Ferran Prados

Longitudinal multiple sclerosis lesion segmentation: Resource and challenge

Abstract 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. HighlightsPublic lesion data base of 21 training data sets and 61 testing data sets.Fully automated evaluation website.Comparison between 14 state‐of‐the‐art algorithms and 2 manual delineators.

Spinal cord grey matter segmentation challenge

ABSTRACT An important image processing step in spinal cord magnetic resonance imaging is the ability to reliably and accurately segment grey and white matter for tissue specific analysis. There are several semi‐ or fully‐automated segmentation methods for cervical cord cross‐sectional area measurement with an excellent performance close or equal to the manual segmentation. However, grey matter segmentation is still challenging due to small cross‐sectional size and shape, and active research is being conducted by several groups around the world in this field. Therefore a grey matter spinal cord segmentation challenge was organised to test different capabilities of various methods using the same multi‐centre and multi‐vendor dataset acquired with distinct 3D gradient‐echo sequences. This challenge aimed to characterize the state‐of‐the‐art in the field as well as identifying new opportunities for future improvements. Six different spinal cord grey matter segmentation methods developed independently by various research groups across the world and their performance were compared to manual segmentation outcomes, the present gold‐standard. All algorithms provided good overall results for detecting the grey matter butterfly, albeit with variable performance in certain quality‐of‐segmentation metrics. The data have been made publicly available and the challenge web site remains open to new submissions. No modifications were introduced to any of the presented methods as a result of this challenge for the purposes of this publication. HighlightsFirst grey matter spinal cord segmentation challenge.Six institutions participated in the challenge and compared their methods.Public available dataset from multiple vendors and sites.The challenge web site remains open to new submissions.

Measuring brain atrophy with a generalized formulation of the boundary shift integral

Brain atrophy measured using structural magnetic resonance imaging (MRI) has been widely used as an imaging biomarker for disease diagnosis and tracking of pathologic progression in neurodegenerative diseases. In this work, we present a generalized and extended formulation of the boundary shift integral (gBSI) using probabilistic segmentations to estimate anatomic changes between 2 time points. This method adaptively estimates a non-binary exclusive OR region of interest from probabilistic brain segmentations of the baseline and repeat scans to better localize and capture the brain atrophy. We evaluate the proposed method by comparing the sample size requirements for a hypothetical clinical trial of Alzheimers disease to that needed for the current implementation of BSI as well as a fuzzy implementation of BSI. The gBSI method results in a modest but reduced sample size, providing increased sensitivity to disease changes through the use of the probabilistic exclusive OR region.

Fully automated grey and white matter spinal cord segmentation

Axonal loss in the spinal cord is one of the main contributing factors to irreversible clinical disability in multiple sclerosis (MS). In vivo axonal loss can be assessed indirectly by estimating a reduction in the cervical cross-sectional area (CSA) of the spinal cord over time, which is indicative of spinal cord atrophy, and such a measure may be obtained by means of image segmentation using magnetic resonance imaging (MRI). In this work, we propose a new fully automated spinal cord segmentation technique that incorporates two different multi-atlas segmentation propagation and fusion techniques: The Optimized PatchMatch Label fusion (OPAL) algorithm for localising and approximately segmenting the spinal cord, and the Similarity and Truth Estimation for Propagated Segmentations (STEPS) algorithm for segmenting white and grey matter simultaneously. In a retrospective analysis of MRI data, the proposed method facilitated CSA measurements with accuracy equivalent to the inter-rater variability, with a Dice score (DSC) of 0.967 at C2/C3 level. The segmentation performance for grey matter at C2/C3 level was close to inter-rater variability, reaching an accuracy (DSC) of 0.826 for healthy subjects and 0.835 people with clinically isolated syndrome MS.

Relationship of grey and white matter abnormalities with distance from the surface of the brain in multiple sclerosis

Objective To assess the association between proximity to the inner (ventricular and aqueductal) and outer (pial) surfaces of the brain and the distribution of normal appearing white matter (NAWM) and grey matter (GM) abnormalities, and white matter (WM) lesions, in multiple sclerosis (MS). Methods 67 people with relapse-onset MS and 30 healthy controls were included in the study. Volumetric T1 images and high-resolution (1 mm3) magnetisation transfer ratio (MTR) images were acquired and segmented into 12 bands between the inner and outer surfaces of the brain. The first and last bands were discarded to limit partial volume effects with cerebrospinal fluid. MTR values were computed for all bands in supratentorial NAWM, cerebellar NAWM and brainstem NA tissue, and deep and cortical GM. Band WM lesion volumes were also measured. Results Proximity to the ventricular surfaces was associated with progressively lower MTR values in the MS group but not in controls in supratentorial and cerebellar NAWM, brainstem NA and in deep and cortical GM. The density of WM lesions was associated with proximity to the ventricles only in the supratentorial compartment, and no link was found with distance from the pial surfaces. Conclusions In MS, MTR abnormalities in NAWM and GM are related to distance from the inner and outer surfaces of the brain, and this suggests that there is a common factor underlying their spatial distribution. A similar pattern was not found for WM lesions, raising the possibility that different factors promote their formation.

Spinal cord atrophy as a primary outcome measure in phase II trials of progressive multiple sclerosis

Objectives: To measure the development of spinal cord (SC) atrophy over 1 year in patients with progressive multiple sclerosis (PMS) and determine the sample sizes required to demonstrate a reduction in spinal cord cross-sectional area (SC-CSA) as an outcome measure in clinical trials. Methods: In total, 44 PMS patients (26 primary progressive multiple sclerosis (PPMS), 18 secondary progressive multiple sclerosis (SPMS)) and 29 healthy controls (HCs) were studied at baseline and 12 months. SC-CSA was measured using the three-dimensional (3D) fast field echo sequences acquired at 3T and the active surface model. Multiple linear regressions were used to investigate changes in imaging measurements. Results: PPMS patients had shorter disease duration, lower Expanded Disability Status Scale (EDSS) and larger SC-CSA than SPMS patients. All patients together showed a significantly greater decrease in percentage SC-CSA change than HCs, which was driven by the PPMS. All patients deteriorated over 1 year, but no association was found between percentage SC-CSA change and clinical changes. The sample size per arm required to detect a 50% treatment effect over 1 year, at 80% power, was 57 for PPMS and 546 for SPMS. Conclusion: SC-CSA may become an outcome measure in trials of PPMS patients, when they are at an early stage of the disease, have moderate disability and modest SC atrophy.

A Modality-Agnostic Patch-Based Technique for Lesion Filling in Multiple Sclerosis

Multiple Sclerosis lesions influence the process of image analysis, leading to tissue segmentation problems and biased morphometric estimates. With the aim of reducing this bias, existing techniques fill segmented lesions as normal appearing white matter. However, due to lesion segmentation errors or the presence of neighbouring structures, such as the ventricles and deep grey matter structures, filling all lesions as white matter like intensities is prone to introduce errors and artefacts. In this paper, we present a novel lesion filling strategy based on in-painting techniques for image completion. This technique makes use of a patch-based Non-Local Means algorithm that fills the lesions with the most plausible texture, rather than normal appearing white matter. We demonstrate that this strategy introduces less bias and fewer artefacts and spurious edges than previous techniques. The advantages of the proposed methodology are that it preserves both anatomical structure and signal-to-noise characteristics even when the lesions are neighbouring grey matter and cerebrospinal fluid, and avoids excess blurring or rasterisation due to the choice of segmentation plane, and lesion shape, size and/or position.

A multi-time-point modality-agnostic patch-based method for lesion filling in multiple sclerosis.

Multiple sclerosis lesions influence the process of image analysis, leading to tissue segmentation problems and biased morphometric estimates. Existing techniques try to reduce this bias by filling all lesions as normal-appearing white matter on T1-weighted images, considering each time-point separately. However, due to lesion segmentation errors and the presence of structures adjacent to the lesions, such as the ventricles and deep grey matter nuclei, filling all lesions with white matter-like intensities introduces errors and artefacts. In this paper, we present a novel lesion filling strategy inspired by in-painting techniques used in computer graphics applications for image completion. The proposed technique uses a five-dimensional (5D), patch-based (multi-modality and multi-time-point), Non-Local Means algorithm that fills lesions with the most plausible texture. We demonstrate that this strategy introduces less bias, fewer artefacts and spurious edges than the current, publicly available techniques. The proposed method is modality-agnostic and can be applied to multiple time-points simultaneously. In addition, it preserves anatomical structures and signal-to-noise characteristics even when the lesions are neighbouring grey matter or cerebrospinal fluid, and avoids excess of blurring or rasterisation due to the choice of the segmentation plane, shape of the lesions, and their size and/or location.

Value of the central vein sign at 3T to differentiate MS from seropositive NMOSD

Objective To assess the value of the central vein sign (CVS) on a clinical 3T scanner to distinguish between multiple sclerosis (MS) and neuromyelitis optica spectrum disorder (NMOSD). Methods Eighteen aquaporin-4-antibody-positive patients with NMOSD, 18 patients with relapsing-remitting MS, and 25 healthy controls underwent 3T MRI. The presence of a central vein in white matter lesions on susceptibility-weighted imaging, defined as a thin hypointense line or a small dot, was recorded. Results The proportion of lesions with the CVS was higher in MS than NMOSD (80% vs 32%, p < 0.001). A greater proportion of lesions with the CVS predicted the diagnosis of MS, rather than NMOSD (odds ratio 1.10, 95% confidence interval [CI] 1.04 to 1.16, p = 0.001), suggesting that each percent unit increase in the proportion of lesions with the CVS in an individual patient was associated with a 10% increase in the risk of the same patient having MS. If more than 54% of the lesions on any given scan show the CVS, then the patient can be given a diagnosis of MS with an accuracy of 94% (95% CIs 81.34, 99.32, p < 0.001, sensitivity/specificity 90%/100%). Conclusion The clinical value of the CVS in the context of the differential diagnosis between MS and NMOSD, previously suggested using 7T scanners, is now extended to clinical 3T scanners, thereby making a step towards the use of CVS in clinical practice. Classification of evidence This study provides Class III evidence that the CVS on 3T MRI accurately distinguishes patients with MS from those with seropositive NMOSD.

Deep gray matter volume loss drives disability worsening in multiple sclerosis

Gray matter (GM) atrophy occurs in all multiple sclerosis (MS) phenotypes. We investigated whether there is a spatiotemporal pattern of GM atrophy that is associated with faster disability accumulation in MS.