Search

BACKGROUND: The apolipoprotein (APOE) e4 allele is a known risk factor for Alzheimer’s disease (AD), while the e2 allele is thought to be protective against AD. Few studies have examined the relationship between brain pathologies, atrophy, and white matter hyperintensities (WMHs) and APOE status in those with the e2e4 genotype and results are inconsistent for those with an e2 allele. METHODS: We analyzed Alzheimer’s Disease Neuroimaging participants that had APOE genotyping and at least one of the following metrics: regional WMH load, ventricle size, hippocampal (HC) and entorhinal cortex (EC) volume, amyloid level (i.e., AV-45), and phosphorylated tau (pTau). Participants were divided into one of four APOE allele profiles (E4=e4e4 or e3e4; E2=e2e2 or e2e3; E3=e3e3; or E24=e2e4, Fig.1). Linear mixed models examined the relationship between APOE profiles and each pathology (i.e., regional WMHs, ventricle size, hippocampal and entorhinal cortex volume, amyloid level, and phosphorylated tau measures). while controlling for age, sex, education, and diagnostic status at baseline and over time. RESULTS: APOE ε4 is associated with increased pathology while ε2 positivity is associated with reduced baseline and lower accumulation of pathologies and rates of neurodegeneration. APOE ε2ε4 is similar to ε4 (increased neurodegeneration) but with a slower rate of change. CONCLUSIONS: The strong associations observed between APOE and pathology in this study show the importance of how genetic factors influence structural brain changes. These findings suggest that ε2ε4 genotype is related to increased declines associated with the ε4 as opposed to the protective effects of the ε2. These findings have important implications for initiating treatments and interventions. Given that people who have the ε2ε4 genotype can expect to have increased atrophy, they must be included (alongside those with an ε4 profile) in targeted interventions to reduce brain changes that occur with AD.

Copyright © The Author(s) 2022. Connections among brain regions allow pathological perturbations to spread from a single source region to multiple regions. Patterns of neurodegeneration in multiple diseases, including behavioural variant of frontotemporal dementia (bvFTD), resemble the large-scale functional systems, but how bvFTD-related atrophy patterns relate to structural network organization remains unknown. Here we investigate whether neurodegeneration patterns in sporadic and genetic bvFTD are conditioned by connectome architecture. Regional atrophy patterns were estimated in both genetic bvFTD (75 patients, 247 controls) and sporadic bvFTD (70 patients, 123 controls). First, we identified distributed atrophy patterns in bvFTD, mainly targeting areas associated with the limbic intrinsic network and insular cytoarchitectonic class. Regional atrophy was significantly correlated with atrophy of structurally- and functionally-connected neighbours, demonstrating that network structure shapes atrophy patterns. The anterior insula was identified as the predominant group epicentre of brain atrophy using data-driven and simulation-based methods, with some secondary regions in frontal ventromedial and antero-medial temporal areas. We found that FTD-related genes, namely C9orf72 and TARDBP, confer local transcriptomic vulnerability to the disease, modulating the propagation of pathology through the connectome. Collectively, our results demonstrate that atrophy patterns in sporadic and genetic bvFTD are jointly shaped by global connectome architecture and local transcriptomic vulnerability, providing an explanation as to how heterogenous pathological entities can lead to the same clinical syndrome. Canada First Research Excellence Fund, awarded to McGill University for the Healthy Brains for Healthy Lives initiative. B.M. acknowledges support from the Natural Sciences and Engineering Research Council of Canada (NSERC Discovery Grant RGPIN #017-04265) and from the Canada Research Chairs Program. S.D. receives salary support from the Fonds de Recherche du Québec—Santé (FRQS). G.S. acknowledges support from the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Fonds de recherche du Québec—Nature et Technologies (FRQNT). V.B. acknowledges support from the Fonds de recherche du Québec—Nature et Technologies (FRQNT). FTLDNI data collection and sharing was funded by the Frontotemporal Lobar Degeneration Neuroimaging Initiative (National Institutes of Health Grant R01 AG032306) and is coordinated through the University of California, San Francisco, Memory and Aging Center. FTLDNI data are disseminated by the Laboratory for Neuro Imaging at the University of Southern California.

© Author(s) (or their employer(s)) 2021. No commercial re-use. See rights and permissions. Published by BMJ., Introduction: Structural brain imaging is paramount for the diagnosis of behavioural variant of frontotemporal dementia (bvFTD), but it has low sensitivity leading to erroneous or late diagnosis. Methods: A total of 515 subjects from two different bvFTD cohorts (training and independent validation cohorts) were used to perform voxel-wise morphometric analysis to identify regions with significant differences between bvFTD and controls. A random forest classifier was used to individually predict bvFTD from deformation-based morphometry differences in isolation and together with semantic fluency. Tenfold cross validation was used to assess the performance of the classifier within the training cohort. A second held-out cohort of genetically confirmed bvFTD cases was used for additional validation. Results: Average 10-fold cross-validation accuracy was 89% (82% sensitivity, 93% specificity) using only MRI and 94% (89% sensitivity, 98% specificity) with the addition of semantic fluency. In the separate validation cohort of definite bvFTD, accuracy was 88% (81% sensitivity, 92% specificity) with MRI and 91% (79% sensitivity, 96% specificity) with added semantic fluency scores. Conclusion: Our results show that structural MRI and semantic fluency can accurately predict bvFTD at the individual subject level within a completely independent validation cohort coming from a different and independent database., Data collection and sharing for this project was funded by the Frontotemporal Lobar Degeneration Neuroimaging Initiative (National Institutes of Health Grant R01 AG032306). The study is coordinated through the University of California, San Francisco, Memory and Aging Center. FTLDNI data are disseminated by the Laboratory for Neuro Imaging at the University of Southern California. Brain scan acquisition at the McConnell Brain Imaging was supported by the Brain Canada Foundation with support from Health Canada and the Canada Foundation for Innovation (CFI Project 34874). This work was supported by Italian Ministry of Health (CoEN015 and Ricerca Corrente).

INTRODUCTION: Lateral ventricles are reliable and sensitive indicators of brain atrophy and disease progression in behavioral variant frontotemporal dementia (bvFTD). We aimed to investigate whether an automated tool using ventricular features could improve diagnostic accuracy in bvFTD across neurodegenerative diseases. METHODS: Using 678 subjects (2750 timepoints), differences in ventricular features were assessed between bvFTD, normal controls and other dementia cohorts. RESULTS: Ventricular antero-posterior ratio (APR) was the only feature that was significantly different and increased faster in bvFTD compared to all other cohorts. We achieved a 10-fold cross-validation accuracy of 80% (77% sensitivity, 82% specificity) in differentiating bvFTD from all other cohorts with ventricular features, and 76% accuracy using only the single APR feature. DISCUSSION: APR could be a useful and easy-to-implement feature to aid bvFTD diagnosis. We have made our ventricle feature estimation and bvFTD diagnostic tool publicly available, allowing application of our model in other studies.

INTRODUCTION: Structural brain imaging is paramount for the diagnosis of behavioural variant of frontotemporal dementia (bvFTD), but it has low sensitivity leading to erroneous or late diagnosis. METHODS: A total of 515 subjects from two different bvFTD cohorts (training and independent validation cohorts) were used to perform voxel-wise morphometric analysis to identify regions with significant differences between bvFTD and controls. A random forest classifier was used to individually predict bvFTD from deformation-based morphometry differences in isolation and together with semantic fluency. Tenfold cross validation was used to assess the performance of the classifier within the training cohort. A second held-out cohort of genetically confirmed bvFTD cases was used for additional validation. RESULTS: Average 10-fold cross-validation accuracy was 89% (82% sensitivity, 93% specificity) using only MRI and 94% (89% sensitivity, 98% specificity) with the addition of semantic fluency. In the separate validation cohort of definite bvFTD, accuracy was 88% (81% sensitivity, 92% specificity) with MRI and 91% (79% sensitivity, 96% specificity) with added semantic fluency scores. CONCLUSION: Our results show that structural MRI and semantic fluency can accurately predict bvFTD at the individual subject level within a completely independent validation cohort coming from a different and independent database.

Neuroimaging stands to benefit from emerging ultrahigh-resolution histological atlases of the human brain; the first of which is "BigBrain". Ongoing research aims to characterise regional differentiation of cytoarchitecture with BigBrain and to optimise registration of BigBrain with standard neuroimaging templates. Together, this work paves the way for multi-scale investigations of brain organisation. However, working with BigBrain can present new challenges for neuroimagers, including dealing with cellular resolution neuroanatomy and complex transformation procedures. To simplify workflows and support adoption of best practices, we developed BigBrainWarp, a toolbox for integration of BigBrain with multimodal neuroimaging. The primary BigBrainWarp function wraps multiple state-of-the-art deformation matrices into one line of code, allowing users to easily map data between BigBrain and standard MRI spaces. Additionally, the toolbox contains ready-to-use cytoarchitectural features to improve accessibility of histological information. The present article discusses recent contributions to BigBrain-MRI integration and demonstrates the utility of BigBrainWarp for further investigations.

INTRODUCTION: Heterogeneity in the progression of Alzheimer's disease makes it challenging to predict the rate of cognitive and functional decline for individual patients. Tools for short-term prediction could help enrich clinical trial designs and focus prevention strategies on the most at-risk patients. METHOD: We built a prognostic model using baseline cognitive scores and MRI-based features to determine which subjects with mild cognitive impairment remained stable and which functionally declined (measured by a two-point increase in CDR-SB) over 2 and 3-year follow-up periods, periods typical of the length of clinical trials. RESULTS: Combining both sets of features yields 77% accuracy (81% sensitivity and 75% specificity) to predict cognitive decline at 2 years (74% accuracy at 3 years with 75% sensitivity and 73% specificity). Using this tool to select trial participants yields a 3.8-fold decrease in the required sample size for a 2-year study (2.8-fold decrease for a 3-year study) for a hypothesized 25% treatment effect to reduce cognitive decline. DISCUSSION: This cohort enrichment tool could accelerate treatment development by increasing power in clinical trials.

Supplemental material, sj-pdf-1-jcb-10.1177_0271678X211035625 for Preclinical in vivo longitudinal assessment of KG207-M as a disease-modifying Alzheimer’s disease therapeutic by Min Su Kang, Monica Shin, Julie Ottoy, Arturo Aliaga Aliaga, Sulantha Mathotaarachchi, Kely Quispialaya, Tharick A Pascoal, D Louis Collins, M. Mallar Chakravarty, Axel Mathieu, Åsa Sandelius, Kaj Blennow, Henrik Zetterberg, Gassan Massarweh, Jean-Paul Soucy, A Claudio Cuello, Serge Gauthier, Michael Waterston, Nathan Yoganathan, Etienne Lessard, Arsalan Haqqani, Kerry Rennie, Danica Stanimirovic, Balu Chakravarthy and Pedro Rosa-Neto in Journal of Cerebral Blood Flow & Metabolism

We describe the creation of an open science dataset from a cohort of cognitively unimpaired aging individuals with a parental or multiple-sibling history of Alzheimer’s disease (AD). Our purpose was to enable PResymptomatic EValuation of Novel or Experimental Treatments for AD (“PREVENT-AD”). To characterize this population, possibly progressing in the pre-symptomatic phase of AD, we studied genetic variants and obtained longitudinal measures of cognition, brain structure and function, blood and cerebral fluid biochemistry and neurosensory capacities. Two nested prevention trials were also conducted. Data were hosted in LORIS, a platform that facilitates data organization, curation and sharing. We initially assessed 425 individuals, 385 meeting criteria for sustained investigation and 330 remaining active for longitudinal follow-ups. Between 2011 and 2017, we obtained quality-controlled data from 1704 MRI scans, 532 CSF samples, and 1882 cognitive evaluations. To date, 310 active participants (94%) have agreed that their data be openly shared. In addition to being a living resource for continued data acquisition, therefore, PREVENT-AD offers shared data to facilitate understanding of AD pathogenesis.

Stroke is the second leading cause of disability worldwide. In order to minimize disability, the goal of stroke treatment is to preserve tissue in the area where blood supply is decreased but sufficient to stave off cell death. Thrombectomy has been shown to offer fast and efficient reperfusion with high recanalization rates. However, due to the inherent risks of thrombectomy, indications including evidence of good collateral circulation should be present. Currently, methods for evaluating collateral circulation are limited. In this work, we present an automated technique to compute a collateral circulation score based on differences seen in mean intensities between left and right cerebral hemispheres in 4D angiography images. In this preliminary study, we analyzed scans of 28 subjects: 4 with normal collateral circulation, 7 with good, 10 with intermediate, and 7 with poor collateral circulation, (based on radiologist score) and found a good correlation between the computed score and radiologist score (r2 = 0.71) and good separation between good and intermediate/poor groups. This work shows promise as the first step in building an automated collateral circulation score., Comment: 8 pages, 7 7 figures, MICCAI Computing and Visualization for Intravascular Imaging (CVII) Workshop

Brain gray matter (GM) morphometric changes are prevalent in both aging and Alzheimers disease (AD), though disentangling these two processes has proved challenging. Using independent component analysis, we derived morphometric networks from a large, multi-cohort dataset, and investigated how GM volume within these networks differs in young adulthood, old adulthood, and AD. Aging and AD contributed additive effects on GM loss in nearly all networks, except frontal lobe networks, where GM reductions were more specific to aging. While no networks show GM loss highly specific to AD, a higher degree of variability in the whole-brain pattern of GM volume characterized AD only. Preservation of the whole-brain GM pattern in cognitively normal older adults was related to better cognition and lower risk of developing cognitive impairment. These results suggest both aging and AD involve widespread atrophy, but that cognitive impairment is uniquely associated with disruption of morphometric organization.

Quantification of cerebral white matter hyperintensities (WMH) of presumed vascular origin is of key importance in many neurological research studies. Currently, measurements are often still obtained from manual segmentations on brain MR images, which is a laborious procedure. Automatic WMH segmentation methods exist, but a standardized comparison of the performance of such methods is lacking. We organized a scientific challenge, in which developers could evaluate their method on a standardized multi-center/-scanner image dataset, giving an objective comparison: the WMH Segmentation Challenge (https://wmh.isi.uu.nl/). Sixty T1+FLAIR images from three MR scanners were released with manual WMH segmentations for training. A test set of 110 images from five MR scanners was used for evaluation. Segmentation methods had to be containerized and submitted to the challenge organizers. Five evaluation metrics were used to rank the methods: (1) Dice similarity coefficient, (2) modified Hausdorff distance (95th percentile), (3) absolute log-transformed volume difference, (4) sensitivity for detecting individual lesions, and (5) F1-score for individual lesions. Additionally, methods were ranked on their inter-scanner robustness. Twenty participants submitted their method for evaluation. This paper provides a detailed analysis of the results. In brief, there is a cluster of four methods that rank significantly better than the other methods, with one clear winner. The inter-scanner robustness ranking shows that not all methods generalize to unseen scanners. The challenge remains open for future submissions and provides a public platform for method evaluation.

Background Impulsivity is a risk factor for obesity. It has different underlying facets that can be assessed using questionnaires. Impulsivity can be further refined by the use of food-specific questionnaires, which measure a tendency to uncontrolled eating. We examined how these impulsivity measures relate to each other, to obesity, and to brain anatomy. Methods We assessed students in their first year of university - a risky period for weight gain- at the beginning (N = 2214) and at the end of the school year (N = 1145) using questionnaire measures of impulsivity, personality, stress reactivity and eating-specific traits. A subset of participants (N = 72) underwent T1-weighted MRI to investigate the brain correlates of impulsivity. Results Using factor analysis, we show that impulsivity can be stratified into three domains, which we label stress reactivity, reward sensitivity and self-control, while eating questionnaires resolve into a single latent factor - uncontrolled eating. A watershed model shows that uncontrolled eating mediates the effect of impulsivity traits on BMI. Self-control and stress reactivity scores are associated with a thinner lateral orbitofrontal cortex. In addition, stress reactivity correlates positively with amygdala and negatively with hippocampal volume. Longitudinally, lack of self-control, not uncontrolled eating, correlates with weight gain, while stress reactivity correlates with weight loss in male students. Conclusions The brain-impulsivity-obesity relationship is hierarchical. Structural brain differences relate to differences in impulsivity domains which affect BMI via uncontrolled eating. However, longitudinally, low self-control, not uncontrolled eating, is a predictor of weight gain in this sample.

Linear registration to stereotaxic space is a common first step in many automated image-processing tools for analysis of human brain MRI scans. This step is crucial for the success of the following image-processing steps. Several well-established algorithms are commonly used in the field of neuroimaging for this task, but none of them has a 100% success rate. Manual assessment of the registration is commonly used as part of quality control.We propose a completely automatic quality control method based on deep learning that replaces human rater and accurately performs quality control assessment for stereotaxic registration of T1w brain scans.In a recently published study from our group comparing linear registration methods, we used a database of 9693 MRI scans from several publically available datasets and applied five linear registration tools. In this study, the resulting images that were assessed and labeled by a human rater are used to train a deep neural network to detect cases when registration failed.Our method was able to achieve 88% accuracy and 11% false positive rate in detecting scans that should pass quality control, better than a manual QC rater.

We propose a tissue classification method for MRI scans of 6-month old infants, and used it to process the iSeg-2017 challenge data. The method relies on a deep-learning 3D U-Net network, trained with MRI scans of 216 infants, aged 6-24 months, from the ACE-IBIS longitudinal database.

In the field of computation anatomy, the diffeomorphic framework is widely used to perform analysis of human brain anatomy in both healthy and diseased populations. While useful for analysis, the framework imposes certain implementation constraints that do not necessarily result in improved accuracy of inter-subject co-registration in case of average anatomical template (AAT) construction – a common technique used in large population studies. In this work, we evaluated several state-of-the-art non-diffeomorphic and diffeomorphic non-linear registration frameworks in terms of their ability to build AATs. While all methods generated well behaved transforms, we found that the diffeomorphic framework does not automatically guarantee an increase of accuracy in average anatomical template construction.

IntroductionFluid-attenuated Inversion Recovery (FLAIR) and dual T2w and proton density (PD) magnetic resonance images (MRIs) are considered to be the optimum sequences for detecting white matter hyperintensities (WMHs) in aging and Alzheimer's disease populations. However, many existing large multisite studies forgo their acquisition in favor of other MRI sequences due to economic and time constraints.MethodsIn this article, we have investigated whether FLAIR and T2w/PD sequences are necessary to detect WMHs in Alzheimer's and aging studies, compared to using only T1w images. Using a previously validated automated tool based on a Random Forests classifier, WMHs were segmented for the baseline visits of subjects from ADC, ADNI1, and ADNI2/GO studies with and without T2w/PD and FLAIR information. The obtained WMH loads (WMHLs) in different lobes were then correlated with manually segmented WMHLs, each other, age, cognitive, and clinical measures to assess the strength of the correlations with and without using T2w/PD and FLAIR information.ResultsThe WMHLs obtained from T1w-Only segmentations correlated with the manual WMHLs (ADNI1: r = .743, p

Structural MRI is widely used for investigating brain atrophy in many neurodegenerative disorders, with several research groups developing and publishing techniques to provide quantitative assessments of this longitudinal change. Often techniques are compared through computation of required sample size estimates for future clinical trials. However interpretation of such comparisons is rendered complex because, despite using the same publicly available cohorts, the various techniques have been assessed with different data exclusions and different statistical analysis models. We created the MIRIAD atrophy challenge in order to test various capabilities of atrophy measurement techniques. The data consisted of 69 subjects (46 Alzheimer's disease, 23 control) who were scanned multiple (up to twelve) times at nine visits over a follow-up period of one to two years, resulting in 708 total image sets. Nine participating groups from 6 countries completed the challenge by providing volumetric measurements of key structures (whole brain, lateral ventricle, left and right hippocampi) for each dataset and atrophy measurements of these structures for each time point pair (both forward and backward) of a given subject. From these results, we formally compared techniques using exactly the same dataset. First, we assessed the repeatability of each technique using rates obtained from short intervals where no measurable atrophy is expected. For those measures that provided direct measures of atrophy between pairs of images, we also assessed symmetry and transitivity. Then, we performed a statistical analysis in a consistent manner using linear mixed effect models. The models, one for repeated measures of volume made at multiple time-points and a second for repeated “direct” measures of change in brain volume, appropriately allowed for the correlation between measures made on the same subject and were shown to fit the data well. From these models, we obtained estimates of the distribution of atrophy rates in the Alzheimer's disease (AD) and control groups and of required sample sizes to detect a 25% treatment effect, in relation to healthy ageing, with 95% significance and 80% power over follow-up periods of 6, 12, and 24 months. Uncertainty in these estimates, and head-to-head comparisons between techniques, were carried out using the bootstrap. The lateral ventricles provided the most stable measurements, followed by the brain. The hippocampi had much more variability across participants, likely because of differences in segmentation protocol and less distinct boundaries. Most methods showed no indication of bias based on the short-term interval results, and direct measures provided good consistency in terms of symmetry and transitivity. The resulting annualized rates of change derived from the model ranged from, for whole brain: − 1.4% to − 2.2% (AD) and − 0.35% to − 0.67% (control), for ventricles: 4.6% to 10.2% (AD) and 1.2% to 3.4% (control), and for hippocampi: − 1.5% to − 7.0% (AD) and − 0.4% to − 1.4% (control). There were large and statistically significant differences in the sample size requirements between many of the techniques. The lowest sample sizes for each of these structures, for a trial with a 12 month follow-up period, were 242 (95% CI: 154 to 422) for whole brain, 168 (95% CI: 112 to 282) for ventricles, 190 (95% CI: 146 to 268) for left hippocampi, and 158 (95% CI: 116 to 228) for right hippocampi. This analysis represents one of the most extensive statistical comparisons of a large number of different atrophy measurement techniques from around the globe. The challenge data will remain online and publicly available so that other groups can assess their methods., Highlights • We compared numerous brain atrophy measurement techniques using multiple metrics. • Each participant produced measures on the exact same dataset, blinded to disease. • A central statistical analysis using linear mixed effect models was performed. • Head to head comparisons for each region were performed using sample size estimates. • Brain and ventricle measures were more consistent across groups than for hippocampi.

Successful automated detection of short needles during an intervention is necessary to allow the physician identify and correct any misalignment of the needle and the target at early stages, which reduces needle passes and improves health outcomes. In this paper, we present a novel approach to detect needle voxels in 3D ultrasound volume with high precision using convolutional neural networks. Each voxel is classified from locally-extracted raw data of three orthogonal planes centered on it. We propose a bootstrap re-sampling approach to enhance the training in our highly imbalanced data. The proposed method successfully detects 17G and 22G needles with a single trained network, showing a robust generalized approach. Extensive ex-vivo evaluations on 3D ultrasound datasets of chicken breast show 25% increase in F1-score over the state-of-the-art feature-based method. Furthermore, very short needles inserted for only 5 mm in the volume are detected with tip localization errors of \({

Central post-stroke pain of thalamic origin is an extremely distressing and often refractory disorder. There are no well-established predictors for pain development after thalamic stroke, and the role of different thalamic nuclei is unclear. Here, we used structural magnetic resonance imaging to identify the thalamic nuclei, specifically implicated in the generation of central post-stroke pain of thalamic origin. Lesions of 10 patients with central post-stroke pain of thalamic origin and 10 control patients with thalamic strokes without pain were identified as volumes of interest on magnetic resonance imaging data. Non-linear deformations were estimated to match each image with a high-resolution template and were applied to each volume of interest. By using a digital atlas of the thalamus, we elucidated the involvement of different nuclei with respect to each lesion. Patient and control volumes of interest were summed separately to identify unique areas of involvement. Voxelwise odds ratio maps were calculated to localize the anatomical site where lesions put patients at risk of developing central post-stroke pain of thalamic origin. In the patients with pain, mainly lateral and posterior thalamic nuclei were affected, whereas a more anterior-medial lesion pattern was evident in the controls. The lesions of 9 of 10 pain patients overlapped at the border of the ventral posterior nucleus and the pulvinar, coinciding with the ventrocaudalis portae nucleus. The lesions of this area showed an odds ratio of 81 in favour of developing thalamic pain. The high odds ratio at the ventral posterior nucleus-pulvinar border zone indicates that this area is crucial in the pathogenesis of thalamic pain and demonstrates the feasibility of identifying patients at risk of developing central post-stroke pain of thalamic origin early after thalamic insults. This provides a basis for pre-emptive treatment studies

INTRODUCTION: White matter hyperintensities (WMHs) are areas of abnormal signal on magnetic resonance images (MRIs) that characterize various types of histopathological lesions. The load and location of WMHs are important clinical measures that may indicate the presence of small vessel disease in aging and Alzheimer’s disease (AD) patients. Manually segmenting WMHs is time consuming and prone to inter-rater and intra-rater variabilities. Automated tools that can accurately and robustly detect these lesions can be used to measure the vascular burden in individuals with AD or the elderly population in general. Many WMH segmentation techniques use a classifier in combination with a set of intensity and location features to segment WMHs, however, the optimal choice of classifier is unknown. METHODS: We compare 10 different linear and nonlinear classification techniques to identify WMHs from MRI data. Each classifier is trained and optimized based on a set of features obtained from co-registered MR images containing spatial location and intensity information. We further assessed the performance of the classifiers using different combinations of MRI contrast information. The performances of the different classifiers were compared on three heterogeneous multi-site datasets, including images acquired with different scanners and different scan-parameters. These included data from the ADC study from University of California Davis, the NACC database and the ADNI study. The classifiers (naïve Bayes, logistic regression, decision trees, random forests, support vector machines, k-nearest neighbors, bagging, and boosting) were evaluated using a variety of voxel-wise and volumetric similarity measures such as Dice Kappa similarity index (SI), Intra-Class Correlation (ICC), and sensitivity as well as computational burden and processing times. These investigations enable meaningful comparisons between the performances of different classifiers to determine the most suitable classifiers for segmentation of WMHs. In the spirit of open-source science, we also make available a fully automated tool for segmentation of WMHs with pre-trained classifiers for all these techniques. RESULTS: Random Forests yielded the best performance among all classifiers with mean Dice Kappa (SI) of 0.66±0.17 and ICC=0.99 for the ADC dataset (using T1w, T2w, PD, and FLAIR scans), SI=0.72±0.10, ICC=0.93 for the NACC dataset (using T1w and FLAIR scans), SI=0.66±0.23, ICC=0.94 for ADNI1 dataset (using T1w, T2w, and PD scans) and SI=0.72±0.19, ICC=0.96 for ADNI2/GO dataset (using T1w and FLAIR scans). Not using the T2w/PD information did not change the performance of the Random Forest classifier (SI=0.66±0.17, ICC=0.99). However, not using FLAIR information in the ADC dataset significantly decreased the Dice Kappa, but the volumetric correlation did not drastically change (SI=0.47±0.21, ICC=0.95). CONCLUSION: Our investigations showed that with appropriate features, most off-the-shelf classifiers are able to accurately detect WMHs in presence of FLAIR scan information, while Random Forests had the best performance across all datasets. However, we observed that the performances of most linear classifiers and some nonlinear classifiers drastically decline in absence of FLAIR information, with Random Forest still retaining the best performance.

Introduction: Hippocampal volume is a core biomarker of Alzheimer's disease (AD). However, its contribution over the standard diagnostic workup is unclear. Methods: Three hundred fifty-six patients, under clinical evaluation for cognitive impairment, with suspected AD and Mini-Mental State Examination >= 20, were recruited across 17 European memory clinics. After the traditional diagnostic workup, diagnostic confidence of AD pathology (DCAD) was estimated by the physicians in charge. The latter were provided with the results of automated hippocampal volumetry in standardized format and DCAD was reassessed. Results: An increment of one interquartile range in hippocampal volume was associated with a mean change of DCAD of 28.0% (95% credible interval: [211.5, 25.0]). Automated hippocampal volumetry showed a statistically significant impact on DCAD beyond the contributions of neuropsychology, 18 F-fluorodeoxyglucose positron emission tomography/single-photon emission computed tomography, and cerebrospinal fluid markers (28.5, CrI: [211.5, 25.6]; 214.1, CrI: [219.3, 28.8]; 210.6, CrI: [214.6, 26.1], respectively). Discussion: There is a measurable effect of hippocampal volume on DCAD even when used on top of the traditional diagnostic workup. (C) 2017 the Alzheimer's Association. Published by Elsevier Inc. All rights reserved.

International audience; Template-based analysis of multi-parametric MRI data of the spinal cord sets the foundation for multi-center studies with minimum bias, thereby helping the discovery of new biomarkers of spinal-related diseases. In this study, we introduce a spinal cord MRI template, the PAM50, which is anatomically compatible with the ICBM152 brain template and uses the same coordinate system. The fusion of the PAM50 and ICBM152 templates facilitates group studies and multi-center studies of combined brain and spinal cord MRI and also allows the use of existing atlases of the brainstem compatible with the ICBM template.

In this paper we propose a method to solve nonrigid image registration through a learning approach, instead of via iterative optimization of a predefined dissimilarity metric. We design a Convolutional Neural Network (CNN) architecture that, in contrast to all other work, directly estimates the displacement vector field (DVF) from a pair of input images. The proposed RegNet is trained using a large set of artificially generated DVFs, does not explicitly define a dissimilarity metric, and integrates image content at multiple scales to equip the network with contextual information. At testing time nonrigid registration is performed in a single shot, in contrast to current iterative methods. We tested RegNet on 3D chest CT follow-up data. The results show that the accuracy of RegNet is on par with a conventional B-spline registration, for anatomy within the capture range. Training RegNet with artificially generated DVFs is therefore a promising approach for obtaining good results on real clinical data, thereby greatly simplifying the training problem. Deformable image registration can therefore be successfully casted as a learning problem.

Large-scale collaborative analysis of brain imaging data, in psychiatry and neurology, offers a new source of statistical power to discover features that boost accuracy in disease classification, differential diagnosis, and outcome prediction. However, due to data privacy regulations or limited accessibility to large datasets across the world, it is challenging to efficiently integrate distributed information. Here we propose a novel classification framework through multi-site weighted LASSO: each site performs an iterative weighted LASSO for feature selection separately. Within each iteration, the classification result and the selected features are collected to update the weighting parameters for each feature. This new weight is used to guide the LASSO process at the next iteration. Only the features that help to improve the classification accuracy are preserved. In tests on data from five sites (299 patients with major depressive disorder (MDD) and 258 normal controls), our method boosted classification accuracy for MDD by 4.9% on average. This result shows the potential of the proposed new strategy as an effective and practical collaborative platform for machine learning on large scale distributed imaging and biobank data.

There are varying, often conflicting, reports with respect to altered striatal volume and morphometry in the major psychoses due to the influences of antipsychotic medications on striatal volume. Thus, disassociating disease effects from those of medication become exceedingly difficult. For the first time, using a longitudinally studied sample of structural magnetic resonance images from patients with childhood onset schizophrenia (COS; neurobiologically contiguous with the adult onset form of schizophrenia), their nonpsychotic siblings (COSSIBs), and novel shape mapping algorithms that are volume independent, we report the familial contribution of striatal morphology in schizophrenia. The results of our volumetric analyses demonstrate age-related increases in overall striatal volumes specific only to COS. However, both COS and COSSIBs showed overlapping shape differences in the striatal head, which normalized in COSSIBs by late adolescence. These results mirror previous studies from our group, demonstrating cortical thickness deficits in COS and COSSIBs as these deficits normalize in COSSIBs in the same age range as our striatal findings. Finally, there is a single region of nonoverlapping outward displacement in the dorsal aspect of the caudate body, potentially indicative of a response to medication. Striatal shape may be considered complimentary to volume as an endophenotype, and, in some cases may provide information that is not detectable using standard volumetric techniques. Our striatal shape findings demonstrate the striking localization of abnormalities in striatal the head. The neuroanatomical localization of these findings suggest the presence of abnormalities in the striatal-prefrontal circuits in schizophrenia and resilience mechanisms in COSSIBs with age dependent normalization.

National audience; – Les méthodes de segmentation automatique sont des outils importants pour l'analyse des images par résonance magnétique. Dans ce travail, nous introduisons une nouvelle méthode basée sur l'utilisation de patchs pour effectuer une segmentation de structures anatomiques. La méthode proposée est basée sur l'algorithme PatchMatch, optimisé pour la fusion d'étiquettes. Elle est nommée OPAL (pour Optimized PAtchMatch Label fusion), et fournit une précision de segmentation très compétitive en quasi temps-réel (moins d'1s de traitement par sujet). – Automatic segmentation methods are important tools for analysis of magnetic resonance images. In this paper, we introduce a new patch-based method using the PatchMatch algorithm to perform the segmentation of anatomical structures. Based on an Optimized PAtchMatch Label fusion (OPAL) strategy, the proposed method provides very competitive segmentation accuracy in near real-time (less than 1s per subject).

Background The European Alzheimer's Disease Consortium and Alzheimer's Disease Neuroimaging Initiative (ADNI) Harmonized Protocol (HarP) is a Delphi definition of manual hippocampal segmentation from magnetic resonance imaging (MRI) that can be used as the standard of truth to train new tracers, and to validate automated segmentation algorithms. Training requires large and representative data sets of segmented hippocampi. This work aims to produce a set of HarP labels for the proper training and certification of tracers and algorithms. Methods Sixty-eight 1.5 T and 67 3 T volumetric structural ADNI scans from different subjects, balanced by age, medial temporal atrophy, and scanner manufacturer, were segmented by five qualified HarP tracers whose absolute interrater intraclass correlation coefficients were 0.953 and 0.975 (left and right). Labels were validated as HarP compliant through centralized quality check and correction. Results Hippocampal volumes (mm 3 ) were as follows: controls: left = 3060 (standard deviation [SD], 502), right = 3120 (SD, 897); mild cognitive impairment (MCI): left = 2596 (SD, 447), right = 2686 (SD, 473); and Alzheimer's disease (AD): left = 2301 (SD, 492), right = 2445 (SD, 525). Volumes significantly correlated with atrophy severity at Scheltens' scale (Spearman's ρ = P = Cerebrospinal fluid spaces (mm 3 ) were as follows: controls: left = 23 (32), right = 25 (25); MCI: left = 15 (13), right = 22 (16); and AD: left = 11 (13), right = 20 (25). Five subjects (3.7%) presented with unusual anatomy. Conclusions This work provides reference hippocampal labels for the training and certification of automated segmentation algorithms. The publicly released labels will allow the widespread implementation of the standard segmentation protocol.

BackgroundAn international Delphi panel has defined a harmonized protocol (HarP) for the manual segmentation of the hippocampus on MR. The aim of this study is to study the concurrent validity of the HarP toward local protocols, and its major sources of variance.MethodsFourteen tracers segmented 10 Alzheimer's Disease Neuroimaging Initiative (ADNI) cases scanned at 1.5 T and 3T following local protocols, qualified for segmentation based on the HarP through a standard web-platform and resegmented following the HarP. The five most accurate tracers followed the HarP to segment 15 ADNI cases acquired at three time points on both 1.5 T and 3T.ResultsThe agreement among tracers was relatively low with the local protocols (absolute left/right ICC 0.44/0.43) and much higher with the HarP (absolute left/right ICC 0.88/0.89). On the larger set of 15 cases, the HarP agreement within (left/right ICC range: 0.94/0.95 to 0.99/0.99) and among tracers (left/right ICC range: 0.89/0.90) was very high. The volume variance due to different tracers was 0.9% of the total, comparing favorably to variance due to scanner manufacturer (1.2), atrophy rates (3.5), hemispheric asymmetry (3.7), field strength (4.4), and significantly smaller than the variance due to atrophy (33.5%, P < .001), and physiological variability (49.2%, P < .001).ConclusionsThe HarP has high measurement stability compared with local segmentation protocols, and good reproducibility within and among human tracers. Hippocampi segmented with the HarP can be used as a reference for the qualification of human tracers and automated segmentation algorithms. published

Structural MRI is widely used for investigating brain atrophy in many neurodegenerative disorders, with several research groups developing and publishing techniques to provide quantitative assessments of this longitudinal change. Often techniques are compared through computation of required sample size estimates for future clinical trials. However interpretation of such comparisons is rendered complex because, despite using the same publicly available cohorts, the various techniques have been assessed with different data exclusions and different statistical analysis models. We created the MIRIAD atrophy challenge in order to test various capabilities of atrophy measurement techniques. The data consisted of 69 subjects (46 Alzheimer's disease, 23 control) who were scanned multiple (up to twelve) times at nine visits over a follow-up period of one to two years, resulting in 708 total image sets. Nine participating groups from 6 countries completed the challenge by providing volumetric measurements of key structures (whole brain, lateral ventricle, left and right hippocampi) for each dataset and atrophy measurements of these structures for each time point pair (both forward and backward) of a given subject. From these results, we formally compared techniques using exactly the same dataset. First, we assessed the repeatability of each technique using rates obtained from short intervals where no measurable atrophy is expected. For those measures that provided direct measures of atrophy between pairs of images, we also assessed symmetry and transitivity. Then, we performed a statistical analysis in a consistent manner using linear mixed effect models. The models, one for repeated measures of volume made at multiple time-points and a second for repeated "direct" measures of change in brain volume, appropriately allowed for the correlation between measures made on the same subject and were shown to fit the data well. From these models, we obtained estimates of the distribution of atrophy rates in the Alzheimer's disease (AD) and control groups and of required sample sizes to detect a 25% treatment effect, in relation to healthy ageing, with 95% significance and 80% power over follow-up periods of 6, 12, and 24 months. Uncertainty in these estimates, and head-to-head comparisons between techniques, were carried out using the bootstrap. The lateral ventricles provided the most stable measurements, followed by the brain. The hippocampi had much more variability across participants, likely because of differences in segmentation protocol and less distinct boundaries. Most methods showed no indication of bias based on the short-term interval results, and direct measures provided good consistency in terms of symmetry and transitivity. The resulting annualized rates of change derived from the model ranged from, for whole brain: -1.4% to -2.2% (AD) and -0.35% to -0.67% (control), for ventricles: 4.6% to 10.2% (AD) and 1.2% to 3.4% (control), and for hippocampi: -1.5% to -7.0% (AD) and -0.4% to -1.4% (control). There were large and statistically significant differences in the sample size requirements between many of the techniques. The lowest sample sizes for each of these structures, for a trial with a 12 month follow-up period, were 242 (95% CI: 154 to 422) for whole brain, 168 (95% CI: 112 to 282) for ventricles, 190 (95% CI: 146 to 268) for left hippocampi, and 158 (95% CI: 116 to 228) for right hippocampi.

Structural MRI is an important imaging biomarker in Alzheimer’s disease as the cerebral atrophy has been shown to closely correlate with cogni-tive symptoms. Recognizing this, numerous methods have been developed for quantifying the disease related atrophy from MRI over the past decades. Special effort has been dedicated to separate AD related modifications from normal ag-ing for the purpose of early detection and prediction. Several groups have re-ported promising results using automatic methods; however, it is very difficult to compare these methods due to varying cohorts and different validation frameworks. To address this issue, the public challenge on Computer-Aided Di-agnosis of Dementia based on structural MRI data (CADDementia) was pro-posed. The challenge calls for accurate classification of 354 MRI scans collect-ed among AD patients, subjects with mild cognitive impairment and cognitively normal control. The true diagnosis is hidden from the participating groups, thus making the validation truly objective. This paper describes our proposed meth-od to automatically classify the challenge data along with a validation on 30 scans with known diagnosis also provided for the challenge.

The relationship between anxious/depressed traits and neuromaturation remains largely unstudied. Characterizing this relationship during healthy neurodevelopment is critical to understanding processes associated with the emergence of child/adolescent onset mood/anxiety disorders. In this study, mixed-effects models were used to determine longitudinal cortical thickness correlates of Child Behavior Checklist (CBCL) and Young Adult Self Report Anxious/Depressed scores in healthy children. Analyses included 341 subjects from 4.9 to 22.3 year-old with repeated MRI at up to 3 time points, at 2-year intervals (586 MRI scans). There was a significant “CBCL Anxious/Depressed by Age” interaction on cortical thickness in the right ventromedial prefrontal cortex (vmPFC), including the medial orbito-frontal, gyrus rectus, and subgenual anterior cingulate areas. Anxious/Depressed scores were negatively associated with thickness at younger ages (

Amnestic mild cognitive impairment (MCI) is considered a condition where patients are at risk of developing clinically definite Alzheimer’s disease (AD) with an annual conversion rate of approximately 15%[1]. AD is characterized by progressive brain atrophy with major impact on the cerebral cortex and medial temporal lobe structures such as hippocampus. Understanding the structural pattern of cortical atrophy at different stages of MCI, before AD can be diagnosed, may help in patient monitoring and prognosis. We used data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) to calculate and visualize the cortical atrophy at different stages in patients who eventually converted to clinically definite AD.We selected patients with a diagnosis of MCI from the ADNI database who converted to AD during the follow-up period. T1-weighted MRI scans were collected at time of conversion(n=140) and at intervals prior to the date of conversion: 6 months(n=121), 12 months(n=128), 24 months(n=62) and 36 months(n=29). As control group we collected all baseline scans of healthy controls (HC)(n=226). Images were denoised[2], bias field corrected[3], linearly co-registered[4], skull-stripped[5], and cortical thickness were measured using FACE[6] and mapped to an average cortical surface. Statistical maps of differences in cortical thickness between groups of MCI patients and HC were constructed and corrected for multiple comparisons.Three years prior to clinically definite AD, the MCI patients show signs of atrophy in the parahippocampal gyrus, the cingulate gyrus and temporal lobes (figure). At two years before conversion, atrophy has spread to large parts of the temporal and frontal lobes, much of the cingulate gyrus and parts of the orbitofrontal region. Twelve months before conversion, atrophy is present over all the cortex except the occipital lobe and the sensory-motor areas. Six months later and at conversion this pattern persists, but with more severe atrophy.Focal atrophy can be observed in patients with MCI three years before they are diagnosed with clinically definite AD. The atrophy accelerates during the MCI stage and affects the entire cortex except the occipital and sensory-motor cortex at the time of diagnosis - not unlike the progressive patterns of amyloid deposition described by Braak[7]. These patterns may be used to assess the disease progression in patients with MCI.

Classically, model-based segmentation procedures match magnetic resonance imaging (MRI) volumes to an expertly labeled atlas using nonlinear registration. The accuracy of these techniques are limited due to atlas biases, misregistration, and resampling error. Multi-atlas-based approaches are used as a remedy and involve matching each subject to a number of manually labeled templates. This approach yields numerous independent segmentations that are fused using a voxel-by-voxel label-voting procedure. In this article, we demonstrate how the multi-atlas approach can be extended to work with input atlases that are unique and extremely time consuming to construct by generating a library of multiple automatically generated templates of different brains (MAGeT Brain). We demonstrate the efficacy of our method for the mouse and human using two different nonlinear registration algorithms (ANIMAL and ANTs). The input atlases consist a high-resolution mouse brain atlas and an atlas of the human basal ganglia and thalamus derived from serial histological data. MAGeT Brain segmentation improves the identification of the mouse anterior commissure (mean Dice Kappa values (κ = 0.801), but may be encountering a ceiling effect for hippocampal segmentations. Applying MAGeT Brain to human subcortical structures improves segmentation accuracy for all structures compared to regular model-based techniques (κ = 0.845, 0.752, and 0.861 for the striatum, globus pallidus, and thalamus, respectively). Experiments performed with three manually derived input templates suggest that MAGeT Brain can approach or exceed the accuracy of multi-atlas label-fusion segmentation (κ = 0.894, 0.815, and 0.895 for the striatum, globus pallidus, and thalamus, respectively).

International audience; The human medial temporal lobe (MTL) is an important part of the limbic system, and its substructures play key roles in learning, memory, and neurodegeneration. The MTL includes the hippocampus (HC), amygdala (AG), parahippocampal cortex (PHC), entorhinal cortex, and perirhinal cortex-structures that are complex in shape and have low between-structure intensity contrast, making them difficult to segment manually in magnetic resonance images. This article presents a new segmentation method that combines active appearance modeling and patch-based local refinement to automatically segment specific substructures of the MTL including HC, AG, PHC, and entorhinal/perirhinal cortex from MRI data. Appearance modeling, relying on eigend-ecomposition to analyze statistical variations in image intensity and shape information in study population, is used to capture global shape characteristics of each structure of interest with a generative model. Patch-based local refinement, using nonlocal means to compare the image local intensity properties, is applied to locally refine the segmentation results along the structure borders to improve structure delimitation. In this manner, nonlocal regularization and global shape constraints could allow more accurate segmentations of structures. Validation experiments against manually defined labels demonstrate that this new segmentation method is computationally efficient, robust, and accurate. In a leave-one-out validation on 54 normal young adults, the method yielded a mean Dice κ of 0.87 for the HC, 0.81 for the AG, 0.73 for the anterior parts of the parahippocampal gyrus (entorhinal and perirhinal cortex), and 0.73 for the posterior parahippocampal gyrus. Hum Brain Mapp, 2012. © 2012 Wiley Periodicals, Inc.

International audience; In this paper, the method for the creation of an anatomically and mechanically realistic brain phantom from polyvinyl alcohol cryogel (PVA-C) is proposed for validation of image processing methods such as segmentation, reconstruction, registration, and denoising. PVA-C is material widely used in medical imaging phantoms because of its mechanical similarities to soft tissues. The phantom was cast in a mold designed using the left hemisphere of the Colin27 brain dataset [C. Holmes et al., "Enhancement of MR images using registration for signal averaging," J. Comput. Assist. Tomogr. 22(2), 324 (1998)]. Marker spheres and inflatable catheters were also implanted to enable good registration comparisons and to simulate tissue deformation, respectively. The phantom contained deep sulci, a complete insular region, and an anatomically accurate left ventricle. It was found to provide good contrast in triple modality imaging, consisting of computed tomography, ultrasound, and magnetic resonance imaging. Multiple sets of multimodal data were acquired from this phantom. The methods for building the anatomically accurate, multimodality phantom were described in this work. All multimodal data are made available freely to the image processing community (http://pvabrain.inria.fr). We believe the phantom images could allow for the validation and further aid in the development of novel medical image processing techniques.

Background:The atrophy of medial temporal lobe structures, such as the hippocampus (HC) and entorhinal cortex, is potentially specific and may serve as early biomarkers of Alzheimer’s disease [1]. In particular, the atrophy of the HC can be used as a marker of AD progression since changes in HC are closely related to changes in cognitive performance of the subject [1]. The evaluation of HC atrophy is usually estimated by volumetric studies on anatomical MRI, requiring a segmentation step that can be very time consuming when done manually. This limitation can be overcome by using automatic segmentation methods. However, the complex relationship between the intensity of the MR signal and the HC boundaries makes it difficult to accurately achieve this task. In the past few years, label fusion methods have demonstrated high performances in anatomical structure segmentation [2-4]. These methods use manual segmentations of anatomical MRIs from a training library to perform the segmentation of a subject. Recently, we have proposed a new nonlocal patch-based label fusion method and demonstrated that this approach can accurately segment the HC on young adults through an efficient fusion of manual segmentations [3]. By taking advantage of the redundancy of information present within the subject’s image, as well as the redundancy across subjects, the patch-based nonlocal means scheme enables the robust use of a large number of samples during estimation. Contrary to classical label fusion working at the structure level, our method deals with a finer level by comparing subparts of the structure (i.e., small 3D cubes). In this study, we propose to validate our nonlocal patch-based method on the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database by segmenting the HC of Cognitively Normal (CN) subjects and patients with early Alzheimer’s Disease (AD). Moreover, we evaluate the impact of the training library composition (i.e., only AD, only CN or a mix of AD and CN) on the segmentation accuracy of our algorithm. Methods:In this study, we used images obtained from the ADNI database (www.loni.ucla.edu/ADNI). This database contains both 1.5T and 3.0T T1-w MRI scans. For our experiments, we randomly selected 10 1.5T MRI baseline scans of CN subjects and 10 1.5T MRI baseline scans of patients with AD. After a linear registration to the stereotaxic space by using the ICBM-152 template as target image, the left and right hippocampi of each selected image were manually segmented by an expert at our centre.The first experiment was designed to evaluate the accuracy of our segmentation method on the ADNI dataset with a group constituting of CN subjects and patients with AD. The accuracy of the label fusion method varies according to the number of templates. In this first experiment we used a training library of 16 templates (8 CN subjects and 8 patients with AD). The second experiment was designed to investigate the impact of the composition of the training library. Three different training populations were built by combining the selected 20 MRI scans. The CN population was composed of 8 normal subjects, the AD population was composed of 8 patients and the mixed AD / CN population was composed of 4 CN subjects and 4 AD patients. This way, the number of templates in the training library was the same for all the training libraries compared.Through a leave-one-out procedure, our automatic segmentation method was applied on each of the 20 MRI scans. During the experiments, the default parameters were used for all the images (i.e., a patch-size of 7x7x7 voxels and a search area of 9x9x9 voxels). The quality of the obtained automatic segmentations was evaluated by estimating the Dice Kappa similarity index. This index measures the overlap between the manual segmentation and the segmentation produced by our patch-based method. Finally, a pair-wise multi-comparison test was used to detect significant differences between segmentation quality according to the training library for both populations CN and AD.Results: For the first experiment, the median Dice Kappa values are presented in Table 1. The segmentation accuracy was significantly better (p-value=0.002) for CN subjects (median k=0.883 for both HC) than for patients with AD (median k=0.838 for both HC). A median Dice Kappa value superior to 0.8 indicates a high correlation between manual and automatic segmentations. A median Kappa value superior 0.88 is similar to the highest published values in literature [2-4]. The difference in segmentation quality between populations might come from two sources. First, the higher anatomical variability of patients with AD makes the segmentation more difficult and may require a larger training library. Second, the smaller volumes of HC of patients with AD, due to the HC atrophy, can negatively bias the Dice Kappa index measure. For the second experiment, Figure 1 and Table 2 show the Dice Kappa similarity index for both studied populations according to the training library composition. As expected, by using smaller training libraries, the segmentation accuracy slightly decreased compared to experiment 1. For both populations, the best median Kappa values were obtained with the mixed CN / AD training library (k=0.875 for CN population and k=0.835 AD population). Moreover, for both populations, results were statistically similar by using a training library with the same property than the population segmented, compared to a mixed training library. When the entire training library is composed of templates with dissimilar property (i.e., segmenting AD subjects with CN templates), the results are significantly degraded; especially in the case of CN population (see left of Fig. 1). These results indicate that a mixed training library, containing a large number of patch variants, is better suited for segmentation than a non-population specific training library. Similarly to the first experiment, segmentation accuracy was significantly lower (p-value=0.004) for patients with AD than for CN subjects. Conclusions:The goals of this study were to validate our patch-based label fusion on ADNI data and to investigate the influence of the composition of the training library. First, we demonstrated that our patch-based method provides high segmentation accuracy for both CN and AD populations despite the small number of involved manual segmentation templates. During this experiment, the size and the anatomical variability of HC in AD population tended to decrease the segmentation accuracy compared to CN population. Second, we showed that the characteristic of the training library has a significant impact on the segmentation accuracy. By using a population specific training library or a mixed training library, better results are obtained. In conclusion, when the subject’s status is unknown, the results of our experiments suggest that using a mixed training library, covering the different scenarios, is the best strategy to adopt.

BackgroundRapid brain loss is characteristic for the patients with mild cognitive impairment (MCI) and Alzheimer disease (AD) [1]. Increase of the lateral ventricular volume is strongly correlated with the progression of the disease. High variability in the degree of atrophy for subjects with AD and MCI makes use of a single disease-specific template challenging. We propose a novel approach to generate a continuous four-dimensional template, where the 4th dimension is a surrogate measure of overall brain atrophy.MethodsWe used MRI scans obtained from the ADNI database (www.loni.ucla.edu/ADNI). Automated methods to estimate intracranial capacity (ICC) and lateral ventricles volume (LVV) [2] was applied to all available datasets at base line. The ratio between LVV and ICC (RLVV) was used as a surrogate measure of overall brain atrophy with mean(standard deviation) value of 2.46(0.87)%. Subsets from all subjects (CN, MCI and AD) were selected with uniform distribution of RLVV from 1.0 to 6.0% , resulting in a total of 160 subjects. Our algorithm [3] was modified to perform simultaneous 1) creation of the template and 2) linear regression of image intensity and shape versus RLVV.ResultsThe ratio between LVV and ICC yielded values of mean(sd) 2.13(0.72)% for NC, 2.45(0.84)% for MCI and 2.84(0.91)% for AD. The continuous, four dimensional anatomical template was created. For a given RLVV, an appropriate three dimensional anatomical template may be constructed, reflecting the average shape of the brain and the contrast between different tissue types for the given level of atrophy. Figure 1 shows images through 6 example values of increasing RLVV.ConclusionsThe proposed method and resulting template will be useful tools for the development of robust automatic image processing methods targeted to the study of the populations with high degree of variability of atrophy. Furthermore, method presented is not limited to be used only with MCI and AD subjects, but can also be easily adopted for other neuro-degenerative studies.[1] Carlson, NE, et al. Neurology, 2008. 70(11). [2] Fonov, VS, et al. Alzheimer's and Dementia, 2010. 6(4, Supplement 1). [3] Fonov, V, et al. NeuroImage, 2011. 54(1).

International audience; PURPOSE: To propose an innovative approach to better detect Alzheimer's Disease (AD) based on a finer detection of hippocampus (HC) atrophy patterns. METHOD: In this paper, we propose a new approach to simultaneously perform segmentation and grading of the HC to better capture the patterns of pathology occurring during AD. Based on a patch-based framework, the novel proposed grading measure estimates the similarity of the patch surrounding the voxel under study with all the patches present in different training populations. The training library used during our experiments was composed by 2 populations, 50 Cognitively Normal subjects (CN) and 50 patients with AD. Tests were completed in a leave-one-out framework. RESULTS: First, the evaluation of HC segmentation accuracy yielded a Dice's Kappa of 0.88 for CN and 0.84 for AD. Second, the proposed HC grading enables detection of AD with a success rate of 89%. Finally, a comparison of several biomarkers was investigated using a linear discriminant analysis. CONCLUSION: Using the volume and the grade of the HC at the same time resulted in an efficient patient classification with a success rate of 90%.

Background: The European Union AddNeuroMed project and the US-based Alzheimer Disease Neuroimaging Initiative (ADNI) are two large multi-centre initiatives designed to analyse and validate biomarkers for AD. This study aims to compare and combine magnetic resonance imaging (MRI) data from the two study cohorts using an automated image analysis pipeline and multivariate data analysis. Methods: A total of 664 subjects were included in this study (AddNeuroMed: 126 AD, 115 CTL, ADNI: 194 AD, 229 CTL) Data acquisition for the AddNeuroMed project was set up to be compatible with the ADNI study and the high resolution sagital 3D T1w MP-RAGE datasets used for image analysis. Regional segmentation of the brain was carried out using the multi-scale ANIMAL image analysis technique (Automated Non-linear Image Matching and Anatomical Labeling). Cortical thickness measurements were performed using CLASP. A total of 24 measures were pooled together for multivariate analysis using the OPLS method (orthogonal partial least squares). Models were created for the two cohorts and for the combined cohorts to discriminate between AD patients and controls. Finally the ADNI cohort was used as a replication dataset to validate the model created for the AddNeuroMed cohort. Results: Using cross-validation, we achieved the following values: AddNeuroMed cohort: sensitivity = 79%, specificity = 86%; ADNI cohort: sensitivity = 79%, specificity = 87%; both cohorts combined: sensitivity = 83%, specificity = 83%. Using the AddNeuroMed cohort as a training set and validating the model with the ADNI cohort resulted in a sensitivity of 78% and specificity of 87%. All three models created showed very similar results. Examples of important variables for discriminating between AD and CTL included temporal lobe grey matter volume, total CSF volume and mean cortical thickness. Conclusions: Multivariate data analysis is a powerful tool for distinguishing between different patient groups. The AddNeuroMed, ADNI and combined cohorts showed similar patterns of atrophy and the predictive power was very similar. This demonstrates that the methods used are robust and that large data sets can be combined if MRI imaging protocols are carefully aligned.

In Magnetic Resonance Imaging low-resolution images are routinely interpolated to decrease voxel size and improve apparent resolution. However, classical interpolation techniques are not able to recover the high frequency information lost during the acquisition process. In the present paper a new superresolution method is proposed to recover such information using coplanar high resolution images. The proposed methodology takes benefit from the fact that in typical clinical settings both high and low-resolution images of different types are taken from the same subject. These available high resolution images can be used to improve effectively the resolution of other coplanar lower resolution images. Experiments on synthetic and real data are supplied to show the effectiveness of the proposed approach. A comparison with classical interpolation techniques is presented to demonstrate the improved performance of the proposed methodology over previous State-of-the-art methods.

International audience; PURPOSE: To adapt the so-called nonlocal means filter to deal with magnetic resonance (MR) images with spatially varying noise levels (for both Gaussian and Rician distributed noise). MATERIALS AND METHODS: Most filtering techniques assume an equal noise distribution across the image. When this assumption is not met, the resulting filtering becomes suboptimal. This is the case of MR images with spatially varying noise levels, such as those obtained by parallel imaging (sensitivity-encoded), intensity inhomogeneity-corrected images, or surface coil-based acquisitions. We propose a new method where information regarding the local image noise level is used to adjust the amount of denoising strength of the filter. Such information is automatically obtained from the images using a new local noise estimation method. RESULTS: The proposed method was validated and compared with the standard nonlocal means filter on simulated and real MRI data showing an improved performance in all cases. CONCLUSION: The new noise-adaptive method was demonstrated to outperform the standard filter when spatially varying noise is present in the images.

International audience; Graph Cuts have been shown as a powerful interactive segmentation technique in several medical domains.We propose to automate the Graph Cuts in order to automatically segment Multiple Sclerosis (MS) lesions in MRI. We replace the manual interaction with a robust EM-based approach in order to discriminate between MS lesions and the Normal Appearing Brain Tissues (NABT). Evaluation is performed in synthetic and real images showing good agreement between the automatic segmentation and the target segmentation. We compare our algorithm with the state of the art techniques and with several manual segmentations. An advantage of our algorithm over previously published ones is the possibility to semi-automatically improve the segmentation due to the Graph Cuts interactive feature.

International audience; We present a new tool for segmenting multiple sclerosis lesions that can take advantage of the complementary modalities we usually use for this purpose. Based on the integration of multi channel information in a scale-space paradigm, its optimization by graph cuts provides a powerful and accurate tool. After presenting the mathematical and physical background on which is based the spectral gradient, we introduce its computational optimization by the graph cuts approach. The validation on synthetic and real data shows both its accuracy and reliability on different sets of standard MRI of sequences.

International audience; A new algorithm for segmentation of white matter lesions and normal appearing brain tissues in Multiple Sclerosis (MS) is presented. Two different segmentation methods are combined in order to have a better and more meaningful segmentation. On the one hand, a local segmentation approach, the Mean Shift, is used to generate local regions in our images. On the other hand, a variant of the Expectation Maximization is employed to classify these regions as Normal Appearing Brain Tissues (NABT) or lesions. Validation of this method is performed with synthetic and real data. The output is a more powerful algorithm that employs at the same time global and local information to improve image segmentation.

Cortical multiple sclerosis lesions are difficult to detect in magnetic resonance images due to poor contrast with surrounding grey matter, spatial variation in healthy grey matter and partial volume effects. We propose using an observer-independent laminar profile-based parcellation method to detect cortical lesions. Following cortical surface extraction, profiles are extended from the white matter surface to the grey matter surface. The cortex is parcellated according to profile intensity and shape features using a k-means classifier. The method is applied to a high-resolution quantitative magnetic resonance data set from a fixed post mortem multiple sclerosis brain, and validated using histology.