG recorded inside the scanner is greatly deteriorated compared to the usual EEG obtained outside the scanner. This is one of main impediments to the widespread use of EEG-fMRI in epilepsy. We propose a deep learning based semi-automatic IED detector that can find the candidate IEDs in the EEG recorded inside the scanner which resemble sample IEDs marked in the EEG recorded outside the scanner. The manual marking burden is greatly reduced as the expert need only edit candidate IEDs. The model is trained on data from 30 patients. Validation of IEDs detection accuracy on another 37 consecutive patients shows our method can improve the median sensitivity from 50.0% for the previously proposed template-based method to 84.2%, with false positive rate as 5 events/min. Reproducibility validation on 15 patients is applied to evaluate if our method can produce similar hemodynamic response maps compared with the manual marking ground truth results. We explore the concordance between the maximum hemodynamic response and the intracerebral EEG defined EZ and find that both methods produce similar percentage of concordance (76.9%, 10 out of 13 patients, electrode was absent in the maximum hemodynamic response in two patients). This tool will make EEG-fMRI analysis more practical for clinical usage., Highlights • A deep learning based epileptic discharge detector for EEG-fMRI is proposed. • The burden of manually marking epileptic discharges is greatly reduced. • Our method can produce similar EEG-fMRI results compared with traditional method.

SummaryObjective Intracranial electroencephalography (EEG), performed presurgically in patients with drug-resistant and difficult-to-localize focal epilepsy, samples only a small fraction of brain tissue and thus requires strong hypotheses regarding the possible localization of the epileptogenic zone. EEG/fMRI (functional magnetic resonance imaging), a noninvasive tool resulting in hemodynamic responses, could contribute to the generation of these hypotheses. This study assessed how these responses, despite their interictal origin, predict the seizure-onset zone (SOZ). Methods We retrospectively studied 37 consecutive patients who underwent stereo-EEG (SEEG) and EEG/fMRI that resulted in significant hemodynamic responses. Hemodynamic response maps were co-registered to postimplantation anatomic imaging, allowing inspection of these responses in relation to SEEG electrode's location. The area containing the most significant t-value (primary cluster) explored with an electrode was assessed for concordance with SEEG-defined SOZ. Discriminant analysis was performed to distinguish the primary clusters having a high probability of localizing the SOZ. Results Thirty-one patients had at least one study with primary cluster explored with an electrode, and 24 (77%) had at least one study with primary cluster concordant with the SOZ. Each patient could have multiple types of interictal discharge and therefore multiple studies. Among 59 studies from the 37 patients, 44 had a primary cluster explored with an electrode and 30 (68%) were concordant with the SOZ. Discriminant analysis showed that the SOZ is predictable with high confidence (>90%) if the primary cluster is highly significant and if the next significant cluster is much less significant or absent. Significance The most significant hemodynamic response to interictal discharges delineates the subset of the irritative zone that generates seizures in a high proportion of patients with difficult-to-localize focal epilepsy. EEG/fMRI generates responses that are valuable targets for electrode implantation and may reduce the need for implantation in patients in whom the most significant response satisfies the condition of our discriminant analysis.

Purpose Recent studies have applied the new magnetic resonance encephalography (MREG) sequence to the study of interictal epileptic discharges (IEDs) in the electroencephalogram (EEG) of epileptic patients. However, there are no criteria to quantitatively evaluate different processing methods, to properly use the new sequence. Methods We evaluated different processing steps of this new sequence under the common generalized linear model (GLM) framework by assessing the reliability of results. A bootstrap sampling technique was first used to generate multiple replicated data sets; a GLM with different processing steps was then applied to obtain activation maps, and the reliability of these maps was assessed. Results We applied our analysis in an event-related GLM related to IEDs. A higher reliability was achieved by using a GLM with head motion confound regressor with 24 components rather than the usual 6, with an autoregressive model of order 5 and with a canonical hemodynamic response function (HRF) rather than variable latency or patient-specific HRFs. Comparison of activation with IED field also favored the canonical HRF, consistent with the reliability analysis. Conclusion The reliability analysis helps to optimize the processing methods for this fast fMRI sequence, in a context in which we do not know the ground truth of activation areas. Magn Reson Med 78:370–382, 2017. © 2016 International Society for Magnetic Resonance in Medicine

Automatic and reliable segmentation of subcortical structures is an important but difficult task in quantitative brain image analysis. Multi-atlas based segmentation methods have attracted great interest due to their promising performance. Under the multi-atlas based segmentation framework, using deformation fields generated for registering atlas images onto a target image to be segmented, labels of the atlases are first propagated to the target image space and then fused to get the target image segmentation based on a label fusion strategy. While many label fusion strategies have been developed, most of these methods adopt predefined weighting models that are not necessarily optimal. In this study, we propose a novel local label learning strategy to estimate the target image's segmentation label using statistical machine learning techniques. In particular, we use a L1-regularized support vector machine (SVM) with a k nearest neighbor (kNN) based training sample selection strategy to learn a classifier for each of the target image voxel from its neighboring voxels in the atlases based on both image intensity and texture features. Our method has produced segmentation results consistently better than state-of-the-art label fusion methods in validation experiments on hippocampal segmentation of over 100 MR images obtained from publicly available and in-house datasets. Volumetric analysis has also demonstrated the capability of our method in detecting hippocampal volume changes due to Alzheimer's disease. Hum Brain Mapp 35:2674-2697, 2014. (c) 2013 Wiley Periodicals, Inc.

Recent studies have applied the new magnetic resonance encephalography (MREG) sequence to the study of interictal epileptic discharges (IEDs) in the electroencephalogram (EEG) of epileptic patients. However, there are no criteria to quantitatively evaluate different processing methods, to properly use the new sequence.We evaluated different processing steps of this new sequence under the common generalized linear model (GLM) framework by assessing the reliability of results. A bootstrap sampling technique was first used to generate multiple replicated data sets; a GLM with different processing steps was then applied to obtain activation maps, and the reliability of these maps was assessed.We applied our analysis in an event-related GLM related to IEDs. A higher reliability was achieved by using a GLM with head motion confound regressor with 24 components rather than the usual 6, with an autoregressive model of order 5 and with a canonical hemodynamic response function (HRF) rather than variable latency or patient-specific HRFs. Comparison of activation with IED field also favored the canonical HRF, consistent with the reliability analysis.The reliability analysis helps to optimize the processing methods for this fast fMRI sequence, in a context in which we do not know the ground truth of activation areas. Magn Reson Med 78:370-382, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

For subcortical structure segmentation, multi-atlas based segmentation methods have attracted great interest due to their competitive performance. Under this framework, using deformation fields generated for registering atlas images to the target image, labels of the atlases are first propagated to the target image space and further fused somehow to get the target segmentation. Many label fusion strategies have been proposed and most of them adopt predefined weighting models which are not necessarily optimal. In this paper, we propose a local label learning (L3) strategy to estimate the target image's label using statistical machine learning techniques. Specifically, we use Support Vector Machine (SVM) to learn a classifier for each of the target image voxels using its neighboring voxels in the atlases as a training dataset. Each training sample has dozens of image features extracted around its neighborhood and these features are optimally combined by the SVM learning method to classify the target voxel. The key contribution of this method is the development of a locally specific classifier for each target voxel based on informative texture features. The validation experiment on 57 MR images has demonstrated that our method generates segmentation results of hippocampal with a dice overlap of 0.908±0.023 to manual segmentations, statistically significantly better than state-of-the-art segmentation algorithms.