Surface-based multi-template automated hippocampal segmentation: Application to temporal lobe epilepsy

Connectome biomarkers of drug-resistant epilepsy

Drug-resistant epilepsy (DRE) considerably affects patient health, cognition, and well-being, and disproportionally contributes to the overall burden of epilepsy. The most common DRE syndromes are temporal lobe epilepsy related to mesiotemporal sclerosis and extratemporal epilepsy related to cortical malformations. Both syndromes have been traditionally considered as "focal," and most patients benefit from brain surgery for long-term seizure control. However, increasing evidence indicates that many DRE patients also present with widespread structural and functional network disruptions. These anomalies have been suggested to relate to cognitive impairment and prognosis, highlighting their importance for patient management. The advent of multimodal neuroimaging and formal methods to quantify complex systems has offered unprecedented ability to profile structural and functional brain networks in DRE patients. Here, we performed a systematic review on existing DRE network biomarker candidates and their contribution to three key application areas: (1) modeling of cognitive impairments, (2) localization of the surgical target, and (3) prediction of clinical and cognitive outcomes after surgery. Although network biomarkers hold promise for a range of clinical applications, translation of neuroimaging biomarkers to the patient's bedside has been challenged by a lack of clinical and prospective studies. We therefore close by highlighting conceptual and methodological strategies to improve the evaluation and accessibility of network biomarkers, and ultimately guide clinically actionable decisions.

Accuracy and practical aspects of semi- and fully automatic segmentation methods for resected brain areas

Purpose

Precise segmentation of brain lesions is essential for neurological research. Specifically, resection volume estimates can aid in the assessment of residual postoperative tissue, e.g. following surgery for glioma. Furthermore, behavioral lesion-symptom mapping in epilepsy relies on accurate delineation of surgical lesions. We sought to determine whether semi- and fully automatic segmentation methods can be applied to resected brain areas and which approach provides the most accurate and cost-efficient results.

Methods

We compared a semi-automatic (ITK-SNAP) with a fully automatic (lesion_GNB) method for segmentation of resected brain areas in terms of accuracy with manual segmentation serving as reference. Additionally, we evaluated processing times of all three methods. We used T1w, MRI-data of epilepsy patients (n = 27; 11 m; mean age 39 years, range 16–69) who underwent temporal lobe resections (17 left).

Results

The semi-automatic approach yielded superior accuracy (p < 0.001) with a median Dice similarity coefficient (mDSC) of 0.78 and a median average Hausdorff distance (maHD) of 0.44 compared with the fully automatic approach (mDSC 0.58, maHD 1.32). There was no significant difference between the median percent volume difference of the two approaches (p > 0.05). Manual segmentation required more human input (30.41 min/subject) and therefore inferring significantly higher costs than semi- (3.27 min/subject) or fully automatic approaches (labor and cost approaching zero).

Conclusion

Semi-automatic segmentation offers the most accurate results in resected brain areas with a moderate amount of human input, thus representing a viable alternative compared with manual segmentation, especially for studies with large patient cohorts.

Functional Networks in Epilepsy Presurgical Evaluation

Continuing advancements in neuroimaging methodology allow for increasingly detailed in vivo characterization of structural and functional brain networks, leading to the recognition of epilepsy as a disorder of large-scale networks. In surgical candidates, analysis of functional networks has proved invaluable for the identification of eloquent brain areas, such as hemispherical language dominance. More recently, connectome-based biomarkers have demonstrated potential to further inform clinical decision making in drug-refractory epilepsy. This article summarizes current evidence on epilepsy as a network disorder, emphasizing potential benefits of network analysis techniques for preoperative assessments and resection planning.

Recommendations of the Russian League Against Epilepsy (RLAE) on the use of magnetic resonance imaging in the diagnosis of epilepsy

Introduction. The MRI method has revolutionized the diagnosis of epilepsy. However, the widespread adoption of MRI in clinical practice is slowed by an insufficient number of high-field MRI scanners, a shortage of trained specialists, and the lack of standard examination protocols. The aim of this article is to present the Recommendations of the Russian League Against Epilepsy (RLAE) on the use of magnetic resonance imaging in the diagnosis of epilepsy.

Materials and methods. As a structural element of the International League Against Epilepsy (ILAE), the RLAE considers it important to adapt the Protocol developed by ILAE for specialists in Russia and EAEU countries. The working group analyzed and generalized the clinical practice existing in the Russian Federation, the Republic of Kazakhstan, the Republic of Belarus and the Republic of Uzbekistan. These recommendations are intended for doctors in specialized centers of epilepsy surgery, and for doctors in general medical centers. The recommendations are applicable primarily to adult patients, but the general principles are relevant to children as well.

Results. In all patients with convulsive seizures shortly after the first seizure, or patients diagnosed with epilepsy who have an unexplained increase in the frequency of seizures, rapid decrease in cognitive functions or the appearance / worsening of neuropsychiatric symptoms, the RLAE recommends using a unified MR protocol for the neuroimaging of structural sequences in epilepsy with three-dimensional pulse sequences T1 and T2 FLAIR with isotropic voxel 1 × 1 × 1 mm3 and two-dimensional T2- weighted pulse sequences with a pixel size of 1 × 1 mm2 or less. The MRI examination should be combined with EEG or EEG-video monitoring. Using this protocol allows one to set a unified standard for examining patients with epilepsy in order to detect (with high sensitivity) brain lesions playing a key role in the occurrence of seizures. Here, all 13 recommendations are presented.

Conclusion. Implementation of these recommendations in clinical practice will improve the access to high-tech medical care and optimize health care costs. 

Recommendations for the use of structural magnetic resonance imaging in the care of patients with epilepsy: A consensus report from the International League Against Epilepsy Neuroimaging Task Force

Structural magnetic resonance imaging (MRI) is of fundamental importance to the diagnosis and treatment of epilepsy, particularly when surgery is being considered. Despite previous recommendations and guidelines, practices for the use of MRI are variable worldwide and may not harness the full potential of recent technological advances for the benefit of people with epilepsy. The International League Against Epilepsy Diagnostic Methods Commission has thus charged the 2013-2017 Neuroimaging Task Force to develop a set of recommendations addressing the following questions: (1) Who should have an MRI? (2) What are the minimum requirements for an MRI epilepsy protocol? (3) How should magnetic resonance (MR) images be evaluated? (4) How to optimize lesion detection? These recommendations target clinicians in established epilepsy centers and neurologists in general/district hospitals. They endorse routine structural imaging in new onset generalized and focal epilepsy alike and describe the range of situations when detailed assessment is indicated. The Neuroimaging Task Force identified a set of sequences, with three-dimensional acquisitions at its core, the harmonized neuroimaging of epilepsy structural sequences-HARNESS-MRI protocol. As these sequences are available on most MR scanners, the HARNESS-MRI protocol is generalizable, regardless of the clinical setting and country. The Neuroimaging Task Force also endorses the use of computer-aided image postprocessing methods to provide an objective account of an individual's brain anatomy and pathology. By discussing the breadth and depth of scope of MRI, this report emphasizes the unique role of this noninvasive investigation in the care of people with epilepsy.

Hippocampal subfields segmentation in brain MR images using generative adversarial networks

Background

Segmenting the hippocampal subfields accurately from brain magnetic resonance (MR) images is a challenging task in medical image analysis. Due to the small structural size and the morphological complexity of the hippocampal subfields, the traditional segmentation methods are hard to obtain the ideal segmentation result.

Methods

In this paper, we proposed a hippocampal subfields segmentation method using generative adversarial networks. The proposed method can achieve the pixel-level classification of brain MR images by building an UG-net model and an adversarial model and training the two models against each other alternately. UG-net extracts local information and retains the interrelationship features between pixels. Moreover, the adversarial training implements spatial consistency among the generated class labels and smoothens the edges of class labels on segmented region.

Results

The evaluation has performed on the dataset obtained from center for imaging of neurodegenerative diseases (CIND) for CA1, CA2, DG, CA3, Head, Tail, SUB, ERC and PHG in hippocampal subfields, resulting in the dice similarity coefficient (DSC) of 0.919, 0.648, 0.903, 0.673, 0.929, 0.913, 0.906, 0.884 and 0.889 respectively. For the large subfields, such as Head and CA1 of hippocampus, the DSC was increased by 3.9% and 9.03% than state-of-the-art approaches, while for the smaller subfields, such as ERC and PHG, the segmentation accuracy was significantly increased 20.93% and 16.30% respectively.

Conclusion

The results show the improvement in performance of the proposed method, compared with other methods, which include approaches based on multi-atlas, hierarchical multi-atlas, dictionary learning and sparse representation and CNN. In implementation, the proposed method provides better results in hippocampal subfields segmentation.

Multi-Template Mesiotemporal Lobe Segmentation: Effects of Surface and Volume Feature Modeling

Numerous neurological disorders are associated with atrophy of mesiotemporal lobe structures, including the hippocampus (HP), amygdala (AM), and entorhinal cortex (EC). Accurate segmentation of these structures is, therefore, necessary for understanding the disease process and patient management. Recent multiple-template segmentation algorithms have shown excellent performance in HP segmentation. Purely surface-based methods precisely describe structural boundary but their performance likely depends on a large template library, as segmentation suffers when the boundaries of template and individual MRI are not well aligned while volume-based methods are less dependent. So far only few algorithms attempted segmentation of entire mesiotemporal structures including the parahippocampus. We compared performance of surface- and volume-based approaches in segmenting the three mesiotemporal structures and assess the effects of different environments (i.e., size of templates, under pathology). We also proposed an algorithm that combined surface- with volume-derived similarity measures for optimal template selection. To further improve the method, we introduced two new modules: (1) a non-linear registration that is driven by volume-based intensities and features sampled on deformable template surfaces; (2) a shape averaging based on regional weighting using multi-scale global-to-local icosahedron sampling. Compared to manual segmentations, our approach, namely HybridMulti showed high accuracy in 40 healthy controls (mean Dice index for HP/AM/EC = 89.7/89.3/82.9%) and 135 patients with temporal lobe epilepsy (88.7/89.0/82.6%). This accuracy was comparable across two different datasets of 1.5T and 3T MRI. It resulted in the best performance among tested multi-template methods that were either based on volume or surface data alone in terms of accuracy and sensitivity to detect atrophy related to epilepsy. Moreover, unlike purely surface-based multi-template segmentation, HybridMulti could maintain accurate performance even with a 50% template library size.

Hippocampal involvement in nonpathological déjà vu: Subfield vulnerability rather than temporal lobe epilepsy equivalent

INTRODUCTION

Morphological correlates of nonpathological déjà vu (DV) have been identified recently within the human brain. Significantly reduced gray matter volume (GMV) within a set of cortical and subcortical regions reported in subjects experiencing DV seems to mirror the distribution of GMV reduction in mesial temporal lobe epilepsy (MTLE) patients but vary in terms of the hippocampus. Another condition associated with hippocampal GMV reduction and DV alike disturbance in memory processing is schizophrenia (SCH). Here, we tested the hypothesis that hippocampal involvement in nonpathological DV resembles more closely the pattern of GMV decrease observed in MTLE compared with that occurring in SCH.

METHODS

Using automated segmentation of the MRI data we compared the medians of GMV within 12 specific hippocampal subfields in healthy individuals that do (DV+; N = 87) and do not report déjà vu experience (DV-; N = 26), and patients with MTLE (N = 47) and SCH (N = 29). By Pearson correlation, we then evaluated the similarity of MTLE and SCH groups to DV+ group with respect to spatial distribution of GMV deviation from DV- group.

RESULTS

Significant GMV decrease was found in MTLE group in most of the subfields. There were just trends in the hippocampal GMV decrease found in DV+ or SCH groups. Concerning the spatial distribution of GMV decrease, we revealed statistically significant correlation for the left hippocampus for SCH vs DV+. Otherwise there was no statistically significant correlation.

CONCLUSIONS

Our findings reveal structural features of hippocampal involvement in nonpathological DV, MTLE, and SCH. Despite our expectations, the pattern of GMV reduction in the DV+ relative to the DV- group does not resemble the pattern observed in MTLE any more than that observed in SCH. The highly similar patterns of the three clinical groups rather suggest an increased vulnerability of certain hippocampal subfields; namely, Cornu Ammonis (CA)4, CA3, dentate gyrus granular cell layer (GC-DG), hippocampal-amygdaloid transition area (HATA) and subiculum.

Egocentric and allocentric visuospatial working memory in premotor Huntington's disease: A double dissociation with caudate and hippocampal volumes

Our brains represent spatial information in egocentric (self-based) or allocentric (landmark-based) coordinates. Rodent studies have demonstrated a critical role for the caudate in egocentric navigation and the hippocampus in allocentric navigation. We administered tests of egocentric and allocentric working memory to individuals with premotor Huntington's disease (pmHD), which is associated with early caudate nucleus atrophy, and controls. Each test had 80 trials during which subjects were asked to remember 2 locations over 1-sec delays. The only difference between these otherwise identical tests was that locations could only be coded in self-based or landmark-based coordinates. We applied a multiatlas-based segmentation algorithm and computed point-wise Jacobian determinants to measure regional variations in caudate and hippocampal volumes from 3T MRI. As predicted, the pmHD patients were significantly more impaired on egocentric working memory. Only egocentric accuracy correlated with caudate volumes, specifically the dorsolateral caudate head, right more than left, a region that receives dense efferents from dorsolateral prefrontal cortex. In contrast, only allocentric accuracy correlated with hippocampal volumes, specifically intermediate and posterior regions that connect strongly with parahippocampal and posterior parietal cortices. These results indicate that the distinction between egocentric and allocentric navigation applies to working memory. The dorsolateral caudate is important for egocentric working memory, which can explain the disproportionate impairment in pmHD. Allocentric working memory, in contrast, relies on the hippocampus and is relatively spared in pmHD.

Incomplete Hippocampal Inversion: A Comprehensive MRI Study of Over 2000 Subjects

The incomplete-hippocampal-inversion (IHI), also known as malrotation, is an atypical anatomical pattern of the hippocampus, which has been reported in healthy subjects in different studies. However, extensive characterization of IHI in a large sample has not yet been performed. Furthermore, it is unclear whether IHI are restricted to the medial-temporal lobe or are associated with more extensive anatomical changes. Here, we studied the characteristics of IHI in a community-based sample of 2008 subjects of the IMAGEN database and their association with extra-hippocampal anatomical variations. The presence of IHI was assessed on T1-weighted anatomical magnetic resonance imaging (MRI) using visual criteria. We assessed the association of IHI with other anatomical changes throughout the brain using automatic morphometry of cortical sulci. We found that IHI were much more frequent in the left hippocampus (left: 17%, right: 6%, χ(2)-test, p < 10(-28)). Compared to subjects without IHI, subjects with IHI displayed morphological changes in several sulci located mainly in the limbic lobe. Our results demonstrate that IHI are a common left-sided phenomenon in normal subjects and that they are associated with morphological changes outside the medial temporal lobe.

Local manifold learning for multiatlas segmentation: application to hippocampal segmentation in healthy population and Alzheimer's disease

AIMS

Automated hippocampal segmentation is an important issue in many neuroscience studies.

METHODS

We presented and evaluated a novel segmentation method that utilized a manifold learning technique under the multiatlas-based segmentation scenario. A manifold representation of local patches for each voxel was achieved by applying an Isomap algorithm, which can then be used to obtain spatially local weights of atlases for label fusion. The obtained atlas weights potentially depended on all pairwise similarities of the population, which is in contrast to most existing label fusion methods that only rely on similarities between the target image and the atlases. The performance of the proposed method was evaluated for hippocampal segmentation and compared with two representative local weighted label fusion methods, that is, local majority voting and local weighted inverse distance voting, on an in-house dataset of 28 healthy adolescents (age range: 10-17 years) and two ADNI datasets of 100 participants (age range: 60-89 years). We also implemented hippocampal volumetric analysis and evaluated segmentation performance using atlases from a different dataset.

RESULTS

The median Dice similarities obtained by our proposed method were approximately 0.90 for healthy subjects and above 0.88 for two mixed diagnostic groups of ADNI subjects.

CONCLUSION

The experimental results demonstrated that the proposed method could obtain consistent and significant improvements over label fusion strategies that are implemented in the original space.

Magnetic resonance imaging pattern learning in temporal lobe epilepsy: classification and prognostics

OBJECTIVE

In temporal lobe epilepsy (TLE), although hippocampal atrophy lateralizes the focus, the value of magnetic resonance imaging (MRI) to predict postsurgical outcome is rather modest. Prediction solely based on the hippocampus may be hampered by widespread mesiotemporal structural damage shown by advanced imaging. Increasingly complex and high-dimensional representation of MRI metrics motivates a shift to machine learning to establish objective, data-driven criteria for pathogenic processes and prognosis.

METHODS

We applied clustering to 114 consecutive unilateral TLE patients using 1.5T MRI profiles derived from surface morphology of hippocampus, amygdala, and entorhinal cortex. To evaluate the diagnostic validity of the classification, we assessed its yield to predict outcome in 79 surgically treated patients. Reproducibility of outcome prediction was assessed in an independent cohort of 27 patients evaluated on 3.0T MRI.

RESULTS

Four similarly sized classes partitioned our cohort; in all, alterations spanned over the 3 mesiotemporal structures. Compared to 46 controls, TLE-I showed marked bilateral atrophy; in TLE-II atrophy was ipsilateral; TLE-III showed mild bilateral atrophy; whereas TLE-IV showed hypertrophy. Classes differed with regard to histopathology and freedom from seizures. Classwise surface-based classifiers accurately predicted outcome in 92 ± 1% of patients, outperforming conventional volumetry. Predictors of relapse were distributed bilaterally across structures. Prediction accuracy was similarly high in the independent cohort (96%), supporting generalizability.

INTERPRETATION

We provide a novel description of individual variability across the TLE spectrum. Class membership was associated with distinct patterns of damage and outcome predictors that did not spatially overlap, emphasizing the ability of machine learning to disentangle the differential contribution of morphology to patient phenotypes, ultimately refining the prognosis of epilepsy surgery.

Automated voxel-by-voxel tissue classification for hippocampal segmentation: methods and validation

The hippocampus is an important structural biomarker for Alzheimer's disease (AD) and has a primary role in the pathogenesis of other neurological and psychiatric diseases. This study presents a fully automated pattern recognition system for an accurate and reproducible segmentation of the hippocampus in structural Magnetic Resonance Imaging (MRI). The method was validated on a mixed cohort of 56 T1-weighted structural brain images, and consists of three processing levels: (a) Linear registration: all brain images were registered to a standard template and an automated method was applied to capture the global shape of the hippocampus. (b) Feature extraction: all voxels included in the previously selected volume were characterized by 315 features computed from local information. (c) Voxel classification: a Random Forest algorithm was used to classify voxels as belonging or not belonging to the hippocampus. In order to improve the classification performance, an adaptive learning method based on the use of the Pearson's correlation coefficient was developed. The segmentation results (Dice similarity index = 0.81 ± 0.03) compare well with other state-of-the art approaches. A validation study was conducted on an independent dataset of 100 T1-weighted brain images, achieving significantly better results than those obtained with FreeSurfer.

Hippocampal surface deformation accuracy in T1-weighted volumetric MRI sequences in subjects with epilepsy

BACKGROUND AND PURPOSE

To demonstrate the accuracy across different acquisition and analysis methods, we evaluated the variability in hippocampal volumetric and surface displacement measurements resulting from two different MRI (magnetic resonance imaging) acquisition protocols.

METHODS

Nine epilepsy patients underwent two independent T1-weighted magnetization prepared spoiled gradient sequences during a single 3T MRI session. Using high-dimension mapping-large deformation (HDM-LD) segmentation, we calculated volumetric estimates and generated a vector-based 3-dimensional surface model of each subject's hippocampi, and evaluated volume and surface changes, the latter using a cluster-based noise estimation model.

RESULTS

Mean hippocampal volumes and standard deviations for the left hippocampi were 2,750 (826) mm3 and 2,782 (859) mm3 (P = .13), and for the right hippocampi were 2,558 (750) mm3 and 2,547 (692) mm3 (P = .76), respectively for the MPR1 and MPR2 sequences. Average Dice coefficient comparing overlap for segmentations was 86%. There was no significant effect of MRI sequence on volume estimates and no significant hippocampal surface change between sequences.

CONCLUSION

Statistical comparison of hippocampal volumes and statistically thresholded HDM-LD surfaces in TLE patients showed no differences between the segmentations obtained in the two MRI acquisition sequences. This validates the robustness across MRI sequences of the HDM-LD technique for estimating volume and surface changes in subjects with epilepsy.

Quantitative analysis of structural neuroimaging of mesial temporal lobe epilepsy

Mesial temporal lobe epilepsy (MTLE) is the most common of the surgically remediable drug-resistant epilepsies. MRI is the primary diagnostic tool to detect anatomical abnormalities and, when combined with EEG, can more accurately identify an epileptogenic lesion, which is often hippocampal sclerosis in cases of MTLE. As structural imaging technology has advanced the surgical treatment of MTLE and other lesional epilepsies, so too have the analysis techniques that are used to measure different structural attributes of the brain. These techniques, which are reviewed here and have been used chiefly in basic research of epilepsy and in studies of MTLE, have identified different types and the extent of anatomical abnormalities that can extend beyond the affected hippocampus. These results suggest that structural imaging and sophisticated imaging analysis could provide important information to identify networks capable of generating spontaneous seizures and ultimately help guide surgical therapy that improves postsurgical seizure-freedom outcomes.