Instance-level quantitative saliency in multiple sclerosis lesion segmentation.

White matter lesions (WML) are common in a variety of brain pathologies, including ischemia affecting blood vessels deeper inside the brain’s white matter, and show an abnormal signal in T1-weighted and FLAIR images. The emergence of personalized medicine requires quantification and analysis of differential characteristics of WML across different brain regions. Manual segmentation and analysis of WMLs is laborious and time-consuming; therefore, automated methods providing robust, reproducible, and fast WML segmentation and analysis are highly desirable. In this study, we tackled the segmentation problem as a voxel-based classification problem. We developed an ensemble of different classification models, including six models of support vector machine, trained on handcrafted and transfer learning features, and five models of Residual neural network, trained on varying window sizes. The output of these models was combined through majority-voting. A series of image processing operations was applied to remove false positives in a post-processing step. Moreover, images were mapped to a standard atlas template to quantify the spatial distribution of WMLs, and a radiomic analysis of all the lesions across different brain regions was carried out. The performance of the method on multi-institutional WML Segmentation Challenge dataset (n = 150) comprising T1-weighted and FLAIR images was >90% within data of each institution, multi-institutional data pooled together, and across-institution training–testing. Forty-five percent of lesions were found in the temporal lobe of the brain, and these lesions were easier to segment (95.67%) compared to lesions in other brain regions. Lesions in different brain regions were characterized by their differential characteristics of signal strength, size/shape, heterogeneity, and texture (p < 0.001). The proposed multimodal ensemble-based segmentation of WML showed effective performance across all scanners. Further, the radiomic characteristics of WMLs of different brain regions provide an in vivo portrait of phenotypic heterogeneity in WMLs, which points to the need for precision diagnostics and personalized treatment.

Automatic segmentation of different-sized white matter lesions by voxel probability estimation

The segmentation of MRI scans of patients with white matter lesions (WML) is difficult because the MRI characteristics of WML are similar to those of gray matter. Intensity-based statistical classification techniques misclassify some WML as gray matter and some gray matter as WML.We developed a fast elastic matching algorithm that warps a reference data set containing information about the location of the gray matter into the approximate shape of the patient's brain. The region of white matter was segmented after segmenting the cortex and deep gray matter structures. The cortex was identified by using a three-dimensional, region-growing algorithm that was constrained by anatomical, intensity gradient, and tissue class parameters. White matter and WML were then segmented without interference from gray matter by using a two-class minimum-distance classifier.Analysis of double-echo spin-echo MRI scans of 16 patients with clinically determined multiple sclerosis (MS) was carried out. The segmentation of the cortex and deep gray matter structures provided anatomical context. This was found to improve the segmentation of MS lesions by allowing correct classification of the white matter region despite the overlapping tissue class distributions of gray matter and MS lesion.

White matter lesions (WMLs) in the human brain are generally diagnosed by using magnetic resonance (MR) images. Doctors working on WMLs generally need to calculate the volume of lesions for each patient at regular intervals in order to observe the course of diseases and manage the treatment process. This paper introduces an unsupervised automatic approach for segmentation of WMLs in the human brain. The approach consists of skull stripping, preprocessing, and lesion detection steps. Three skull stripping methods are proposed to increase successful stripping probability on various qualities of MR image data. After preprocessing and segmenting lesions, the system applies volumetric calculation and 3D visualization of lesions. This volumetric information can be used by doctors to observe changes in the lesions against regularly scanned MR images of patients. GPU-based parallel image processing techniques are utilized on Nvidia CUDA environment in order to improve performance by 40–50 times. Therefore, the developed system saves the time of doctors by providing them a fast automatic segmentation method for WMLs.

Background Cortical lesions develop early in multiple sclerosis (MS) and play a major role in disease progression. MRI at 7.0 T shows high sensitivity for detection of cortical lesions as well as better spatial resolution and signal-to-noise ratio compared with lower field strengths. Purpose To longitudinally characterize (a) the development and evolution of cortical lesions in multiple sclerosis across the cortical width, sulci, and gyri; (b) their relation with white matter lesion accrual; and (c) the contribution of 7.0-T cortical and white matter lesion load and cortical thickness to neurologic disability. Materials and Methods Twenty participants with relapsing-remitting MS and 13 with secondary progressive MS, along with 10 age-matched healthy controls, were prospectively recruited from 2010 to 2016 to acquire, in two imaging sessions (mean interval, 1.5 years), 7.0-T MRI T2*-weighted gradient-echo images (0.33 × 0.33 × 1.0 mm3) for cortical and white matter lesion segmentation and 3.0-T T1-weighted images for cortical surface reconstruction and cortical thickness estimation. Cortical lesions were sampled through the cortex to quantify cortical lesion distribution. The Expanded Disability Status Scale (EDSS) was used to assess neurologic disability. Nonparametric statistics assessed differences between and within groups in MRI metrics of cortical and white matter lesion burden; regression analysis explored associations of disability with MRI metrics. Results Twenty-five of 31 (81%) participants developed new cortical lesions per year (intracortical, 1.3 ± 1.7 vs leukocortical, 0.7 ± 1.9; P = .04), surpassing white matter lesion accrual (cortical, 2.0 ± 2.8 vs white matter, 0.7 ± 0.6; P = .01). In contrast to white matter lesions, cortical lesion accrual was greater in participants with secondary progressive MS than with relapsing-remitting MS (3.6 lesions/year ± 4.2 vs 1.1 lesions/year ± 0.9, respectively; P = .03) and preferentially localized in sulci. Total cortical lesion volume independently predicted baseline EDSS (β = 1.5, P < .001) and EDSS changes at follow-up (β = 0.5, P = .003). Conclusion Cortical lesions predominantly develop intracortically and within sulci, suggesting an inflammatory cerebrospinal fluid-mediated lesion pathogenesis. Cortical lesion accumulation was prominent at 7.0 T and independently predicted neurologic disability progression. © RSNA, 2019 Online supplemental material is available for this article. See also the editorial by Filippi and Rocca in this issue.

The cause of late-onset first epileptic seizures (LOFES) in older age is often not readily evident. In absence of probable causes, it has been suggested that cerebral small vessel disease (CSVD), which is common with increasing age, may be crucial. We aimed to further investigate the impact of white matter lesion (WML) burden and distribution pattern on LOFES. We retrospectively compared structural MRI of LOFES patients (n = 39) aged 60 years or older to controls with a transient ischemic attack (TIA, n = 38) and to patient controls (n = 35). WML segmentation was performed on FLAIR images using the SPM based automated lesion prediction algorithm of the LST toolbox and careful manual adjustment. Further, a dichotomization of WML was achieved by use of the BIANCA masking function. A voxel-based morphometry (VBM) analysis was additionally performed on T1 weighted sequences using the automated SPM12 based CAT12 software. Comparing intrapersonal volume ratios adjusted for the effects of gender and age, we found that the WML distribution was shifted to the juxtacortical compartment in LOFES patients. Among several influencing variables a path analysis could additionally show that this juxtacortical weighting of WML was a significant predictor for LOFES (β = 0.509, p < 0.001). With regard to total WML volume, LOFES and TIA patients did not differ significantly. Compared to TIA group, LOFES patients gray matter volume was regionally decreased in the right pre- and postcentral gyrus. By using algorithm-based automated lesion segmentation software tools and VBM analysis we could highlight that a juxtacortical weighting of WML distribution and regionally decreased gray matter volume distinguished LOFES from TIA and PC groups in our sample.

In this paper we describe and validate a longitudinal method for whole-brain segmentation of longitudinal MRI scans. It builds upon an existing whole-brain segmentation method that can handle multi-contrast data and robustly analyze images with white matter lesions. This method is here extended with subject-specific latent variables that encourage temporal consistency between its segmentation results, enabling it to better track subtle morphological changes in dozens of neuroanatomical structures and white matter lesions. We validate the proposed method on multiple datasets of control subjects and patients suffering from Alzheimer’s disease and multiple sclerosis, and compare its results against those obtained with its original cross-sectional formulation and two benchmark longitudinal methods. The results indicate that the method attains a higher test–retest reliability, while being more sensitive to longitudinal disease effect differences between patient groups. An implementation is publicly available as part of the open-source neuroimaging package FreeSurfer.

A method to automatically segment cerebrospinal fluid, gray matter, white matter and white matter lesions is presented. The method uses magnetic resonance brain images from proton density. T1-weighted and fluid-attenuated inversion recovery sequences. The method is based on an automatically trained k-nearest neighbour classifier extended with an additional step for the segmentation of white matter lesions. On six datasets, segmentations are quantitatively compared with manual segmentations, which have been carried out by two expert observers. For the tissues, similarity indices between method and observers approximate those between manual segmentations. Reasonably good lesion segmentation results are obtained compared to interobserver variability

In this work, we present a comparison of a shallow and a deep learning architecture for the automated segmentation of white matter lesions in MR images of multiple sclerosis patients. In particular, we train and test both methods on early stage disease patients, to verify their performance in challenging conditions, more similar to a clinical setting than what is typically provided in multiple sclerosis segmentation challenges. Furthermore, we evaluate a prototype naive combination of the two methods, which refines the final segmentation. All methods were trained on 32 patients, and the evaluation was performed on a pure test set of 73 cases. Results show low lesion-wise false positives (30%) for the deep learning architecture, whereas the shallow architecture yields the best Dice coefficient (63%) and volume difference (19%). Combining both shallow and deep architectures further improves the lesion-wise metrics (69% and 26% lesion-wise true and false positive rate, respectively).

We propose a fast, learning-based algorithm for segmenting white matter (WM) lesions for magnetic resonance (MR) brain images. The inputs to the algorithm are T1, T2, and FLAIR images. Unlike most of the previously reported learning-based algorithms, which treat expert labeled lesion map as ground truth in the training step, the proposed algorithm only requires the user to provide a few regions of interest (ROI's) containing lesions. An unsupervised clustering algorithm is applied to segment these ROI's into areas. Based on the assumption that lesion voxels have higher intensity on FLAIR image, areas corresponding to lesions are identified and their probability distributions in T1, T2, and FLAIR images are computed. The lesion segmentation in 3D is done by using the probability distributions to generate a confidence map of lesion and applying a graph based segmentation algorithm to label lesion voxels. The initial lesion label is used to further refine the probability distribution estimation for the final lesion segmentation. The advantages of the proposed algorithm are: 1. By using the weak labels, we reduced the dependency of the segmentation performance on the expert discrimination of lesion voxels in the training samples; 2. The training can be done using labels generated by users with only general knowledge of brain anatomy and image characteristics of WM lesion, instead of these carefully labeled by experienced radiologists; 3. The algorithm is fast enough to make interactive segmentation possible. We test the algorithm on nine ACCORD-MIND MRI datasets. Experimental results show that our algorithm agrees well with expert labels and outperforms a support vector machine based WM lesion segmentation algorithm.

LST-AI: A deep learning ensemble for accurate MS lesion segmentation

Magnetic Resonance Imaging (MRI) biomarkers of multiple sclerosis (MS), particularly fluid-attenuated inversion-recovery (FLAIR) sequences, have long been investigated. However, advanced analytical methods, capable of fusing results from different MRI modalities, could be more informative to have enabled joint biomarkers. Here we estimate disease duration (DD) in MS subjects (n=47) based on fusion of information from Myelin Water Imaging (MWI), Diffusion Tensor Imaging (DTI), and resting state functional MRI (rsfMRI) modalities, by adapting a joint Multimodal Statistical Analysis Framework. Using this data driven, multimodal, latent variable (LV) approach, common and unique information in each dataset was acquired and their relationship with DD is analyzed through the Least Absolute Shrinkage and Selection Operator (LASSO) regression. The common components between the three modalities, but not the unique components of each modality, accurately predicted DD. To further investigate the regions importance for estimating DD, we separated the data into two groups: “early” and “late” DD depending upon their relationship to the median duration of illness (120 months). In early disease, DTI information in the Right Tapetum and the Fornix jointly with MWI information in the Left Superior Cerebellar Penducle, Right Cerebral Penducle, and Cingulum were most informative. In contrast, rsfMRI demonstrated altered connectivity throughout disease duration. Our results demonstrate the power of multimodal imaging markers in MS.

Decision fusion is the most important step in ensemble machine learning schemes. One of the greatest challenges of decision fusion is the discrete nature of decisions. This challenge causes decision fusion solutions to become variations of the voting algorithm from statistical perspective. However, increasing the redundancy of decisions imposes a serious computational challenge for real-time systems. Resorting to fewer decisions imposes uncertainty challenges. In this Letter, the authors present a methodology to generate saliency maps for decision fusion. Specifically, they propose a local saliency map for decision fusion using a local majority filter. They choose semantic segmentation via pixel labelling produced from a random decision forest model as a case study. The local saliency map is used to derive three intermediate labelled images that are added to the voting pool and hence rectifying the final decision. The results of the proposed solution reduced the error by 26% and increased robustness by 16% with only two decisions.

White matter hyperintensity (WMHI) lesions on MR images are an important indication of various types of brain diseases that involve inflammation and blood vessel abnormalities. Automated quantification of the WMHI can be valuable for the clinical management of patients, but existing automated software is often developed for a single type of disease and may not be applicable for clinical scans with thick slices and different scanning protocols. The purpose of the study is to develop and validate an algorithm for automatic quantification of white matter hyperintensity suitable for heterogeneous MRI data with different disease types. We developed and evaluated "DeepWML", a deep learning method for fully automated white matter lesion (WML) segmentation of multicentre FLAIR images. We used MRI from 507 patients, including three distinct white matter diseases, obtained in 9 centres, with a wide range of scanners and acquisition protocols. The automated delineation tool was evaluated through quantitative parameters of Dice similarity, sensitivity and precision compared to manual delineation (gold standard). The overall median Dice similarity coefficient was 0.78 (range 0.64 ~ 0.86) across the three disease types and multiple centres. The median sensitivity and precision were 0.84 (range 0.67 ~ 0.94) and 0.81 (range 0.64 ~ 0.92), respectively. The tool's performance increased with larger lesion volumes. DeepWML was successfully applied to a wide spectrum of MRI data in the three white matter disease types, which has the potential to improve the practical workflow of white matter lesion delineation.

White matter lesion (WML) segmentation applied to magnetic resonance imaging (MRI) scans of people with multiple sclerosis has been an area of extensive research in recent years. As with most tasks in medical imaging, deep learning (DL) methods have proven very effective and have quickly replaced existing methods. Despite the improvement offered by these networks, there are still shortcomings with these DL approaches. In this work, we compare several DL algorithms, as well as methods for ensembling the results of those algorithms, for performing MS lesion segmentation. An ensemble approach is shown to best estimate total WML and has the highest agreement with manual delineations.