Annemie Ribbens

Assessing age-related gray matter decline with voxel-based morphometry depends significantly on segmentation and normalization procedures

Healthy ageing coincides with a progressive decline of brain gray matter (GM) ultimately affecting the entire brain. For a long time, manual delineation-based volumetry within predefined regions of interest (ROI) has been the gold standard for assessing such degeneration. Voxel-Based Morphometry (VBM) offers an automated alternative approach that, however, relies critically on the segmentation and spatial normalization of a large collection of images from different subjects. This can be achieved via different algorithms, with SPM5/SPM8, DARTEL of SPM8 and FSL tools (FAST, FNIRT) being three of the most frequently used. We complemented these voxel based measurements with a ROI based approach, whereby the ROIs are defined by transforms of an atlas (containing different tissue probability maps as well as predefined anatomic labels) to the individual subject images in order to obtain volumetric information at the level of the whole brain or within separate ROIs. Comparing GM decline between 21 young subjects (mean age 23) and 18 elderly (mean age 66) revealed that volumetric measurements differed significantly between methods. The unified segmentation/normalization of SPM5/SPM8 revealed the largest age-related differences and DARTEL the smallest, with FSL being more similar to the DARTEL approach. Method specific differences were substantial after segmentation and most pronounced for the cortical structures in close vicinity to major sulci and fissures. Our findings suggest that algorithms that provide only limited degrees of freedom for local deformations (such as the unified segmentation and normalization of SPM5/SPM8) tend to overestimate between-group differences in VBM results when compared to methods providing more flexible warping. This difference seems to be most pronounced if the anatomy of one of the groups deviates from custom templates, a finding that is of particular importance when results are compared across studies using different VBM methods.

Reliable measurements of brain atrophy in individual patients with multiple sclerosis

As neurodegeneration is recognized as a major contributor to disability in multiple sclerosis (MS), brain atrophy quantification could have a high added value in clinical practice to assess treatment efficacy and disease progression, provided that it has a sufficiently low measurement error to draw meaningful conclusions for an individual patient.

Unsupervised Segmentation, Clustering, and Groupwise Registration of Heterogeneous Populations of Brain MR Images

Population analysis of brain morphology from magnetic resonance images contributes to the study and understanding of neurological diseases. Such analysis typically involves segmentation of a large set of images and comparisons of these segmentations between relevant subgroups of images (e.g., “normal” versus “diseased”). The images of each subgroup are usually selected in advance in a supervised way based on clinical knowledge. Their segmentations are typically guided by one or more available atlases, assumed to be suitable for the images at hand. We present a data-driven probabilistic framework that simultaneously performs atlas-guided segmentation of a heterogeneous set of brain MR images and clusters the images in homogeneous subgroups, while constructing separate probabilistic atlases for each cluster to guide the segmentation. The main benefits of integrating segmentation, clustering and atlas construction in a single framework are that: 1) our method can handle images of a heterogeneous group of subjects and automatically identifies homogeneous subgroups in an unsupervised way with minimal prior knowledge, 2) the subgroups are formed by automatical detection of the relevant morphological features based on the segmentation, 3) the atlases used by our method are constructed from the images themselves and optimally adapted for guiding the segmentation of each subgroup, and 4) the probabilistic atlases represent the morphological pattern that is specific for each subgroup and expose the groupwise differences between different subgroups. We demonstrate the feasibility of the proposed framework and evaluate its performance with respect to image segmentation, clustering and atlas construction on simulated and real data sets including the publicly available BrainWeb and ADNI data. It is shown that combined segmentation and atlas construction leads to improved segmentation accuracy. Furthermore, it is demonstrated that the clusters generated by our unsupervised framework largely coincide with the clinically determined subgroups in case of disease-specific differences in brain morphology and that the differences between the cluster-specific atlases are in agreement with the expected disease-specific patterns, indicating that our method is capable of detecting the different modes in a population. Our method can thus be seen as a comprehensive image-driven population analysis framework that can contribute to the detection of novel subgroups and distinctive image features, potentially leading to new insights in the brain development and disease.

SPARC: Unified framework for automatic segmentation, probabilistic atlas construction, registration and clustering of brain MR images

Automated methods for image segmentation, image registration, clustering of images and probabilistic atlas construction are of great interest in medical image analysis. In this work, we propose a model where these different aspects are combined in one comprehensive probabilistic framework. The framework is formulated as an EM optimization algorithm. Validation is performed on simulated and real images in terms of segmentation, clustering and atlas construction. Accurate segmentations are obtained and the different modes in a population of normal controls and Huntington disease patients are discovered. Furthermore, our method reveals the localization of cluster specific morphological differences for each image in the population.

Two Time Point MS Lesion Segmentation in Brain MRI : An Expectation-Maximization Framework

Purpose: Lesion volume is a meaningful measure in multiple sclerosis (MS) prognosis. Manual lesion segmentation for computing volume in a single or multiple time points is time consuming and suffers from intra and inter-observer variability. Methods: In this paper, we present MSmetrix-long: a joint expectation-maximization (EM) framework for two time point white matter (WM) lesion segmentation. MSmetrix-long takes as input a 3D T1-weighted and a 3D FLAIR MR image and segments lesions in three steps: (1) cross-sectional lesion segmentation of the two time points; (2) creation of difference image, which is used to model the lesion evolution; (3) a joint EM lesion segmentation framework that uses output of step (1) and step (2) to provide the final lesion segmentation. The accuracy (Dice score) and reproducibility (absolute lesion volume difference) of MSmetrix-long is evaluated using two datasets. Results: On the first dataset, the median Dice score between MSmetrix-long and expert lesion segmentation was 0.63 and the Pearson correlation coefficient (PCC) was equal to 0.96. On the second dataset, the median absolute volume difference was 0.11 ml. Conclusions: MSmetrix-long is accurate and consistent in segmenting MS lesions. Also, MSmetrix-long compares favorably with the publicly available longitudinal MS lesion segmentation algorithm of Lesion Segmentation Toolbox.

Semisupervised probabilistic clustering of brain MR images including prior clinical information

Accurate morphologic clustering of subjects and detection of population specific differences in brain MR images, due to e.g. neurological diseases, is of great interest in medical image analysis. In previous work, we proposed a probabilistic framework for unsupervised image clustering that allows exposing cluster specific morphological differences in each image. In this paper, we extend this framework to also accommodate semisupervised clustering approaches which provides the possibility of including prior knowledge about cluster memberships, group-level morphological differences and clinical prior knowledge. The method is validated on three different data sets and a comparative study between the supervised, semisupervised and unsupervised methods is performed. We show that the use of a limited amount of prior knowledge about cluster memberships can contribute to a better clustering performance in certain applications, while on the other hand the semisupervised clustering is quite robust to incorrect prior clustering knowledge.

HLA genotype as a marker of multiple sclerosis prognosis: A pilot study

OBJECTIVE The identification of a biomarker with prognostic value is an unmet need in multiple sclerosis (MS). The objective of this study was to investigate a possible association of HLA genotype with disease status and progression in MS, based on comprehensive and sensitive clinical and magnetic resonance imaging (MRI) parameters to measure disease effects. METHOD A total of 118 MS patients (79 females, 39 males) underwent HLA typing. Patient MS status was assessed at two time points in a 2-year interval, based on clinical scores (including EDSS, MSSS, T25FW, 9-HPT, SDMT, BVMT, CVLT-II) and MRI evaluations. Quantitative brain MRI values were obtained for whole brain atrophy, FLAIR lesion volume change and number of new lesions using MSmetrix. Predefined HLA patient groups were compared as of disease status and progression. Global assessment was achieved by an overall t-statistic and assessment per measurement by a Welch test and/or Mann Whitney U test. The effects of multiple covariates, including age, gender and disease duration as well as scan parameters, were also evaluated using a regression analysis. RESULTS The HLA-A*02 allele was associated with better outcomes in terms of MSSS, EDSS and new lesion count (Welch test p-value<0.05). The HLA-B*07 and HLA-B*44 alleles showed a global negative effect on disease status, although none of the measurements reached significance (p-value<0.05). Results for the HLA-DRB1*15, HLA-DQB1*06 and HLA-B*08 alleles were inconclusive. The influence of the confounding variables on the statistical analysis was limited.

Bayesian and grAphical Models for Biomedical Imaging

Although N3 is perhaps the most widely used method for MRI bias field correction, its underlying mechanism is in fact not well understood. Specifically, the method relies on a relatively heuristic recipe of alternating iterative steps that does not optimize any particular objective function. In this paper we explain the successful bias field correction properties of N3 by showing that it implicitly uses the same generative models and computational strategies as expectation maximization (EM) based bias field correction methods. We demonstrate experimentally that purely EM-based methods are capable of producing bias field correction results comparable to those of N3 in less computation time.

005 Filling in the gaps: precision MRI reporting in multiple sclerosis clinical practice

Introduction Clinical multiple sclerosis (MS) magnetic resonance imaging (MRI) brain reports provide important information to neurologists. The quantitative data reported varies between centres and radiologists. Structured MRI reporting and formal quantitative MRI (QMRI) analysis may assist clinicians with patient management. The objective is to compare quantitative data derived from standard clinical reports, structured neuroradiologist reviews, local QMRI analyses and fully-automated MSmetrix QMRI analyses, in a longitudinal clinical MS cohort. Methods Clinical brain MRI scans separated by one-year minimum, from the same patient on the same scanner, were evaluated. Quantitative information was extracted from the clinical reports and structured neuroradiologist reviews. MRI scan-pairs were analysed locally by imaging-analysts and centrally by MSmetrix. Results 50 MS patients, baseline age 39.02 (9.06) years, disease duration 9.11 (6.88) years and Expanded Disability Status Scale score 1.91 (1.62), were included. Compared to clinical reports, structured neuroradiologist reviews provided increased semi-/quantitative data; baseline T2 and T1 lesion burden (50% vs 100%; 2% vs 100%), baseline brain volume-loss (BVL; 72% vs 100%), new T1 lesions (0% vs 100%), regional brain atrophy (BA; 20% vs 100%). Lesion and brain volumes were not provided in radiology reports/reviews. Comparison of local and central QMRI revealed moderate-strong Pearson correlations for most metrics; Intra-class correlations varied more widely. Statistical consistency existed across all methods in detecting new T2 lesions. Radiologist-identified baseline BVL was associated with lower quantitatively-measured brain volumes. Local QMRI longitudinal BA rates >0.4% and >0.8%, were 48% and 26% respectively. Neuroradiologist review identified BA in 12%. Conclusion Structured neuroradiology review provided additional quantitative information over standard clinical reports. Quantitative data measured using local and MSmetrix pipelines were generally well associated but are not interchangeable. Longitudinal whole brain and regional atrophy is not reliably identified by radiologists and QMRI analysis provides a clear advantage in this regard. Structured reporting, and formal QMRI analysis, provide additional quantitative MRI data that may assist clinical decision-making in MS.

Targeting phosphocreatine metabolism in relapsing–remitting multiple sclerosis: evaluation with brain MRI, 1 H and 31 P MRS, and clinical and cognitive testing

Background/objectivesFluoxetine and prucalopride might change phosphocreatine (PCr) levels via the cAMP–PKA pathway, an interesting target in the neurodegenerative mechanisms of MS.MethodsWe conducted a two-center double-blind, placebo-controlled, randomized trial including 48 relapsing–remitting MS patients. Patients were randomized to receive placebo (n = 13), fluoxetine (n = 15), or prucalopride (n = 14) for 6 weeks. Proton (1H) and phosphorus (31P) magnetic resonance spectroscopy (MRS) as well as volumetric and perfusion MR imaging were performed at weeks 0, 2, and 6. Clinical and cognitive testing were evaluated at weeks 0 and 6.ResultsNo significant changes were observed for both 31P and 1H MRS indices. We found a significant effect on white matter volume and a trend towards an increase in grey matter and whole brain volume in the fluoxetine group at week 2; however, these effects were not sustained at week 6 for white matter and whole brain volume. Fluoxetine and prucalopride showed a positive effect on 9-HPT, depression, and fatigue scores.ConclusionBoth fluoxetine and prucalopride had a symptomatic effect on upper limb function, fatigue, and depression, but this should be interpreted with caution. No effect of treatment was found on 31P and 1H MRS parameters, suggesting that these molecules do not influence the PCr metabolism.