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

This paper tackles an important aspect of the variational problem underlying active contours: optimization by gradient flows. Classically, the definition of a gradient depends directly on the choice of an inner product structure. This consideration is largely absent from the active contours literature. Most authors, explicitely or implicitely, assume that the space of admissible deformations is ruled by the canonical L 2 inner product. The classical gradient flows reported in the literature are relative to this particular choice. Here, we investigate the relevance of using (i) other inner products, yielding other gradient descents, and (ii) other minimizing flows not deriving from any inner product. In particular, we show how to induce different degrees of spatial consistency into the minimizing flow, in order to decrease the probability of getting trapped into irrelevant local minima. We report numerical experiments indicating that the sensitivity of the active contours method to initial conditions, which seriously limits its applicability and efficiency, is alleviated by our application-specific spatially coherent minimizing flows. We show that the choice of the inner product can be seen as a prior on the deformation fields and we present an extension of the definition of the gradient toward more general priors.

The introduction of arterial spin labelling (ASL) techniques in magnetic resonance imaging (MRI) has made feasible a non-invasive measurement of the cerebral blood flow (CBF). However, to date, the low signal-to-noise ratio of ASL gives us no option but to repeat the acquisition to accumulate enough data in order to get a reliable signal. The perfusion signal is then usually extracted by averaging across the repetitions. But the sample mean is very sensitive to outliers. A single incorrect observation can therefore be the source of strong detrimental effects on the perfusion-weighted image estimated with the sample mean. We propose to estimate robust ASL CBF maps with M-estimators to overcome the deleterious effects of outliers. The behavior of this method is compared to z-score thresholding as recommended in Tan et al. (Journal of Magnetic Resonance Imaging 2009;29(5):1134-9.). Validation on simulated and real data is provided. Quantitative validation is undertaken by measuring the correlation with the most widespread technique to measure perfusion with MRI: dynamic susceptibility weighted contrast imaging.

In this paper, patient-specific perfusion abnormalities in Arterial Spin Labeling (ASL) were identified by comparing a single patient to a group of healthy controls using a mixed-effect hierarchical General Linear Model (GLM). Two approaches are currently in use to solve hierarchical GLMs: (1) the homoscedastic approach assumes homogeneous variances across subjects and (2) the heteroscedastic approach is theoretically more efficient in the presence of heterogeneous variances but algorithmically more demanding. In practice, in functional magnetic resonance imaging studies, the superiority of the heteroscedastic approach is still under debate. Due to the low signal-to-noise ratio of ASL sequences, within-subject variances have a significant impact on the estimated perfusion maps and the heteroscedastic model might be better suited in this context. In this paper we studied how the homoscedastic and heteroscedastic approaches behave in terms of specificity and sensitivity in the detection of patient-specific ASL perfusion abnormalities. Validation was undertaken on a dataset of 25 patients diagnosed with brain tumors and 36 healthy volunteers. We showed evidence of heterogeneous within-subject variances in ASL and pointed out an increased false positive rate of the homoscedastic model. In the detection of patient-specific brain perfusion abnormalities with ASL, modeling heterogeneous variances increases the sensitivity at the same specificity level.

This paper presents a sparse representation and an adaptive dictionary learning based method for automated classification of multiple sclerosis (MS) lesions in magnetic resonance (MR) images. Manual delineation of MS lesions is a time-consuming task, requiring neuroradiology experts to analyze huge volume of MR data. This, in addition to the high intra- and inter-observer variability necessitates the requirement of automated MS lesion classification methods. Among many image representation models and classification methods that can be used for such purpose, we investigate the use of sparse modeling. In the recent years, sparse representation has evolved as a tool in modeling data using a few basis elements of an over-complete dictionary and has found applications in many image processing tasks including classification. We propose a supervised classification approach by learning dictionaries specific to the lesions and individual healthy brain tissues, which include white matter (WM), gray matter (GM) and cerebrospinal fluid (CSF). The size of the dictionaries learned for each class plays a major role in data representation but it is an even more crucial element in the case of competitive classification. Our approach adapts the size of the dictionary for each class, depending on the complexity of the underlying data. The algorithm is validated using 52 multi-sequence MR images acquired from 13 MS patients. The results demonstrate the effectiveness of our approach in MS lesion classification.

Measures of brain activity through functional magnetic resonance imaging (fMRI) or electroencephalography (EEG), two complementary modalities, are ground solutions in the context of neurofeedback (NF) mechanisms for brain rehabilitation protocols. While NF-EEG (in which real-time neurofeedback scores are computed from EEG signals) has been explored for a very long time, NF-fMRI (in which real-time neurofeedback scores are computed from fMRI signals) appeared more recently and provides more robust results and more specific brain training. Using fMRI and EEG simultaneously for bi-modal neurofeedback sessions (NF-EEG-fMRI, in which real-time neurofeedback scores are computed from fMRI and EEG) is very promising for the design of brain rehabilitation protocols. However, fMRI is cumbersome and more exhausting for patients. The original contribution of this paper concerns the prediction of bi-modal NF scores from EEG recordings only, using a training phase where EEG signals as well as the NF-EEG and NF-fMRI scores are available. We propose a sparse regression model able to exploit EEG only to predict NF-fMRI or NF-EEG-fMRI in motor imagery tasks. We compared different NF-predictors stemming from the proposed model. We showed that predicting NF-fMRI scores from EEG signals adds information to NF-EEG scores and significantly improves the correlation with bi-modal NF sessions compared to classical NF-EEG scores.

Background and Purpose-Unenhanced time-resolved spin-labeled magnetic resonance angiography enables hemodynamic quantification in arteriovenous malformations (AVMs). Our purpose was to identify quantitative parameters that discriminate among different AVM components and to relate hemodynamic patterns with rupture risk. Methods-Sixteen patients presenting with AVMs (7 women, 9 men; mean age 37.1±15.9 years) were assigned to the high rupture risk or low rupture risk group according to anatomic AVM characteristics and rupture history. High temporal resolution (<70 ms) unenhanced time-resolved spin-labeled magnetic resonance angiography was performed on a 3-T MR system. After dedicated image processing, hemodynamic quantitative parameters were computed. T tests were used to compare quantitative parameters among AVM components, between the high rupture risk and low rupture risk groups, and between the hemorrhagic and nonhemorrhagic groups. Results-Among the quantitative parameters, time-to-peak (P<0.001) and maximum outflow gradient (P=0.01) allowed discriminating various intranidal flow patterns with significantly different values between feeding arteries and draining veins. With 9 AVMs classified into the high rupture risk group (whose 6 were hemorrhagic) and 7 into the low rupture risk group, the observed venous-to-arterial time-to-peak ratio was significantly lower in the high rupture risk (P=0.003) and hemorrhagic (P=0.001) groups. Conclusions-Unenhanced

Études de communication langages, information, médiations 40 | 2013 Epistémologies, théories et pratiques professionnelles en communication des organisations Quand l'information échappe à ses créateurs. Le cas de l'artificialisation des terres agricoles en Languedoc-Roussillon