An Iterative Centroid Approach for Diffeomorphic Online Atlasing.

Online atlasing, i.e., incrementing an atlas with new images as they are acquired, is key when performing studies on very large, or still being gathered, databases. Regular approaches to atlasing however do not focus on this aspect and impose a complete reconstruction of the atlas when adding images. We propose instead a diffeomorphic online atlasing method that allows gradual updates to an atlas. In this iterative centroid approach, we integrate new subjects in the atlas in an iterative manner, gradually moving the centroid of the images towards its final position. This leads to a computationally cheap approach since it only necessitates one additional registration per new subject added. We validate our approach on several experiments with three main goals: 1- to evaluate atlas image quality of the obtained atlases with sharpness and overlap measures, 2- to assess the deviation in terms of transformations with respect to a conventional atlasing method and 3- to compare its computational time with regular approaches of the literature. We demonstrate that the transformations divergence with respect to a state-of-the-art atlas construction method is small and reaches a plateau, that the two construction methods have the same ability to map subject homologous regions onto a common space and produce images of equivalent quality. The computational time of our approach is also drastically reduced for regular updates. Finally, we also present a direct extension of our method to update spatio-temporal atlases, especially useful for developmental studies.

A Clinically-Compatible Workflow for Computer-Aided Assessment of Brain Disease Activity in Multiple Sclerosis Patients.

Over the last 10 years, the number of approved disease modifying drugs acting on the focal inflammatory process in Multiple Sclerosis (MS) has increased from 3 to 10. This wide choice offers the opportunity of a personalized medicine with the objective of no clinical and radiological activity for each patient. This new paradigm requires the optimization of the detection of new FLAIR lesions on longitudinal MRI. In this paper, we describe a complete workflow-that we developed, implemented, deployed, and evaluated-to facilitate the monitoring of new FLAIR lesions on longitudinal MRI of MS patients. This workflow has been designed to be usable by both hospital and private neurologists and radiologists in France. It consists of three main components: (i) a software component that allows for automated and secured anonymization and transfer of MRI data from the clinical Picture Archive and Communication System (PACS) to a processing server (and vice-versa); (ii) a fully automated segmentation core that enables detection of focal longitudinal changes in patients from T1-weighted, T2-weighted and FLAIR brain MRI scans, and (iii) a dedicated web viewer that provides an intuitive visualization of new lesions to radiologists and neurologists. We first present these different components. Then, we evaluate the workflow on 54 pairs of longitudinal MRI scans that were analyzed by 3 experts (1 neuroradiologist, 1 radiologist, and 1 neurologist) with and without the proposed workflow. We show that our workflow provided a valuable aid to clinicians in detecting new MS lesions both in terms of accuracy (mean number of detected lesions per patient and per expert 1.8 without the workflow vs. 2.3 with the workflow, p = 5.10-4) and of time dedicated by the experts (mean time difference 2'45″, p = 10-4). This increase in the number of detected lesions has implications in the classification of MS patients as stable or active, even for the most experienced neuroradiologist (mean sensitivity was 0.74 without the workflow and 0.90 with the workflow, p-value for no difference = 0.003). It therefore has potential consequences on the therapeutic management of MS patients.

Multiple sclerosis lesions segmentation from multiple experts: The MICCAI 2016 challenge dataset.

MRI plays a crucial role in multiple sclerosis diagnostic and patient follow-up. In particular, the delineation of T2-FLAIR hyperintense lesions is crucial although mostly performed manually - a tedious task. Many methods have thus been proposed to automate this task. However, sufficiently large datasets with a thorough expert manual segmentation are still lacking to evaluate these methods. We present a unique dataset for MS lesions segmentation evaluation. It consists of 53 patients acquired on 4 different scanners with a harmonized protocol. Hyperintense lesions on FLAIR were manually delineated on each patient by 7 experts with control on T2 sequence, and gathered in a consensus segmentation for evaluation. We provide raw and preprocessed data and a split of the dataset into training and testing data, the latter including data from a scanner not present in the training dataset. We strongly believe that this dataset will become a reference in MS lesions segmentation evaluation, allowing to evaluate many aspects: evaluation of performance on unseen scanner, comparison to individual experts performance, comparison to other challengers who already used this dataset, etc.

Multiparametric Analysis of Cerebral Development in Preterm Infants Using Magnetic Resonance Imaging.

OBJECTIVES: The severity of neurocognitive impairment increases with prematurity. However, its mechanisms remain poorly understood. Our aim was firstly to identify multiparametric magnetic resonance imaging (MRI) markers that differ according to the degree of prematurity, and secondly to evaluate the impact of clinical complications on these markers. MATERIALS AND METHODS: We prospectively enrolled preterm infants who were divided into two groups according to their degree of prematurity: extremely preterm (<28 weeks' gestational age) and very preterm (28-32 weeks' gestational age). They underwent a multiparametric brain MRI scan at term-equivalent age including morphological, diffusion tensor and arterial spin labeling (ASL) perfusion sequences. We quantified overall and regional volumes, diffusion parameters, and cerebral blood flow (CBF). We then compared the parameters for the two groups. We also assessed the effects of clinical data and potential MRI morphological abnormalities on those parameters. RESULTS: Thirty-four preterm infants were included. Extremely preterm infants (n = 13) had significantly higher frontal relative volumes (p = 0.04), frontal GM relative volumes (p = 0.03), and regional CBF than very preterm infants, but they had lower brainstem and insular relative volumes (respectively p = 0.008 and 0.04). Preterm infants with WM lesions on MRI had significantly lower overall GM CBF (13.3 ± 2 ml/100 g/min versus 17.7 ± 2.5, < ml/100 g/min p = 0.03). CONCLUSION: Magnetic resonance imaging brain scans performed at term-equivalent age in preterm infants provide quantitative imaging parameters that differ with respect to the degree of prematurity, related to brain maturation.

Interpolation and Averaging of Diffusion MRI Multi-Compartment Models.

Multi-compartment models (MCM) are increasingly used to characterize the brain white matter microstructure from diffusion-weighted imaging (DWI). Their use in clinical studies is however limited by the inability to resample an MCM image towards a common reference frame, or to construct atlases from such brain microstructure models. We propose to solve this problem by first identifying that these two tasks amount to the same problem. We propose to tackle it by viewing it as a simplification problem, solved thanks to spectral clustering and the definition of semi-metrics between several usual compartments encountered in the MCM literature. This generic framework is evaluated for two models: the multi-tensor model where individual fibers are modeled as individual tensors and the diffusion direction imaging (DDI) model that differentiates intra- and extra-axonal components of each fiber. Results on simulated data, simulated transformations and real data show the ability of our method to well interpolate MCM images of these types. We finally present as an application an MCM template of normal controls constructed using our approach.

Unsupervised Domain Adaptation With Optimal Transport in Multi-Site Segmentation of Multiple Sclerosis Lesions From MRI Data.

Automatic segmentation of Multiple Sclerosis (MS) lesions from Magnetic Resonance Imaging (MRI) images is essential for clinical assessment and treatment planning of MS. Recent years have seen an increasing use of Convolutional Neural Networks (CNNs) for this task. Although these methods provide accurate segmentation, their applicability in clinical settings remains limited due to a reproducibility issue across different image domains. MS images can have highly variable characteristics across patients, MRI scanners and imaging protocols; retraining a supervised model with data from each new domain is not a feasible solution because it requires manual annotation from expert radiologists. In this work, we explore an unsupervised solution to the problem of domain shift. We present a framework, Seg-JDOT, which adapts a deep model so that samples from a source domain and samples from a target domain sharing similar representations will be similarly segmented. We evaluated the framework on a multi-site dataset, MICCAI 2016, and showed that the adaptation toward a target site can bring remarkable improvements in a model performance over standard training.

Regional brain development analysis through registration using anisotropic similarity, a constrained affine transformation.

We propose a novel method to quantify brain growth in 3 arbitrary orthogonal directions of the brain or its sub-regions through linear registration. This is achieved by introducing a 9 degrees of freedom (dof) transformation called anisotropic similarity which is an affine transformation with constrained scaling directions along arbitrarily chosen orthogonal vectors. This gives the opportunity to extract scaling factors describing brain growth along those directions by registering a database of subjects onto a common reference. This information about directional growth brings insights that are not usually available in longitudinal volumetric analysis. The interest of this method is illustrated by studying the anisotropic regional and global brain development of 308 healthy subjects betwen 0 and 19 years old. A gender comparison of those scaling factors is also performed for four age-intervals. We demonstrate through these applications the stability of the method to the chosen reference and its ability to highlight growth differences accros regions and gender.

White matter abnormalities in depression: A categorical and phenotypic diffusion MRI study.

Mood depressive disorder is one of the most disabling chronic diseases with a high rate of everyday life disability that affects 350 million people around the world. Recent advances in neuroimaging have reported widespread structural abnormalities, suggesting a dysfunctional frontal-limbic circuit involved in the pathophysiological mechanisms of depression. However, a variety of different white matter regions has been implicated and is sought to suffer from lack of reproducibility of such categorical-based biomarkers. These inconsistent results might be attributed to various factors: actual categorical definition of depression as well as clinical phenotype variability. In this study, we 1/ examined WM changes in a large cohort (114 patients) compared to a healthy control group and 2/ sought to identify specific WM alterations in relation to specific depressive phenotypes such as anhedonia (i.e. lack of pleasure), anxiety and psychomotor retardation -three core symptoms involved in depression. Consistent with previous studies, reduced white matter was observed in the genu of the corpus callosum extending to the inferior fasciculus and posterior thalamic radiation, confirming a frontal-limbic circuit abnormality. Our analysis also reported other patterns of increased fractional anisotropy and axial diffusivity as well as decreased apparent diffusion coefficient and radial diffusivity in the splenium of the corpus callosum and posterior limb of the internal capsule. Moreover, a positive correlation between FA and anhedonia was found in the superior longitudinal fasciculus as well as a negative correlation in the cingulum. Then, the analysis of the anxiety and diffusion metric revealed that increased anxiety was associated with greater FA values in genu and splenium of corpus callosum, anterior corona radiata and posterior thalamic radiation. Finally, the motor retardation analysis showed a correlation between increased Widlöcher depressive retardation scale scores and reduced FA in the body and genu of the corpus callosum, fornix, and superior striatum. Through this twofold approach (categorical and phenotypic), this study has underlined the need to move forward to a symptom-based research area of biomarkers, which help to understand the pathophysiology of mood depressive disorders and to stratify precise phenotypes of depression with targeted therapeutic strategies.

Objective Evaluation of Multiple Sclerosis Lesion Segmentation using a Data Management and Processing Infrastructure.

We present a study of multiple sclerosis segmentation algorithms conducted at the international MICCAI 2016 challenge. This challenge was operated using a new open-science computing infrastructure. This allowed for the automatic and independent evaluation of a large range of algorithms in a fair and completely automatic manner. This computing infrastructure was used to evaluate thirteen methods of MS lesions segmentation, exploring a broad range of state-of-theart algorithms, against a high-quality database of 53 MS cases coming from four centers following a common definition of the acquisition protocol. Each case was annotated manually by an unprecedented number of seven different experts. Results of the challenge highlighted that automatic algorithms, including the recent machine learning methods (random forests, deep learning, …), are still trailing human expertise on both detection and delineation criteria. In addition, we demonstrate that computing a statistically robust consensus of the algorithms performs closer to human expertise on one score (segmentation) although still trailing on detection scores.

Measurement of pediatric regional cerebral blood flow from 6 months to 15 years of age in a clinical population.

OBJECTIVES: To investigate changes in cerebral blood flow (CBF) in gray matter (GM) between 6 months and 15 years of age and to provide CBF values for the brain, GM, white matter (WM), hemispheres and lobes. METHODS: Between 2013 and 2016, we retrospectively included all clinical MRI examinations with arterial spin labeling (ASL). We excluded subjects with a condition potentially affecting brain perfusion. For each subject, mean values of CBF in the brain, GM, WM, hemispheres and lobes were calculated. GM CBF was fitted using linear, quadratic and cubic polynomial regression against age. Regression models were compared with Akaike's information criterion (AIC), and Likelihood Ratio tests. RESULTS: 84 children were included (44 females/40 males). Mean CBF values were 64.2 ±â€¯13.8 mL/100 g/min in GM, and 29.3 ±â€¯10.0 mL/100 g/min in WM. The best-fit model of brain perfusion was the cubic polynomial function (AIC = 672.7, versus respectively AIC = 673.9 and AIC = 674.1 with the linear negative function and the quadratic polynomial function). A statistically significant difference between the tested models demonstrating the superiority of the quadratic (p = 0.18) or cubic polynomial model (p = 0.06), over the negative linear regression model was not found. No effect of general anesthesia (p = 0.34) or of gender (p = 0.16) was found. CONCLUSION: we provided values for ASL CBF in the brain, GM, WM, hemispheres, and lobes over a wide pediatric age range, approximately showing inverted U-shaped changes in GM perfusion over the course of childhood.

USPIO-positive MS lesions are associated with greater tissue damage than gadolinium-positive-only lesions during 3-year follow-up.

BACKGROUND: Identifying in vivo the processes that determine lesion severity in multiple sclerosis (MS) remains a challenge. OBJECTIVES: To describe the dynamics of ultrasmall superparamagnetic iron oxide (USPIO) enhancement in MS lesions and the relationship between USPIO enhancement and microstructural changes over 3 years. METHODS: Lesion development was assessed at baseline, Months 3, 6, and 9, using magnetic resonance imaging (MRI) with gadolinium and USPIO. Microstructural changes were assessed at baseline, Months 3, 6, 9, 12, 18, 24, and 36, using relaxometry, magnetization transfer, and diffusion-weighted imaging. RESULTS: We included 15 patients with clinically isolated syndrome. In the 52 MRI scans acquired with USPIO, 22 lesions were USPIO and gadolinium positive, and 44 were USPIO negative but gadolinium positive. Lesions no longer exhibited sustained USPIO enhancement 3 months later. At baseline, lesions that were both USPIO and gadolinium positive had lower magnetization transfer ratio values (common language effect size = 0.84, p = 0.0005) and lower fractional anisotropy values (0.83, p = 0.001) than gadolinium-positive-only lesions. USPIO-positive lesions remained associated with greater damage than gadolinium-positive-only lesions throughout the 3-year follow-up. CONCLUSION: USPIO enhancement, mainly reflecting monocyte infiltration, is transient and is associated with persistent tissue damage after 3 years.

Block-Matching Distortion Correction of Echo-Planar Images With Opposite Phase Encoding Directions.

By shortening the acquisition time of MRI, Echo Planar Imaging (EPI) enables the acquisition of a large number of images in a short time, compatible with clinical constraints as required for diffusion or functional MRI. However such images are subject to large, local distortions disrupting their correspondence with the underlying anatomy. The correction of those distortions is an open problem, especially in regions where large deformations occur. We propose a new block-matching registration method to perform EPI distortion correction based on the acquisition of two EPI with opposite phase encoding directions (PED). It relies on new transformations between blocks adapted to the EPI distortion model, and on an adapted optimization scheme to ensure an opposite symmetric transformation. We present qualitative and quantitative results of the block-matching correction using different metrics on a phantom dataset and on in-vivo data. We show the ability of the block-matching to robustly correct EPI distortion even in strongly affected areas.

Longitudinal multiple sclerosis lesion segmentation: Resource and challenge.

In conjunction with the ISBI 2015 conference, we organized a longitudinal lesion segmentation challenge providing training and test data to registered participants. The training data consisted of five subjects with a mean of 4.4 time-points, and test data of fourteen subjects with a mean of 4.4 time-points. All 82 data sets had the white matter lesions associated with multiple sclerosis delineated by two human expert raters. Eleven teams submitted results using state-of-the-art lesion segmentation algorithms to the challenge, with ten teams presenting their results at the conference. We present a quantitative evaluation comparing the consistency of the two raters as well as exploring the performance of the eleven submitted results in addition to three other lesion segmentation algorithms. The challenge presented three unique opportunities: (1) the sharing of a rich data set; (2) collaboration and comparison of the various avenues of research being pursued in the community; and (3) a review and refinement of the evaluation metrics currently in use. We report on the performance of the challenge participants, as well as the construction and evaluation of a consensus delineation. The image data and manual delineations will continue to be available for download, through an evaluation website2 as a resource for future researchers in the area. This data resource provides a platform to compare existing methods in a fair and consistent manner to each other and multiple manual raters.

Imaging biomarkers in multiple Sclerosis: From image analysis to population imaging.

The production of imaging data in medicine increases more rapidly than the capacity of computing models to extract information from it. The grand challenges of better understanding the brain, offering better care for neurological disorders, and stimulating new drug design will not be achieved without significant advances in computational neuroscience. The road to success is to develop a new, generic, computational methodology and to confront and validate this methodology on relevant diseases with adapted computational infrastructures. This new concept sustains the need to build new research paradigms to better understand the natural history of the pathology at the early phase; to better aggregate data that will provide the most complete representation of the pathology in order to better correlate imaging with other relevant features such as clinical, biological or genetic data. In this context, one of the major challenges of neuroimaging in clinical neurosciences is to detect quantitative signs of pathological evolution as early as possible to prevent disease progression, evaluate therapeutic protocols or even better understand and model the natural history of a given neurological pathology. Many diseases encompass brain alterations often not visible on conventional MRI sequences, especially in normal appearing brain tissues (NABT). MRI has often a low specificity for differentiating between possible pathological changes which could help in discriminating between the different pathological stages or grades. The objective of medical image analysis procedures is to define new quantitative neuroimaging biomarkers to track the evolution of the pathology at different levels. This paper illustrates this issue in one acute neuro-inflammatory pathology: Multiple Sclerosis (MS). It exhibits the current medical image analysis approaches and explains how this field of research will evolve in the next decade to integrate larger scale of information at the temporal, cellular, structural and morphological levels.

The DTI Challenge: Toward Standardized Evaluation of Diffusion Tensor Imaging Tractography for Neurosurgery.

BACKGROUND AND PURPOSE: Diffusion tensor imaging (DTI) tractography reconstruction of white matter pathways can help guide brain tumor resection. However, DTI tracts are complex mathematical objects and the validity of tractography-derived information in clinical settings has yet to be fully established. To address this issue, we initiated the DTI Challenge, an international working group of clinicians and scientists whose goal was to provide standardized evaluation of tractography methods for neurosurgery. The purpose of this empirical study was to evaluate different tractography techniques in the first DTI Challenge workshop. METHODS: Eight international teams from leading institutions reconstructed the pyramidal tract in four neurosurgical cases presenting with a glioma near the motor cortex. Tractography methods included deterministic, probabilistic, filtered, and global approaches. Standardized evaluation of the tracts consisted in the qualitative review of the pyramidal pathways by a panel of neurosurgeons and DTI experts and the quantitative evaluation of the degree of agreement among methods. RESULTS: The evaluation of tractography reconstructions showed a great interalgorithm variability. Although most methods found projections of the pyramidal tract from the medial portion of the motor strip, only a few algorithms could trace the lateral projections from the hand, face, and tongue area. In addition, the structure of disagreement among methods was similar across hemispheres despite the anatomical distortions caused by pathological tissues. CONCLUSIONS: The DTI Challenge provides a benchmark for the standardized evaluation of tractography methods on neurosurgical data. This study suggests that there are still limitations to the clinical use of tractography for neurosurgical decision making.

Diffusion MRI abnormalities detection with orientation distribution functions: a multiple sclerosis longitudinal study.

We propose a new algorithm for the voxelwise analysis of orientation distribution functions between one image and a group of reference images. It relies on a generic framework for the comparison of diffusion probabilities on the sphere, sampled from the underlying models. We demonstrate that this method, combined to dimensionality reduction through a principal component analysis, allows for more robust detection of lesions on simulated data when compared to classical tensor-based analysis. We then demonstrate the efficiency of this pipeline on the longitudinal comparison of multiple sclerosis patients at an early stage of the disease: right after their first clinically isolated syndrome (CIS) and three months later. We demonstrate the predictive value of ODF-based scores for the early detection of lesions that will appear or heal.

Predictive value of imaging markers at multiple sclerosis disease onset based on gadolinium- and USPIO-enhanced MRI and machine learning.

OBJECTIVES: A novel characterization of Clinically Isolated Syndrome (CIS) patients according to lesion patterns is proposed. More specifically, patients are classified according to the nature of inflammatory lesions patterns. It is expected that this characterization can infer new prospective figures from the earliest imaging signs of Multiple Sclerosis (MS), since it can provide a classification of different types of lesions across patients. METHODS: The method is based on a two-tiered classification. Initially, the spatio-temporal lesion patterns are classified. The discovered lesion patterns are then used to characterize groups of patients. The patient groups are validated using statistical measures and by correlations at 24-month follow-up with hypointense lesion loads. RESULTS: The methodology identified 3 statistically significantly different clusters of lesion patterns showing p-values smaller than 0.01. Moreover, these patterns defined at baseline correlated with chronic hypointense lesion volumes by follow-up with an R(2) score of 0.90. CONCLUSIONS: The proposed methodology is capable of identifying three major different lesion patterns that are heterogeneously present in patients, allowing a patient classification using only two MRI scans. This finding may lead to more accurate prognosis and thus to more suitable treatments at early stage of MS.

A mathematical framework for the registration and analysis of multi-fascicle models for population studies of the brain microstructure.

Diffusion tensor imaging (DTI) is unable to represent the diffusion signal arising from multiple crossing fascicles and freely diffusing water molecules. Generative models of the diffusion signal, such as multi-fascicle models, overcome this limitation by providing a parametric representation for the signal contribution of each population of water molecules. These models are of great interest in population studies to characterize and compare the brain microstructural properties. Central to population studies is the construction of an atlas and the registration of all subjects to it. However, the appropriate definition of registration and atlasing methods for multi-fascicle models have proven challenging. This paper proposes a mathematical framework to register and analyze multi-fascicle models. Specifically, we define novel operators to achieve interpolation, smoothing and averaging of multi-fascicle models. We also define a novel similarity metric to spatially align multi-fascicle models. Our framework enables simultaneous comparisons of different microstructural properties that are confounded in conventional DTI. The framework is validated on multi-fascicle models from 24 healthy subjects and 38 patients with tuberous sclerosis complex, 10 of whom have autism. We demonstrate the use of the multi-fascicle models registration and analysis framework in a population study of autism spectrum disorder.

Adaptive multi-modal particle filtering for probabilistic white matter tractography.

Particle filtering has recently been introduced to perform probabilistic tractography in conjunction with DTI and Q-Ball models to estimate the diffusion information. Particle filters are particularly well adapted to the tractography problem as they offer a way to approximate a probability distribution over all paths originated from a specified voxel, given the diffusion information. In practice however, they often fail at consistently capturing the multi-modality of the target distribution. For brain white matter tractography, this means that multiple fiber pathways are unlikely to be tracked over extended volumes. We propose to remedy this issue by formulating the filtering distribution as an adaptive M-component non-parametric mixture model. Such a formulation preserves all the properties of a classical particle filter while improving multi-modality capture. We apply this multi-modal particle filter to both DTI and Q-Ball models and propose to estimate dynamically the number of modes of the filtering distribution. We show on synthetic and real data how this algorithm outperforms the previous versions proposed in the literature.