Bayesian classification criterion for forensic multivariate data

This study presents a classification criteria for two-class Cannabis seedlings. As the cultivation of drug type cannabis is forbidden in Switzerland, law enforcement authorities regularly ask laboratories to determine cannabis plant's chemotype from seized material in order to ascertain that the plantation is legal or not. In this study, the classification analysis is based on data obtained from the relative proportion of three major leaf compounds measured by gas-chromatography interfaced with mass spectrometry (GC-MS). The aim is to discriminate between drug type (illegal) and fiber type (legal) cannabis at an early stage of the growth. A Bayesian procedure is proposed: a Bayes factor is computed and classification is performed on the basis of the decision maker specifications (i.e. prior probability distributions on cannabis type and consequences of classification measured by losses). Classification rates are computed with two statistical models and results are compared. Sensitivity analysis is then performed to analyze the robustness of classification criteria.

"Etude rétrospective des cas de vasculite cutanée identifiés au laboratoire de dermopathologie du CHUV de 2006 à 2010"

CONTEXTE Les vasculites représentent un groupe de maladies inflammatoires touchant tous les vaisseaux, des grosses artères aux artères de plus petit calibre, en passant par les capillaires et les veines, de toutes tailles également. La taille des vaisseaux atteints est d'une importance cruciale pour le diagnostic puisque la classification actuelle en dépend. Sur le plan histopathologique, la vasculite est une réaction inflammatoire qui atteint la paroi de vaisseaux, et qui va entraîner alors une ischémie, puis une nécrose du vaisseau lui-­‐ même voire des structures environnantes ; les caractéristiques microscopiques de ces lésions permettent souvent d'aider à préciser le groupe auquel appartient la vasculite. Cependant, il n'est pas toujours facile ni possible d'obtenir un échantillon de tissu, une biopsie, selon l'endroit concerné (cerveau, régions profondes et difficilement accessibles...). OBJECTIFS Dans les atteintes de type « vasculite », la peau est fréquemment un organe cible. Le grand avantage est son accès direct et aisé, notamment dans le but de pratiquer un examen histopathologique par le biais d'une biopsie, qui peut se pratiquer de manière superficielle ou profonde, révélant alors des vaisseaux de différents calibres. Face à un tableau clinique parfois complexe et menaçant, il est important de pouvoir orienter le diagnostic rapidement. L'objectif de ce travail est donc de déterminer quelles sont l'utilité et l'apport de la biopsie cutanée dans le diagnostic des vasculites, et autrement dit de déterminer s'il vaut la peine de pratiquer ce genre d'examen, ou s'il vaut mieux privilégier d'autres approches plus rentables sur ce plan-­‐là. METHODES Grâce à l'aide des dermatologues et à la base de données du Service de dermatopathologie du CHUV, nous nous proposons de faire une étude rétrospective et systématique, sur les 5 années passées, de toutes les biopsies cutanées compatibles avec le diagnostic de vasculite. RESULTATS ESCOMPTES Ils sont de deux ordres : déterminer d'une part si le résultat de la biopsie était compatible avec le diagnostic clinique retenu chez les patients ayant subi ce type d'examen. Cela nous permettra alors d'établir des recommandations pour la réalisation éventuelle d'une biopsie cutanée dans ce type de situation. Il sera en effet capital de savoir si ce geste est utile et, le cas échéant, s'il faut demander une biopsie profonde ou si une biopsie superficielle est suffisante. Ce travail nous permettra d'autre part de revoir les caractéristiques cutanées -­‐ sur le plan histopathologique -­‐ des vasculites, et de déterminer aussi leur fréquence. Dans l'idéal, ce travail devrait permettre de faire la revue et la mise à jour des connaissances dans le domaine des vasculites cutanées, utiles aux praticiens notamment.

Structured output SVM for remote sensing image classification

In the recent years, kernel methods have revealed very powerful tools in many application domains in general and in remote sensing image classification in particular. The special characteristics of remote sensing images (high dimension, few labeled samples and different noise sources) are efficiently dealt with kernel machines. In this paper, we propose the use of structured output learning to improve remote sensing image classification based on kernels. Structured output learning is concerned with the design of machine learning algorithms that not only implement input-output mapping, but also take into account the relations between output labels, thus generalizing unstructured kernel methods. We analyze the framework and introduce it to the remote sensing community. Output similarity is here encoded into SVM classifiers by modifying the model loss function and the kernel function either independently or jointly. Experiments on a very high resolution (VHR) image classification problem shows promising results and opens a wide field of research with structured output kernel methods.

Enquête auprès des usagers de drogue, clients des structures à bas-seuil d'accès (SBS) en Suisse: tendances 1993-2011

[Table des matières] Résumés. - 2. Introduction. - 3. Méthode pour l'enquête clientèle. - 4. Tendances pour l'ensemble des répondants: Caractéristiques sociodémographiques des usagers ; Consommation de substances et traitement de substitution ; Indicateurs de pratique de l'injection ; Activité sexuelle et comportement préventif; Indicateurs de l'état de santé ; Incarcération en lien avec la drogue et consommation en prison. - 5. Tendances pour les injecteurs de drogue au cours de la vie: Caractéristiques sociodémographiques des usagers injecteurs de drogue ; Consommation de substances et traitement de substitution ; Indicateurs de pratique de l'injection ; Partage de seringues et de matériel ; Activité sexuelle et comportement préventif ; Indicateurs de l'état de santé ; Incarcération en lien avec la drogue et consommation en prison. - 6. Tendances pour les consommateurs de drogue qui ne se sont jamais injecté de drogue au cours de la vie: Caractéristiques sociodémographiques des usagers non-injecteurs de drogue ; Consommation de substances et traitement de substitution ; Activité sexuelle et comportement préventif ; Indicateurs de l'état de santé ; Incarcération en lien avec la drogue et consommation en prison. - 7. Références

Rehabilitation needs assessment in persons with spinal cord injury following the 2010 earthquake in Haiti: a pilot study using an ICF-based tool.

OBJECTIVE: The aim of this pilot study was to describe problems in functioning and associated rehabilitation needs in persons with spinal cord injury after the 2010 earthquake in Haiti by applying a newly developed tool based on the International Classification of Functioning, Disability and Health (ICF). DESIGN: Pilot study. SUBJECTS: Eighteen persons with spinal cord injury (11 women, 7 men) participated in the needs assessment. Eleven patients had complete lesions (American Spinal Injury Association Impairment Scale; AIS A), one patient had tetraplegia. METHODS: Data collection included information from the International Spinal Cord Injury Core Data Set and a newly developed needs assessment tool based on ICF Core Sets. This tool assesses the level of functioning, the corresponding rehabilitation need, and required health professional. Data were summarized using descriptive statistics. RESULTS: In body functions and body structures, patients showed typical problems following spinal cord injury. Nearly all patients showed limitations and restrictions in their activities and participation related to mobility, self-care and aspects of social integration. Several environmental factors presented barriers to these limitations and restrictions. However, the availability of products and social support were identified as facilitators. Rehabilitation needs were identified in nearly all aspects of functioning. To address these needs, a multidisciplinary approach would be needed. CONCLUSION: This ICF-based needs assessment provided useful information for rehabilitation planning in the context of natural disaster. Future studies are required to test and, if necessary, adapt the assessment.

Novel H3K4me3 marks are enriched at human- and chimpanzee-specific cytogenetic structures.

Human and chimpanzee genomes are 98.8% identical within comparable sequences. However, they differ structurally in nine pericentric inversions, one fusion that originated human chromosome 2, and content and localization of heterochromatin and lineage-specific segmental duplications. The possible functional consequences of these cytogenetic and structural differences are not fully understood and their possible involvement in speciation remains unclear. We show that subtelomeric regions-regions that have a species-specific organization, are more divergent in sequence, and are enriched in genes and recombination hotspots-are significantly enriched for species-specific histone modifications that decorate transcription start sites in different tissues in both human and chimpanzee. The human lineage-specific chromosome 2 fusion point and ancestral centromere locus as well as chromosome 1 and 18 pericentric inversion breakpoints showed enrichment of human-specific H3K4me3 peaks in the prefrontal cortex. Our results reveal an association between plastic regions and potential novel regulatory elements.

A cross-cultural study of the higher-order structures underlying personality disorders in French-speaking Africa and Switzerland

Most studies about the higher-order dimensions to be considered in order to parsimoniously describe Personality Disorders (PDs) have identified between two and four factors but there is still no consensus about their exact number. In this context, the cultural stability of these structures might be a criterion to be considered. The aim of this study was to identify stable higher-order structures of PD traits in a French-speaking African and Swiss sample (N = 2,711). All subject completed the IPDE screening questionnaire. Using Everett's criterion and conducting a series of principal component analyses, a cross-culturally stable two- and four-factor structure were identified, associated with a total congruence coefficient of respectively .98 and .94 after Procrustes rotation. Moreover, these two structures were also highly replicable across the four African regions considered, North Africa, West Africa, Central Africa, and Mauritius, with a mean total congruence coefficient of respectively .97 and .87. The four-factor structure presented the advantage of being similar to Livesely's four components and of describing the ten PDs more accurately.

Comparing Multilevel Clustering Methods on Weighted Graphs : The Case of Worldwide Air Passenger Traffic 2000-2004

Specific properties emerge from the structure of large networks, such as that of worldwide air traffic, including a highly hierarchical node structure and multi-level small world sub-groups that strongly influence future dynamics. We have developed clustering methods to understand the form of these structures, to identify structural properties, and to evaluate the effects of these properties. Graph clustering methods are often constructed from different components: a metric, a clustering index, and a modularity measure to assess the quality of a clustering method. To understand the impact of each of these components on the clustering method, we explore and compare different combinations. These different combinations are used to compare multilevel clustering methods to delineate the effects of geographical distance, hubs, network densities, and bridges on worldwide air passenger traffic. The ultimate goal of this methodological research is to demonstrate evidence of combined effects in the development of an air traffic network. In fact, the network can be divided into different levels of âeurooecohesionâeuro, which can be qualified and measured by comparative studies (Newman, 2002; Guimera et al., 2005; Sales-Pardo et al., 2007).