Investigation of morphometric variability of subthalamic nucleus, red nucleus, and substantia nigra in advanced Parkinson's disease patients using automatic segmentation and PCA-based analysis

Projected pooling loss for red nucleus segmentation with soft topology constraints

Purpose

Deep learning is the standard for medical image segmentation. However, it may encounter difficulties when the training set is small. Also, it may generate anatomically aberrant segmentations. Anatomical knowledge can be potentially useful as a constraint in deep learning segmentation methods. We propose a loss function based on projected pooling to introduce soft topological constraints. Our main application is the segmentation of the red nucleus from quantitative susceptibility mapping (QSM) which is of interest in parkinsonian syndromes.

Approach

This new loss function introduces soft constraints on the topology by magnifying small parts of the structure to segment to avoid that they are discarded in the segmentation process. To that purpose, we use projection of the structure onto the three planes and then use a series of MaxPooling operations with increasing kernel sizes. These operations are performed both for the ground truth and the prediction and the difference is computed to obtain the loss function. As a result, it can reduce topological errors as well as defects in the structure boundary. The approach is easy to implement and computationally efficient.

Results

When applied to the segmentation of the red nucleus from QSM data, the approach led to a very high accuracy (Dice 89.9%) and no topological errors. Moreover, the proposed loss function improved the Dice accuracy over the baseline when the training set was small. We also studied three tasks from the medical segmentation decathlon challenge (MSD) (heart, spleen, and hippocampus). For the MSD tasks, the Dice accuracies were similar for both approaches but the topological errors were reduced.

Conclusions

We propose an effective method to automatically segment the red nucleus which is based on a new loss for introducing topology constraints in deep learning segmentation.

High-resolution atlasing and segmentation of the subcortex: Review and perspective on challenges and opportunities created by machine learning

This paper reviews almost three decades of work on atlasing and segmentation methods for subcortical structures in human brain MRI. In writing this survey, we have three distinct aims. First, to document the evolution of digital subcortical atlases of the human brain, from the early MRI templates published in the nineties, to the complex multi-modal atlases at the subregion level that are available today. Second, to provide a detailed record of related efforts in the automated segmentation front, from earlier atlas-based methods to modern machine learning approaches. And third, to present a perspective on the future of high-resolution atlasing and segmentation of subcortical structures in in vivo human brain MRI, including open challenges and opportunities created by recent developments in machine learning.

Automated Segmentation of Midbrain Structures in High-Resolution Susceptibility Maps Based on Convolutional Neural Network and Transfer Learning

Background

Accurate delineation of the midbrain nuclei, the red nucleus (RN), substantia nigra (SN) and subthalamic nucleus (STN), is important in neuroimaging studies of neurodegenerative and other diseases. This study aims to segment midbrain structures in high-resolution susceptibility maps using a method based on a convolutional neural network (CNN).

Methods

The susceptibility maps of 75 subjects were acquired with a voxel size of 0.83 × 0.83 × 0.80 mm³ on a 3T MRI system to distinguish the RN, SN, and STN. A deeply supervised attention U-net was pre-trained with a dataset of 100 subjects containing susceptibility maps with a voxel size of 0.63 × 0.63 × 2.00 mm³ to provide initial weights for the target network. Five-fold cross-validation over the training cohort was used for all the models’ training and selection. The same test cohort was used for the final evaluation of all the models. Dice coefficients were used to assess spatial overlap agreement between manual delineations (ground truth) and automated segmentation. Volume and magnetic susceptibility values in the nuclei extracted with automated CNN delineation were compared to those extracted by manual tracing. Consistencies of volume and magnetic susceptibility values by different extraction strategies were assessed by Pearson correlation coefficients and Bland-Altman analyses.

Results

The automated CNN segmentation method achieved mean Dice scores of 0.903, 0.864, and 0.777 for the RN, SN, and STN, respectively. There were no significant differences between the achieved Dice scores and the inter-rater Dice scores (p > 0.05 for each nucleus). The overall volume and magnetic susceptibility values of the nuclei extracted by the automatic CNN method were significantly correlated with those by manual delineation (p < 0.01).

Conclusion

Midbrain structures can be precisely segmented in high-resolution susceptibility maps using a CNN-based method.

Functional Topography of the Human Subthalamic Nucleus: Relevance for Subthalamotomy in Parkinson's Disease

BACKGROUND

The subthalamic nucleus (STN) is considered a key structure in motor, behavioral, and emotional control. Although identification of the functional topography of the STN has therapeutic implications in the treatment of the motor features of Parkinson's disease (PD), the details of its functional and somatotopic organization in humans are not well understood.

OBJECTIVE

The aim of this study was to characterize the functional organization of the STN and its correlation with the motor outcomes induced by subthalamotomy.

METHODS

We used diffusion-weighted imaging to assess STN connectivity patterns in 23 healthy control subjects and 86 patients with PD, of whom 39 received unilateral subthalamotomy. Analytical tractography was used to reconstruct structural cortico-subthalamic connectivity. A diffusion-weighted imaging/functional magnetic resonance imaging-driven somatotopic parcellation of the STN was defined to delineate the representation of the upper and lower limb in the STN.

RESULTS

We confirmed a connectional gradient to sensorimotor, supplementary-motor, associative, and limbic cortical regions, spanning from posterior-dorsal-lateral to anterior-ventral-medial portions of the STN, with intermediate overlapping zones. Functional magnetic resonance imaging-driven parcellation demonstrated dual segregation of motor cortico-subthalamic projections in humans. Moreover, the relationship between lesion topography and functional anatomy of the STN explains specific improvement in bradykinesia, rigidity, and tremor induced by subthalamotomy.

CONCLUSIONS

Our results support an interplay between segregation and integration of cortico-subthalamic projections, suggesting the coexistence of parallel and convergent information processing. Identifying the functional topography of the STN will facilitate better definition of the optimal location for functional neurosurgical approaches, that is, electrode placement and lesion location, and improve specific cardinal features in PD. © 2021 International Parkinson and Movement Disorder Society.

Automated segmentation of deep brain nuclei using convolutional neural networks and susceptibility weighted imaging

The advent of susceptibility-sensitive MRI techniques, such as susceptibility weighted imaging (SWI), has enabled accurate in vivo visualization and quantification of iron deposition within the human brain. Although previous approaches have been introduced to segment iron-rich brain regions, such as the substantia nigra, subthalamic nucleus, red nucleus, and dentate nucleus, these methods are largely unavailable and manual annotation remains the most used approach to label these regions. Furthermore, given their recent success in outperforming other segmentation approaches, convolutional neural networks (CNN) promise better performances. The aim of this study was thus to evaluate state-of-the-art CNN architectures for the labeling of deep brain nuclei from SW images. We implemented five CNN architectures and considered ensembles of these models. Furthermore, a multi-atlas segmentation model was included to provide a comparison not based on CNN. We evaluated two prediction strategies: individual prediction, where a model is trained independently for each region, and combined prediction, which simultaneously predicts multiple closely located regions. In the training dataset, all models performed with high accuracy with Dice coefficients ranging from 0.80 to 0.95. The regional SWI intensities and volumes from the models' labels were strongly correlated with those obtained from manual labels. Performances were reduced on the external dataset, but were higher or comparable to the intrarater reliability and most models achieved significantly better results compared to multi-atlas segmentation. CNNs can accurately capture the individual variability of deep brain nuclei and represent a highly useful tool for their segmentation from SW images.

Probabilistic Atlas of the Mesencephalic Reticular Formation, Isthmic Reticular Formation, Microcellular Tegmental Nucleus, Ventral Tegmental Area Nucleus Complex, and Caudal-Rostral Linear Raphe Nucleus Complex in Living Humans from 7 Tesla Magnetic Resonance Imaging

Introduction: The mesencephalic reticular formation, isthmic reticular formation, microcellular tegmental nucleus, ventral tegmental area-parabrachial pigmented nucleus complex, and caudal-rostral linear nucleus of the raphe are small brainstem regions crucially involved in arousal, sleep, and reward. Yet, these nuclei are difficult to identify with magnetic resonance imaging (MRI) of living humans. In the current work, we developed a probabilistic atlas of these brainstem nuclei in living humans, using noninvasive ultra-high-field MRI. Methods: We acquired single-subject, multicontrast (diffusion and T₂-weighted), 1.1-mm isotropic resolution, 7 Tesla MRI images of 12 healthy subjects. After preprocessing and alignment to the stereotactic space, these images were used to delineate (in each subject) the nuclei of interest based on the image contrast as well as on neighboring nuclei and landmarks. Nucleus labels were averaged across subjects to yield probabilistic labels. The latter were further validated by assessment of the label inter-rater agreement, internal consistency, and volume. Results: Labels were delineated for each nucleus with good overlap across subjects. The inter-rater agreement and internal consistency were below (p < 10^-8) the linear spatial imaging resolution (1.1 mm), thus validating the generated probabilistic atlas labels. The volumes of our labels did not differ from literature volumes (p < 0.05), further validating our atlas. Discussion and Conclusion: The probabilistic atlas of these five mesopontine nuclei expands current in vivo brainstem nuclei atlases and can be used as a tool to identify the location of these areas in conventional (e.g., 3 Tesla) images. This might serve to unravel the brainstem structure-to-function link and thus improve clinical outcomes. Impact statement The mesencephalic reticular formation, isthmic reticular formation, microcellular tegmental nucleus, ventral tegmental area-parabrachial pigmented nucleus complex, and caudal-rostral linear nucleus of the raphe are small brainstem regions crucially involved in arousal, sleep, and reward. In the current work, we developed a probabilistic atlas of these brainstem nuclei in living humans, using noninvasive, ultra-high-field magnetic resonance imaging. The probabilistic atlas of these five mesopontine nuclei expands current in vivo brainstem nuclei atlases and can be used as a tool to identify the location of these areas in conventional (e.g., 3 Tesla) images. This might serve to unravel the brainstem structure-to-function link and thus improve clinical outcomes.

Statistical morphological analysis reveals characteristic paraspinal muscle asymmetry in unilateral lumbar disc herniation

Growing evidence suggests an association of lumbar paraspinal muscle morphology with low back pain (LBP) and lumbar pathologies. Unilateral spinal disorders provide unique models to study this association, with implications for diagnosis, prognosis, and management. Statistical shape analysis is a technique that can identify signature shape variations related to phenotypes but has never been employed in studying paraspinal muscle morphology. We present the first investigation using this technique to reveal disease-related paraspinal muscle asymmetry, using MRIs of patients with a single posterolateral disc herniation at the L5-S1 spinal level and unilateral leg pain. Statistical shape analysis was conducted to reveal disease- and phenotype-related morphological variations in the multifidus and erector spinae muscles at the level of herniation and the one below. With the analysis, shape variations associated with disc herniation were identified in the multifidus on the painful side at the level below the pathology while no pathology-related asymmetry in cross-sectional area (CSA) and fatty infiltration was found in either muscle. The results demonstrate higher sensitivity and spatial specificity for the technique than typical CSA and fatty infiltration measures. Statistical shape analysis holds promise in studying paraspinal muscle morphology to improve our understanding of LBP and various lumbar pathologies.

Characterizing white matter alterations subject to clinical laterality in drug-naïve de novo Parkinson's disease

Parkinson's disease (PD) is a progressive neurodegenerative disorder that is characterized by a range of motor and nonmotor symptoms, often with the motor dysfunction initiated unilaterally. Knowledge regarding disease‐related alterations in white matter pathways can effectively help improve the understanding of the disease and propose targeted treatment strategies. Microstructural imaging techniques, including diffusion tensor imaging (DTI), allows inspection of white matter integrity to study the pathogenesis of various neurological conditions. Previous voxel‐based analyses with DTI measures, such as fractional anisotropy and mean diffusivity have uncovered changes in brain regions that are associated with PD, but the conclusions were inconsistent, partially due to small patient cohorts and the lack of consideration for clinical laterality onset, particularly in early PD. Fixel‐based analysis (FBA) is a recent framework that offers tract‐specific insights regarding white matter health, but very few FBA studies on PD exist. We present a study that reveals strengthened and weakened white matter integrity that is subject to symptom laterality in a large drug‐naïve de novo PD cohort using complementary DTI and FBA measures. The findings suggest that the disease gives rise to tissue degeneration and potential re‐organization in the early stage.

Atlas-independent, N-of-1 tissue activation modeling to map optimal regions of subthalamic deep brain stimulation for Parkinson disease

•

Neuroanatomical variations among patients are obscured in atlas-based VTA modeling.

•

N-of-1 neuroanatomical and VTA modeling enables patient-level precision.

•

Mean optimal stimulation is dorsomedial to the STN, near its posterior half.

•

Individual VTAs deviate from optimal stimulation sites to varying degrees.

•

Optimal stimulation sites for rigidity, bradykinesia, and tremor partially overlap.

Background

Motor outcomes after subthalamic deep brain stimulation (STN DBS) for Parkinson disease (PD) vary considerably among patients and strongly depend on stimulation location. The objective of this retrospective study was to map the regions of optimal STN DBS for PD using an atlas-independent, fully individualized (N-of-1) tissue activation modeling approach and to assess the relationship between patient-level therapeutic volumes of tissue activation (VTAs) and motor improvement.

Methods

The stimulation-induced electric field for 40 PD patients treated with bilateral STN DBS was modeled using finite element analysis. Neurostimulation models were generated for each patient, incorporating their individual STN anatomy, DBS lead position and orientation, anisotropic tissue conductivity, and clinical stimulation settings. A voxel-based analysis of the VTAs was then used to map the optimal location of stimulation. The amount of stimulation in specific regions relative to the STN was measured and compared between STNs with more and less optimal stimulation, as determined by their motor improvement scores and VTA. The relationship between VTA location and motor outcome was then assessed using correlation analysis. Patient variability in terms of STN anatomy, active contact position, and VTA location were also evaluated. Results from the N-of-1 model were compared to those from a simplified VTA model.

Results

Tissue activation modeling mapped the optimal location of stimulation to regions medial, posterior, and dorsal to the STN centroid. These regions extended beyond the STN boundary towards the caudal zona incerta (cZI). The location of the VTA and active contact position differed significantly between STNs with more and less optimal stimulation in the dorsal-ventral and anterior-posterior directions. Therapeutic stimulation spread noticeably more in the dorsal and posterior directions, providing additional evidence for cZI as an important DBS target. There were significant linear relationships between the amount of dorsal and posterior stimulation, as measured by the VTA, and motor improvement. These relationships were more robust than those between active contact position and motor improvement. There was high variability in STN anatomy, active contact position, and VTA location among patients. Spherical VTA modeling was unable to reproduce these results and tended to overestimate the size of the VTA.

Conclusion

Accurate characterization of the spread of stimulation is needed to optimize STN DBS for PD. High variability in neuroanatomy, stimulation location, and motor improvement among patients highlights the need for individualized modeling techniques. The atlas-independent, N-of-1 tissue activation modeling approach presented in this study can be used to develop and evaluate stimulation strategies to improve clinical outcomes on an individual basis.

Methodological Considerations for Neuroimaging in Deep Brain Stimulation of the Subthalamic Nucleus in Parkinson's Disease Patients

Deep brain stimulation (DBS) of the subthalamic nucleus is a neurosurgical intervention for Parkinson's disease patients who no longer appropriately respond to drug treatments. A small fraction of patients will fail to respond to DBS, develop psychiatric and cognitive side-effects, or incur surgery-related complications such as infections and hemorrhagic events. In these cases, DBS may require recalibration, reimplantation, or removal. These negative responses to treatment can partly be attributed to suboptimal pre-operative planning procedures via direct targeting through low-field and low-resolution magnetic resonance imaging (MRI). One solution for increasing the success and efficacy of DBS is to optimize preoperative planning procedures via sophisticated neuroimaging techniques such as high-resolution MRI and higher field strengths to improve visualization of DBS targets and vasculature. We discuss targeting approaches, MRI acquisition, parameters, and post-acquisition analyses. Additionally, we highlight a number of approaches including the use of ultra-high field (UHF) MRI to overcome limitations of standard settings. There is a trade-off between spatial resolution, motion artifacts, and acquisition time, which could potentially be dissolved through the use of UHF-MRI. Image registration, correction, and post-processing techniques may require combined expertise of traditional radiologists, clinicians, and fundamental researchers. The optimization of pre-operative planning with MRI can therefore be best achieved through direct collaboration between researchers and clinicians.

Image Guidance in Deep Brain Stimulation Surgery to Treat Parkinson's Disease: A Comprehensive Review

Deep brain stimulation (DBS) is an effective therapy as an alternative to pharmaceutical treatments for Parkinson's disease (PD). Aside from factors such as instrumentation, treatment plans, and surgical protocols, the success of the procedure depends heavily on the accurate placement of the electrode within the optimal therapeutic targets while avoiding vital structures that can cause surgical complications and adverse neurologic effects. Although specific surgical techniques for DBS can vary, interventional guidance with medical imaging has greatly contributed to the development, outcomes, and safety of the procedure. With rapid development in novel imaging techniques, computational methods, and surgical navigation software, as well as growing insights into the disease and mechanism of action of DBS, modern image guidance is expected to further enhance the capacity and efficacy of the procedure in treating PD. This article surveys the state-of-the-art techniques in image-guided DBS surgery to treat PD, and discusses their benefits and drawbacks, as well as future directions on the topic.

Direct visualization and characterization of the human zona incerta and surrounding structures

The zona incerta (ZI) is a small gray matter region of the deep brain first identified in the 19th century, yet direct in vivo visualization and characterization has remained elusive. Noninvasive detection of the ZI and surrounding region could be critical to further our understanding of this widely connected but poorly understood deep brain region and could contribute to the development and optimization of neuromodulatory therapies. We demonstrate that high resolution (submillimetric) longitudinal (T1) relaxometry measurements at high magnetic field strength (7 T) can be used to delineate the ZI from surrounding white matter structures, specifically the fasciculus cerebellothalamicus, fields of Forel (fasciculus lenticularis, fasciculus thalamicus, and field H), and medial lemniscus. Using this approach, we successfully derived in vivo estimates of the size, shape, location, and tissue characteristics of substructures in the ZI region, confirming observations only previously possible through histological evaluation that this region is not just a space between structures but contains distinct morphological entities that should be considered separately. Our findings pave the way for increasingly detailed in vivo study and provide a structural foundation for precise functional and neuromodulatory investigation.

Kinesthetic Cells within the Subthalamic Nucleus and Deep Brain Stimulation for Parkinson Disease

OBJECTIVE

We sought to determine the location of kinesthetic cell clusters within the subthalamic nucleus (STN) on magnetic resonance imaging, adjusted for interindividual anatomic variability by employing the medial STN border as a reference point.

METHODS

We retrospectively localized microelectrode recording-defined kinesthetic cells on 3-Tesla T2-weighted and susceptibility-weighted images in patients who underwent STN deep brain stimulation for Parkinson disease and averaged the stereotactic coordinates. These locations were calculated relative to the nonindividualized midcommissural point (MCP) and, in order to account for interindividual anatomic variability, also calculated relative to the patient-specific intersection of Bejjani line with the medial STN border. Two example patients were selected in order to visualize the discrepancies between the adjusted and nonadjusted theoretic kinesthetic cell clusters on magnetic resonance imaging.

RESULTS

Relative to the MCP, average kinesthetic cell coordinates were 12.3 ± 1.2 mm lateral, 1.7 ± 1.4 mm posterior, and 2.3 ± 1.5 mm inferior. Relative to the medial STN border, mean coordinates were 3.4 ± 1.0 mm lateral, 1.0 ± 1.4 mm anterior, and 1.7 ± 1.5 mm superior on T2-sequences, and on susceptibility-weighted images mean coordinates were 3.2 ± 1.1 mm lateral, 0.8 ± 1.5 mm anterior, and 2.1 ± 1.5 mm superior. The theoretic kinesthetic cell clusters may appear outside the sensorimotor STN when using the MCP, whereas these clusters fall well within the sensorimotor STN when employing the medial STN border as a reference point.

CONCLUSIONS

By using the medial STN border as a patient-specific anatomic reference point in STN deep brain stimulation for Parkinson disease, we accounted for interindividual anatomic variability and provided accurate insight in the clustering of kinesthetic cells within the dorsolateral STN.

pBrain: A novel pipeline for Parkinson related brain structure segmentation

•

A novel pipeline for Parkinson`s disease structure segmentation is presented.

•

State-of-the-art fast multiatlas patch-based label fusion with systematic error correction is used to accurately and efficiently produce very competitive results in around 5 min.

•

The proposed pipeline works at high resolution (0.5 mm) but it can work also with standard resolution (1 mm) T2 images allowing the analysis of large legacy databases.

•

The proposed pipeline will be made publically available online through our volBrain platform.

Parkinson is a very prevalent neurodegenerative disease impacting the life of millions of people worldwide. Although its cause remains unknown, its functional and structural analysis is fundamental to advance in the search of a cure or symptomatic treatment. The automatic segmentation of deep brain structures related to Parkinson`s disease could be beneficial for the follow up and treatment planning. Unfortunately, there is not broadly available segmentation software to automatically measure Parkinson related structures. In this paper, we present a novel pipeline to segment three deep brain structures related to Parkinson's disease (substantia nigra, subthalamic nucleus and red nucleus). The proposed method is based on the multi-atlas label fusion technology that works on standard and high-resolution T2-weighted images. The proposed method also includes as post-processing a new neural network-based error correction step to minimize systematic segmentation errors. The proposed method has been compared to other state-of-the-art methods showing competitive results in terms of accuracy and execution time.

Deep Learning-Based Deep Brain Stimulation Targeting and Clinical Applications

Background

The purpose of the present study was to evaluate deep learning-based image-guided surgical planning for deep brain stimulation (DBS). We developed deep learning semantic segmentation-based DBS targeting and prospectively applied the method clinically.

Methods

T2^∗ fast gradient-echo images from 102 patients were used for training and validation. Manually drawn ground truth information was prepared for the subthalamic and red nuclei with an axial cut ∼4 mm below the anterior–posterior commissure line. A fully convolutional neural network (FCN-VGG-16) was used to ensure margin identification by semantic segmentation. Image contrast augmentation was performed nine times. Up to 102 original images and 918 augmented images were used for training and validation. The accuracy of semantic segmentation was measured in terms of mean accuracy and mean intersection over the union. Targets were calculated based on their relative distance from these segmented anatomical structures considering the Bejjani target.

Results

Mean accuracies and mean intersection over the union values were high: 0.904 and 0.813, respectively, for the 62 training images, and 0.911 and 0.821, respectively, for the 558 augmented training images when 360 augmented validation images were used. The Dice coefficient converted from the intersection over the union was 0.902 when 720 training and 198 validation images were used. Semantic segmentation was adaptive to high anatomical variations in size, shape, and asymmetry. For clinical application, two patients were assessed: one with essential tremor and another with bradykinesia and gait disturbance due to Parkinson’s disease. Both improved without complications after surgery, and microelectrode recordings showed subthalamic nuclei signals in the latter patient.

Conclusion

The accuracy of deep learning-based semantic segmentation may surpass that of previous methods. DBS targeting and its clinical application were made possible using accurate deep learning-based semantic segmentation, which is adaptive to anatomical variations.

An accurate registration of the BigBrain dataset with the MNI PD25 and ICBM152 atlases

Brain atlases that encompass detailed anatomical or physiological features are instrumental in the research and surgical planning of various neurological conditions. Magnetic resonance imaging (MRI) has played important roles in neuro-image analysis while histological data remain crucial as a gold standard to guide and validate such analyses. With cellular-scale resolution, the BigBrain atlas offers 3D histology of a complete human brain, and is highly valuable to the research and clinical community. To bridge the insights at macro- and micro-levels, accurate mapping of BigBrain and established MRI brain atlases is necessary, but the existing registration is unsatisfactory. The described dataset includes co-registration of the BigBrain atlas to the MNI PD25 atlas and the ICBM152 2009b atlases (symmetric and asymmetric versions) in addition to manual segmentation of the basal ganglia, red nucleus, amygdala, and hippocampus for all mentioned atlases. The dataset intends to provide a bridge between insights from histological data and MRI studies in research and neurosurgical planning. The registered atlases, anatomical segmentations, and deformation matrices are available at: https://osf.io/xkqb3/ .

MRI Atlas of the Human Deep Brain

Mastering detailed anatomy of the human deep brain in clinical neurosciences is challenging. Although numerous pioneering works have gathered a large dataset of structural and topographic information, it is still difficult to transfer this knowledge into practice, even with advanced magnetic resonance imaging techniques. Thus, classical histological atlases continue to be used to identify structures for stereotactic targeting in functional neurosurgery. Physicians mainly use these atlases as a template co-registered with the patient's brain. However, it is possible to directly identify stereotactic targets on MRI scans, enabling personalized targeting. In order to help clinicians directly identify deep brain structures relevant to present and future medical applications, we built a volumetric MRI atlas of the deep brain (MDBA) on a large scale (infra millimetric). Twelve hypothalamic, 39 subthalamic, 36 telencephalic, and 32 thalamic structures were identified, contoured, and labeled. Nineteen coronal, 18 axial, and 15 sagittal MRI plates were created. Although primarily designed for direct labeling, the anatomic space was also subdivided in twelfths of AC-PC distance, leading to proportional scaling in the coronal, axial, and sagittal planes. This extensive work is now available to clinicians and neuroscientists, offering another representation of the human deep brain ([https://hal.archives-ouvertes.fr/] [hal-02116633]). The atlas may also be used by computer scientists who are interested in deciphering the topography of this complex region.

Cortico-basal white matter alterations occurring in Parkinson's disease

Magnetic resonance imaging studies typically use standard anatomical atlases for identification and analyses of (patho-)physiological effects on specific brain areas; these atlases often fail to incorporate neuroanatomical alterations that may occur with both age and disease. The present study utilizes Parkinson's disease and age-specific anatomical atlases of the subthalamic nucleus for diffusion tractography, assessing tracts that run between the subthalamic nucleus and a-priori defined cortical areas known to be affected by Parkinson's disease. The results show that the strength of white matter fiber tracts appear to remain structurally unaffected by disease. Contrary to that, Fractional Anisotropy values were shown to decrease in Parkinson's disease patients for connections between the subthalamic nucleus and the pars opercularis of the inferior frontal gyrus, anterior cingulate cortex, the dorsolateral prefrontal cortex and the pre-supplementary motor, collectively involved in preparatory motor control, decision making and task monitoring. While the biological underpinnings of fractional anisotropy alterations remain elusive, they may nonetheless be used as an index of Parkinson's disease. Moreover, we find that failing to account for structural changes occurring in the subthalamic nucleus with age and disease reduce the accuracy and influence the results of tractography, highlighting the importance of using appropriate atlases for tractography.

Dynamic Atlas-Based Segmentation and Quantification of Neuromelanin-Rich Brainstem Structures in Parkinson Disease

We present a dynamic atlas composed of neuromelanin-enhanced magnetic resonance brain images of 40 healthy subjects. The performance of this atlas is evaluated on the fully automated segmentation of two paired neuromelanin-rich brainstem healthy structures: the substantia nigra pars compacta and the locus coeruleus. We show that our dynamic atlas requires in average 60% less images and, therefore, 60% less computation time than a static multi-image atlas while achieving a similar segmentation performance. Then, we show that by applying our dynamic atlas, composed of healthy subjects, to the segmentation and neuromelanin quantification of a set of brain images of 39 Parkinson disease patients, we are able to find significant quantitative differences in the level of neuromelanin between healthy subjects and Parkinson disease patients, thus opening the door to the use of these structures as image biomarkers in future computer aided diagnosis systems for the diagnosis of Parkinson disease.

Automatic localization of the subthalamic nucleus on patient-specific clinical MRI by incorporating 7 T MRI and machine learning: Application in deep brain stimulation

Deep brain stimulation (DBS) of the subthalamic nucleus (STN) has shown clinical potential for relieving the motor symptoms of advanced Parkinson's disease. While accurate localization of the STN is critical for consistent across-patients effective DBS, clear visualization of the STN under standard clinical MR protocols is still challenging. Therefore, intraoperative microelectrode recordings (MER) are incorporated to accurately localize the STN. However, MER require significant neurosurgical expertise and lengthen the surgery time. Recent advances in 7 T MR technology facilitate the ability to clearly visualize the STN. The vast majority of centers, however, still do not have 7 T MRI systems, and fewer have the ability to collect and analyze the data. This work introduces an automatic STN localization framework based on standard clinical MRIs without additional cost in the current DBS planning protocol. Our approach benefits from a large database of 7 T MRI and its clinical MRI pairs. We first model in the 7 T database, using efficient machine learning algorithms, the spatial and geometric dependency between the STN and its adjacent structures (predictors). Given a standard clinical MRI, our method automatically computes the predictors and uses the learned information to predict the patient-specific STN. We validate our proposed method on clinical T2 W MRI of 80 subjects, comparing with experts-segmented STNs from the corresponding 7 T MRI pairs. The experimental results show that our framework provides more accurate and robust patient-specific STN localization than using state-of-the-art atlases. We also demonstrate the clinical feasibility of the proposed technique assessing the post-operative electrode active contact locations.

Warping an atlas derived from serial histology to 5 high-resolution MRIs

Previous work from our group demonstrated the use of multiple input atlases to a modified multi-atlas framework (MAGeT-Brain) to improve subject-based segmentation accuracy. Currently, segmentation of the striatum, globus pallidus and thalamus are generated from a single high-resolution and -contrast MRI atlas derived from annotated serial histological sections. Here, we warp this atlas to five high-resolution MRI templates to create five de novo atlases. The overall goal of this work is to use these newly warped atlases as input to MAGeT-Brain in an effort to consolidate and improve the workflow presented in previous manuscripts from our group, allowing for simultaneous multi-structure segmentation. The work presented details the methodology used for the creation of the atlases using a technique previously proposed, where atlas labels are modified to mimic the intensity and contrast profile of MRI to facilitate atlas-to-template nonlinear transformation estimation. Dice's Kappa metric was used to demonstrate high quality registration and segmentation accuracy of the atlases. The final atlases are available at https://github.com/CobraLab/atlases/tree/master/5-atlas-subcortical.

Seed point discontinuity-based segmentation method for the substantia nigra and the red nucleus in quantitative susceptibility maps

BACKGROUND

The automatic segmentation of cerebral nuclei in the quantitative susceptibility mapping (QSM) images can provide assistance for surgical treatment and pathological mechanism studies. However, as the most frequently used segmentation method, the atlas method provides unsatisfactory results when segmenting the substantia nigra (SN) and the red nucleus (RN).

PURPOSE

To propose and evaluate an improved automatic method based on seed points-discontinuity for segmentations of the SN and the RN in QSM images.

STUDY TYPE

Prospective.

SUBJECTS

In all, 22 subjects, 11 patients with Parkinson's disease (PD), and 11 healthy subjects (mean age of 68.0 ± 6.9 years) underwent MR scans.

FIELD STRENGTH/SEQUENCE

3T system and a 3D multiecho gradient echo sequence with monopolar readout gradient.

ASSESSMENT

Manual segmentations by two radiologists (both with over 10 years of experience in neuroimaging) were used to establish a baseline for assessment. The Dice coefficient and the center-of-gravity distance was employed to evaluate the segmentation accuracy.

STATISTICAL TESTS

The mean value and standard deviation of the Dice coefficient and center-of-gravity distance were calculated separately to compare segmentation results from the proposed method, the level set method, the atlas method (including the single-atlas method and the multi-atlas majority voting method).

RESULTS

The statistical results of Dice coefficient of the SN and the RN between the ground truth and the segmentation were 0.79 ± 0.14 and 0.77 ± 0.06 for the proposed method, 0.40 ± 0.10 and 0.65 ± 0.09 for the level set method, 0.68 ± 0.09 and 0.64 ± 0.07 for the single-atlas method, 0.70 ± 0.06 and 0.68 ± 0.05 for the multi-atlas majority voting method, respectively. The proposed method also provides the lowest center-of-gravity distance value (1.05 ± 0.71 for the SN and 0.74 ± 0.35 for the RN).

DATA CONCLUSION

The segmentation results of the proposed method performed well on the in vivo data and were closer to the manual segmentation than the atlas method.

LEVEL OF EVIDENCE

1 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2018;48:1112-1119.

7T MRI subthalamic nucleus atlas for use with 3T MRI

Deep brain stimulation (DBS) of the subthalamic nucleus (STN) reduces motor symptoms in most patients with Parkinson disease (PD), yet may produce untoward effects. Investigation of DBS effects requires accurate localization of the STN, which can be difficult to identify on magnetic resonance images collected with clinically available 3T scanners. The goal of this study is to develop a high-quality STN atlas that can be applied to standard 3T images. We created a high-definition STN atlas derived from seven older participants imaged at 7T. This atlas was nonlinearly registered to a standard template representing 56 patients with PD imaged at 3T. This process required development of methodology for nonlinear multimodal image registration. We demonstrate mm-scale STN localization accuracy by comparison of our 3T atlas with a publicly available 7T atlas. We also demonstrate less agreement with an earlier histological atlas. STN localization error in the 56 patients imaged at 3T was less than 1 mm on average. Our methodology enables accurate STN localization in individuals imaged at 3T. The STN atlas and underlying 3T average template in MNI space are freely available to the research community. The image registration methodology developed in the course of this work may be generally applicable to other datasets.

A probabilistic template of human mesopontine tegmental nuclei from in vivo 7T MRI

Mesopontine tegmental nuclei such as the cuneiform, pedunculotegmental, oral pontine reticular, paramedian raphe and caudal linear raphe nuclei, are deep brain structures involved in arousal and motor function. Dysfunction of these nuclei is implicated in the pathogenesis of disorders of consciousness and sleep, as well as in neurodegenerative diseases. However, their localization in conventional neuroimages of living humans is difficult due to limited image sensitivity and contrast, and a stereotaxic probabilistic neuroimaging template of these nuclei in humans does not exist. We used semi-automatic segmentation of single-subject 1.1mm-isotropic 7T diffusion-fractional-anisotropy and T₂-weighted images in healthy adults to generate an in vivo probabilistic neuroimaging structural template of these nuclei in standard stereotaxic (Montreal Neurological Institute, MNI) space. The template was validated through independent manual delineation, as well as leave-one-out validation and evaluation of nuclei volumes. This template can enable localization of five mesopontine tegmental nuclei in conventional images (e.g. 1.5T, 3T) in future studies of arousal and motor physiology (e.g. sleep, anesthesia, locomotion) and pathology (e.g. disorders of consciousness, sleep disorders, Parkinson's disease). The 7T magnetic resonance imaging procedure for single-subject delineation of these nuclei may also prove useful for future 7T studies of arousal and motor mechanisms.

T2-Imaging Changes in the Nigrosome-1 Relate to Clinical Measures of Parkinson's Disease

Background

The nigrosome-1 region of the substantia nigra (SN) undergoes the greatest and earliest dopaminergic neuron loss in Parkinson’s disease (PD). As T2-weighted magnetic resonance imaging (MRI) scans are often collected with routine clinical MRI protocols, this investigation aims to determine whether T2-imaging changes in the nigrosome-1 are related to clinical measures of PD and to assess their potential as a more clinically accessible biomarker for PD.

Methods

Voxel intensity ratios were calculated for T2-weighted MRI scans from 47 subjects from the Parkinson’s Progression Markers Initiative database. Three approaches were used to delineate the SN and nigrosome-1: (1) manual segmentation, (2) automated segmentation, and (3) area voxel-based morphometry. Voxel intensity ratios were calculated from voxel intensity values taken from the nigrosome-1 and two areas of the remaining SN. Linear regression analyses were conducted relating voxel intensity ratios with the Movement Disorder Society-Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) sub-scores for each subject.

Results

For manual segmentation, linear regression tests consistently identified the voxel intensity ratio derived from the dorsolateral SN and nigrosome-1 (IR2) as predictive of nBehav (p = 0.0377) and nExp (p = 0.03856). For automated segmentation, linear regression tests identified IR2 as predictive of Subscore IA (nBehav) (p = 0.01134), Subscore IB (nExp) (p = 0.00336), Score II (mExp) (p = 0.02125), and Score III (mSign) (p = 0.008139). For the voxel-based morphometric approach, univariate simple linear regression analysis identified IR2 as yielding significant results for nBehav (p = 0.003102), mExp (p = 0.0172), and mSign (p = 0.00393).

Conclusion

Neuroimaging biomarkers may be used as a proxy of changes in the nigrosome-1, measured by MDS-UPDRS scores as an indicator of the severity of PD. The voxel intensity ratio derived from the dorsolateral SN and nigrosome-1 was consistently predictive of non-motor complex behaviors in all three analyses and predictive of non-motor experiences of daily living, motor experiences of daily living, and motor signs of PD in two of the three analyses. These results suggest that T2 changes in the nigrosome-1 may relate to certain clinical measures of PD. T2 changes in the nigrosome-1 may be considered when developing a more accessible clinical diagnostic tool for patients with suspected PD.

Subthalamic nucleus volumes are highly consistent but decrease age-dependently-a combined magnetic resonance imaging and stereology approach in humans

The subthalamic nucleus (STN) is a main target structure of deep brain stimulation (DBS) in idiopathic Parkinson's disease. Nevertheless, there is an ongoing discussion regarding human STN volumes and neuron count, which could potentially have an impact on STN-DBS. Moreover, a suspected functional subdivision forms the basis of the tripartite hypothesis, which has not yet been morphologically substantiated. In this study, it was aimed to investigate the human STN by means of combined magnetic resonance imaging (MRI) and stereology. STN volumes were obtained from 14 individuals (ranging from 65 to 96 years, 25 hemispheres) in 3 T MRI and in luxol-stained histology slices. Neuron number and cell densities were investigated stereologically over the entire STN and in pre-defined subregions in anti-human neuronal protein HuC/D-stained slices. STN volumes measured with MRI were smaller than in stereology but appeared to be highly consistent, measuring on average 99 ± 6 mm³ (MRI) and 132 ± 20 mm³ (stereology). The neuron count was 431,088 ± 72,172. Both STN volumes and cell count decreased age-dependently. Neuron density was different for the dorsal, medial and ventral subregion with significantly higher values ventrally than dorsally. Small variations in STN volumes in both MRI and stereology contradict previous findings of large variations in STN size. Age-dependent decreases in STN volumes and neuron numbers might influence the efficacy of STN-DBS in a geriatric population. Though the study is limited in sample size, site-dependent differences for the STN subregions form a morphological basis for the tripartite theory. Hum Brain Mapp 38:909-922, 2017. © 2016 Wiley Periodicals, Inc.

Automated segmentation of the substantia nigra, subthalamic nucleus and red nucleus in 7T data at young and old age

With recent developments in MR acquisition at 7T, smaller brainstem structures such as the red nuclei, substantia nigra and subthalamic nuclei can be imaged with good contrast and resolution. These structures have important roles both in the study of the healthy brain and in diseases such as Parkinson's disease, but few methods have been described to automatically segment them. In this paper, we extend a method that we have previously proposed for segmentation of the striatum and globus pallidus to segment these deeper and smaller structures. We modify the method to allow more direct control over segmentation smoothness by using a Markov random field prior. We investigate segmentation performance in three age groups and show that the method produces consistent results that correspond well with manual segmentations. We perform a vertex-based analysis to identify changes with age in the shape of the structures and present results suggesting that the method may be at least as effective as manual delineation in capturing differences between subjects.

Toward an In Vivo Neuroimaging Template of Human Brainstem Nuclei of the Ascending Arousal, Autonomic, and Motor Systems

Brainstem nuclei (Bn) in humans play a crucial role in vital functions, such as arousal, autonomic homeostasis, sensory and motor relay, nociception, sleep, and cranial nerve function, and they have been implicated in a vast array of brain pathologies. However, an in vivo delineation of most human Bn has been elusive because of limited sensitivity and contrast for detecting these small regions using standard neuroimaging methods. To precisely identify several human Bn in vivo, we employed a 7 Tesla scanner equipped with multi-channel receive-coil array, which provided high magnetic resonance imaging sensitivity, and a multi-contrast (diffusion fractional anisotropy and T2-weighted) echo-planar-imaging approach, which provided complementary contrasts for Bn anatomy with matched geometric distortions and resolution. Through a combined examination of 1.3 mm(3) multi-contrast anatomical images acquired in healthy human adults, we semi-automatically generated in vivo probabilistic Bn labels of the ascending arousal (median and dorsal raphe), autonomic (raphe magnus, periaqueductal gray), and motor (inferior olivary nuclei, two subregions of the substantia nigra compatible with pars compacta and pars reticulata, two subregions of the red nucleus, and, in the diencephalon, two subregions of the subthalamic nucleus) systems. These labels constitute a first step toward the development of an in vivo neuroimaging template of Bn in standard space to facilitate future clinical and research investigations of human brainstem function and pathology. Proof-of-concept clinical use of this template is demonstrated in a minimally conscious patient with traumatic brainstem hemorrhages precisely localized to the raphe Bn involved in arousal.

Multi-atlas segmentation of biomedical images: A survey

Multi-atlas segmentation (MAS), first introduced and popularized by the pioneering work of Rohlfing, et al. (2004), Klein, et al. (2005), and Heckemann, et al. (2006), is becoming one of the most widely-used and successful image segmentation techniques in biomedical applications. By manipulating and utilizing the entire dataset of "atlases" (training images that have been previously labeled, e.g., manually by an expert), rather than some model-based average representation, MAS has the flexibility to better capture anatomical variation, thus offering superior segmentation accuracy. This benefit, however, typically comes at a high computational cost. Recent advancements in computer hardware and image processing software have been instrumental in addressing this challenge and facilitated the wide adoption of MAS. Today, MAS has come a long way and the approach includes a wide array of sophisticated algorithms that employ ideas from machine learning, probabilistic modeling, optimization, and computer vision, among other fields. This paper presents a survey of published MAS algorithms and studies that have applied these methods to various biomedical problems. In writing this survey, we have three distinct aims. Our primary goal is to document how MAS was originally conceived, later evolved, and now relates to alternative methods. Second, this paper is intended to be a detailed reference of past research activity in MAS, which now spans over a decade (2003-2014) and entails novel methodological developments and application-specific solutions. Finally, our goal is to also present a perspective on the future of MAS, which, we believe, will be one of the dominant approaches in biomedical image segmentation.

Patch-based label fusion segmentation of brainstem structures with dual-contrast MRI for Parkinson's disease

PURPOSE

Parkinson's disease (PD) is a neurodegenerative disorder that impairs the motor functions. Both surgical treatment and study of PD require delineation of basal ganglia nuclei morphology. While many automatic volumetric segmentation methods have been proposed for the lentiform nucleus, few have attempted to identify the key brainstem substructures including the subthalamic nucleus (STN), substantia nigra (SN), and red nucleus (RN) due to their small size and poor contrast in conventional T1W MRI.

METHODS

A dual-contrast patch-based label fusion method was developed to segment the SN, STN, and RN using multivariate cross-correlation. Two different MRI contrasts (T2*w and phase) are produced from a multi-contrast multi-echo FLASH MRI sequence, enabling visualization of these nuclei. T1-T2* fusion MRI was used to resolve the issue of poor nuclei (i.e., the STN, SN, and RN) contrast on T1w MRI, and to mitigate susceptibility artifacts that may hinder accurate nonlinear registration on T2*w MRI. Unbiased group-wise registration was used for anatomical normalization between the atlas library and the target subject. The performance of the proposed method was compared with a state-of-the-art single-contrast label fusion technique.

RESULTS

The proposed method outperformed a state-of-the-art single-contrast patch-based method in segmenting the STN, RN and SN, and the results were better than those reported in previous literature.

CONCLUSION

Our dual-contrast patch-based label fusion method was superior to a single-contrast method for segmenting brainstem nuclei using a multi-contrast multi-echo FLASH MRI sequence. The method is promising for the treatment and research of Parkinson's disease. This method can be extended for multiple alternative image contrasts and other fields of applications.