Generating controlled image sets in cognitive neuroscience research.

The investigation of perceptual and cognitive functions with non-invasive brain imaging methods critically depends on the careful selection of stimuli for use in experiments. For example, it must be verified that any observed effects follow from the parameter of interest (e.g. semantic category) rather than other low-level physical features (e.g. luminance, or spectral properties). Otherwise, interpretation of results is confounded. Often, researchers circumvent this issue by including additional control conditions or tasks, both of which are flawed and also prolong experiments. Here, we present some new approaches for controlling classes of stimuli intended for use in cognitive neuroscience, however these methods can be readily extrapolated to other applications and stimulus modalities. Our approach is comprised of two levels. The first level aims at equalizing individual stimuli in terms of their mean luminance. Each data point in the stimulus is adjusted to a standardized value based on a standard value across the stimulus battery. The second level analyzes two populations of stimuli along their spectral properties (i.e. spatial frequency) using a dissimilarity metric that equals the root mean square of the distance between two populations of objects as a function of spatial frequency along x- and y-dimensions of the image. Randomized permutations are used to obtain a minimal value between the populations to minimize, in a completely data-driven manner, the spectral differences between image sets. While another paper in this issue applies these methods in the case of acoustic stimuli (Aeschlimann et al., Brain Topogr 2008), we illustrate this approach here in detail for complex visual stimuli.

Access of extracellular cations to their binding sites in Na,K-ATPase: role of the second extracellular loop of the alpha subunit.

Na,K-ATPase, the main active transport system for monovalent cations in animal cells, is responsible for maintaining Na(+) and K(+) gradients across the plasma membrane. During its transport cycle it binds three cytoplasmic Na(+) ions and releases them on the extracellular side of the membrane, and then binds two extracellular K(+) ions and releases them into the cytoplasm. The fourth, fifth, and sixth transmembrane helices of the alpha subunit of Na,K-ATPase are known to be involved in Na(+) and K(+) binding sites, but the gating mechanisms that control the access of these ions to their binding sites are not yet fully understood. We have focused on the second extracellular loop linking transmembrane segments 3 and 4 and attempted to determine its role in gating. We replaced 13 residues of this loop in the rat alpha1 subunit, from E314 to G326, by cysteine, and then studied the function of these mutants using electrophysiological techniques. We analyzed the results using a structural model obtained by homology with SERCA, and ab initio calculations for the second extracellular loop. Four mutants were markedly modified by the sulfhydryl reagent MTSET, and we investigated them in detail. The substituted cysteines were more readily accessible to MTSET in the E1 conformation for the Y315C, W317C, and I322C mutants. Mutations or derivatization of the substituted cysteines in the second extracellular loop resulted in major increases in the apparent affinity for extracellular K(+), and this was associated with a reduction in the maximum activity. The changes produced by the E314C mutation were reversed by MTSET treatment. In the W317C and I322C mutants, MTSET also induced a moderate shift of the E1/E2 equilibrium towards the E1(Na) conformation under Na/Na exchange conditions. These findings indicate that the second extracellular loop must be functionally linked to the gating mechanism that controls the access of K(+) to its binding site.

The "reversed" treatment approach for synchronous liver metastases and colorectal cancer. Feasability and first results

Background: The « reversed treatment» approach inverts the treatment sequence of¦advanced synchronous colorectal liver metastases - i.e. the liver metastasis is¦treated first, followed by resection of the primary tumor. Chemotherapy is performed¦before and after liver surgery. We recently started to use a reversed treatment¦approach in selected patients. The aim of this study is to critically assess this new¦treatment modality.¦Methods: Nine patients (7 male, 2 female, mean age 62 years) benefited from this¦new treatment between November 2008 and May 2010. The data were collected¦retrospectively.¦Results: All patients responded to the neoadjuvant chemotherapy. The median¦number of liver metastases was 6 (range 1 - 22). The median size of the largest liver¦metastases was 4.3 cm (range 2.6 - 13 cm). Three patients had portal vein¦embolization prior to liver surgery. Two patients could not complete the treatment.¦One had to undergo emergency surgery for occluding colonic tumor. The second one¦showed liver recurrence before starting the adjuvant chemotherapy. The seven¦patients who completed the treatment are still alive after a median time of 27 months¦(range 17 - 37 months). Seven of them had recurrence (1 rectal, 6 liver). The median¦disease-free survival was 9 months (range 0 - 17 months).¦Conclusion: Based on our preliminary experiences, the reversed strategy shows¦encouraging results for the treatment of advanced synchronous colorectal liver¦metastases in well selected patients. The treatment was generally well tolerated and¦long term survival seems to be prolonged.

Advanced Kernel Methods For Remote Sensing Image Classification

Résumé Suite aux recentes avancées technologiques, les archives d'images digitales ont connu une croissance qualitative et quantitative sans précédent. Malgré les énormes possibilités qu'elles offrent, ces avancées posent de nouvelles questions quant au traitement des masses de données saisies. Cette question est à la base de cette Thèse: les problèmes de traitement d'information digitale à très haute résolution spatiale et/ou spectrale y sont considérés en recourant à des approches d'apprentissage statistique, les méthodes à noyau. Cette Thèse étudie des problèmes de classification d'images, c'est à dire de catégorisation de pixels en un nombre réduit de classes refletant les propriétés spectrales et contextuelles des objets qu'elles représentent. L'accent est mis sur l'efficience des algorithmes, ainsi que sur leur simplicité, de manière à augmenter leur potentiel d'implementation pour les utilisateurs. De plus, le défi de cette Thèse est de rester proche des problèmes concrets des utilisateurs d'images satellite sans pour autant perdre de vue l'intéret des méthodes proposées pour le milieu du machine learning dont elles sont issues. En ce sens, ce travail joue la carte de la transdisciplinarité en maintenant un lien fort entre les deux sciences dans tous les développements proposés. Quatre modèles sont proposés: le premier répond au problème de la haute dimensionalité et de la redondance des données par un modèle optimisant les performances en classification en s'adaptant aux particularités de l'image. Ceci est rendu possible par un système de ranking des variables (les bandes) qui est optimisé en même temps que le modèle de base: ce faisant, seules les variables importantes pour résoudre le problème sont utilisées par le classifieur. Le manque d'information étiquétée et l'incertitude quant à sa pertinence pour le problème sont à la source des deux modèles suivants, basés respectivement sur l'apprentissage actif et les méthodes semi-supervisées: le premier permet d'améliorer la qualité d'un ensemble d'entraînement par interaction directe entre l'utilisateur et la machine, alors que le deuxième utilise les pixels non étiquetés pour améliorer la description des données disponibles et la robustesse du modèle. Enfin, le dernier modèle proposé considère la question plus théorique de la structure entre les outputs: l'intègration de cette source d'information, jusqu'à présent jamais considérée en télédétection, ouvre des nouveaux défis de recherche. Advanced kernel methods for remote sensing image classification Devis Tuia Institut de Géomatique et d'Analyse du Risque September 2009 Abstract The technical developments in recent years have brought the quantity and quality of digital information to an unprecedented level, as enormous archives of satellite images are available to the users. However, even if these advances open more and more possibilities in the use of digital imagery, they also rise several problems of storage and treatment. The latter is considered in this Thesis: the processing of very high spatial and spectral resolution images is treated with approaches based on data-driven algorithms relying on kernel methods. In particular, the problem of image classification, i.e. the categorization of the image's pixels into a reduced number of classes reflecting spectral and contextual properties, is studied through the different models presented. The accent is put on algorithmic efficiency and the simplicity of the approaches proposed, to avoid too complex models that would not be used by users. The major challenge of the Thesis is to remain close to concrete remote sensing problems, without losing the methodological interest from the machine learning viewpoint: in this sense, this work aims at building a bridge between the machine learning and remote sensing communities and all the models proposed have been developed keeping in mind the need for such a synergy. Four models are proposed: first, an adaptive model learning the relevant image features has been proposed to solve the problem of high dimensionality and collinearity of the image features. This model provides automatically an accurate classifier and a ranking of the relevance of the single features. The scarcity and unreliability of labeled. information were the common root of the second and third models proposed: when confronted to such problems, the user can either construct the labeled set iteratively by direct interaction with the machine or use the unlabeled data to increase robustness and quality of the description of data. Both solutions have been explored resulting into two methodological contributions, based respectively on active learning and semisupervised learning. Finally, the more theoretical issue of structured outputs has been considered in the last model, which, by integrating outputs similarity into a model, opens new challenges and opportunities for remote sensing image processing.

MBIS: multivariate Bayesian image segmentation tool.

We present MBIS (Multivariate Bayesian Image Segmentation tool), a clustering tool based on the mixture of multivariate normal distributions model. MBIS supports multichannel bias field correction based on a B-spline model. A second methodological novelty is the inclusion of graph-cuts optimization for the stationary anisotropic hidden Markov random field model. Along with MBIS, we release an evaluation framework that contains three different experiments on multi-site data. We first validate the accuracy of segmentation and the estimated bias field for each channel. MBIS outperforms a widely used segmentation tool in a cross-comparison evaluation. The second experiment demonstrates the robustness of results on atlas-free segmentation of two image sets from scan-rescan protocols on 21 healthy subjects. Multivariate segmentation is more replicable than the monospectral counterpart on T1-weighted images. Finally, we provide a third experiment to illustrate how MBIS can be used in a large-scale study of tissue volume change with increasing age in 584 healthy subjects. This last result is meaningful as multivariate segmentation performs robustly without the need for prior knowledge.

Cluster-based active learning for compact image classification

In this paper, we consider active sampling to label pixels grouped with hierarchical clustering. The objective of the method is to match the data relationships discovered by the clustering algorithm with the user's desired class semantics. The first is represented as a complete tree to be pruned and the second is iteratively provided by the user. The active learning algorithm proposed searches the pruning of the tree that best matches the labels of the sampled points. By choosing the part of the tree to sample from according to current pruning's uncertainty, sampling is focused on most uncertain clusters. This way, large clusters for which the class membership is already fixed are no longer queried and sampling is focused on division of clusters showing mixed labels. The model is tested on a VHR image in a multiclass classification setting. The method clearly outperforms random sampling in a transductive setting, but cannot generalize to unseen data, since it aims at optimizing the classification of a given cluster structure.

A statistical shape model of the human second cervical vertebra.

PURPOSE: Statistical shape and appearance models play an important role in reducing the segmentation processing time of a vertebra and in improving results for 3D model development. Here, we describe the different steps in generating a statistical shape model (SSM) of the second cervical vertebra (C2) and provide the shape model for general use by the scientific community. The main difficulties in its construction are the morphological complexity of the C2 and its variability in the population. METHODS: The input dataset is composed of manually segmented anonymized patient computerized tomography (CT) scans. The alignment of the different datasets is done with the procrustes alignment on surface models, and then, the registration is cast as a model-fitting problem using a Gaussian process. A principal component analysis (PCA)-based model is generated which includes the variability of the C2. RESULTS: The SSM was generated using 92 CT scans. The resulting SSM was evaluated for specificity, compactness and generalization ability. The SSM of the C2 is freely available to the scientific community in Slicer (an open source software for image analysis and scientific visualization) with a module created to visualize the SSM using Statismo, a framework for statistical shape modeling. CONCLUSION: The SSM of the vertebra allows the shape variability of the C2 to be represented. Moreover, the SSM will enable semi-automatic segmentation and 3D model generation of the vertebra, which would greatly benefit surgery planning.

Optimization of Radiation Dose and Image Quality in Musculoskeletal CT: Emphasis on Iterative Reconstruction Techniques (Part 1).

Computed tomography (CT) is a modality of choice for the study of the musculoskeletal system for various indications including the study of bone, calcifications, internal derangements of joints (with CT arthrography), as well as periprosthetic complications. However, CT remains intrinsically limited by the fact that it exposes patients to ionizing radiation. Scanning protocols need to be optimized to achieve diagnostic image quality at the lowest radiation dose possible. In this optimization process, the radiologist needs to be familiar with the parameters used to quantify radiation dose and image quality. CT imaging of the musculoskeletal system has certain specificities including the focus on high-contrast objects (i.e., in CT of bone or CT arthrography). These characteristics need to be taken into account when defining a strategy to optimize dose and when choosing the best combination of scanning parameters. In the first part of this review, we present the parameters used for the evaluation and quantification of radiation dose and image quality. In the second part, we discuss different strategies to optimize radiation dose and image quality at CT, with a focus on the musculoskeletal system and the use of novel iterative reconstruction techniques.

Optimization of Radiation Dose and Image Quality in Musculoskeletal CT: Emphasis on Iterative Reconstruction Techniques (Part 2).

Computed tomography (CT) is a modality of choice for the study of the musculoskeletal system for various indications including the study of bone, calcifications, internal derangements of joints (with CT arthrography), as well as periprosthetic complications. However, CT remains intrinsically limited by the fact that it exposes patients to ionizing radiation. Scanning protocols need to be optimized to achieve diagnostic image quality at the lowest radiation dose possible. In this optimization process, the radiologist needs to be familiar with the parameters used to quantify radiation dose and image quality. CT imaging of the musculoskeletal system has certain specificities including the focus on high-contrast objects (i.e., in CT of bone or CT arthrography). These characteristics need to be taken into account when defining a strategy to optimize dose and when choosing the best combination of scanning parameters. In the first part of this review, we present the parameters used for the evaluation and quantification of radiation dose and image quality. In the second part, we discuss different strategies to optimize radiation dose and image quality of CT, with a focus on the musculoskeletal system and the use of novel iterative reconstruction techniques.

Systematic evaluation of matrix effects in hydrophilic interaction chromatography versus reversed phase liquid chromatography coupled to mass spectrometry.

Reversed phase liquid chromatography (RPLC) coupled to mass spectrometry (MS) is the gold standard technique in bioanalysis. However, hydrophilic interaction chromatography (HILIC) could represent a viable alternative to RPLC for the analysis of polar and/or ionizable compounds, as it often provides higher MS sensitivity and alternative selectivity. Nevertheless, this technique can be also prone to matrix effects (ME). ME are one of the major issues in quantitative LC-MS bioanalysis. To ensure acceptable method performance (i.e., trueness and precision), a careful evaluation and minimization of ME is required. In the present study, the incidence of ME in HILIC-MS/MS and RPLC-MS/MS was compared for plasma and urine samples using two representative sets of 38 pharmaceutical compounds and 40 doping agents, respectively. The optimal generic chromatographic conditions in terms of selectivity with respect to interfering compounds were established in both chromatographic modes by testing three different stationary phases in each mode with different mobile phase pH. A second step involved the assessment of ME in RPLC and HILIC under the best generic conditions, using the post-extraction addition method. Biological samples were prepared using two different sample pre-treatments, i.e., a non-selective sample clean-up procedure (protein precipitation and simple dilution for plasma and urine samples, respectively) and a selective sample preparation, i.e., solid phase extraction for both matrices. The non-selective pretreatments led to significantly less ME in RPLC vs. HILIC conditions regardless of the matrix. On the contrary, HILIC appeared as a valuable alternative to RPLC for plasma and urine samples treated by a selective sample preparation. Indeed, in the case of selective sample preparation, the compounds influenced by ME were different in HILIC and RPLC, and lower and similar ME occurrence was generally observed in RPLC vs. HILIC for urine and plasma samples, respectively. The complementary of both chromatographic modes was also demonstrated, as ME was observed only scarcely for urine and plasma samples when selecting the most appropriate chromatographic mode.

The fourth extracellular loop of the alpha subunit of Na,K-ATPase. Functional evidence for close proximity with the second extracellular loop.

Na,K-ATPase is the main active transport system that maintains the large gradients of Na(+) and K(+) across the plasma membrane of animal cells. The crystal structure of a K(+)-occluding conformation of this protein has been recently published, but the movements of its different domains allowing for the cation pumping mechanism are not yet known. The structure of many more conformations is known for the related calcium ATPase SERCA, but the reliability of homology modeling is poor for several domains with low sequence identity, in particular the extracellular loops. To better define the structure of the large fourth extracellular loop between the seventh and eighth transmembrane segments of the alpha subunit, we have studied the formation of a disulfide bond between pairs of cysteine residues introduced by site-directed mutagenesis in the second and the fourth extracellular loop. We found a specific pair of cysteine positions (Y308C and D884C) for which extracellular treatment with an oxidizing agent inhibited the Na,K pump function, which could be rapidly restored by a reducing agent. The formation of the disulfide bond occurred preferentially under the E2-P conformation of Na,K-ATPase, in the absence of extracellular cations. Using recently published crystal structure and a distance constraint reproducing the existence of disulfide bond, we performed an extensive conformational space search using simulated annealing and showed that the Tyr(308) and Asp(884) residues can be in close proximity, and simultaneously, the SYGQ motif of the fourth extracellular loop, known to interact with the extracellular domain of the beta subunit, can be exposed to the exterior of the protein and can easily interact with the beta subunit.

Urban image classification with semisupervised multiscale cluster kernels

This paper presents a semisupervised support vector machine (SVM) that integrates the information of both labeled and unlabeled pixels efficiently. Method's performance is illustrated in the relevant problem of very high resolution image classification of urban areas. The SVM is trained with the linear combination of two kernels: a base kernel working only with labeled examples is deformed by a likelihood kernel encoding similarities between labeled and unlabeled examples. Results obtained on very high resolution (VHR) multispectral and hyperspectral images show the relevance of the method in the context of urban image classification. Also, its simplicity and the few parameters involved make the method versatile and workable by unexperienced users.