Uncertainty Quantification Of A Semi-Supervised Support Vector Regression Reservoir Model

Uncertainty quantification of petroleum reservoir models is one of the present challenges, which is usually approached with a wide range of geostatistical tools linked with statistical optimisation or/and inference algorithms. Recent advances in machine learning offer a novel approach to model spatial distribution of petrophysical properties in complex reservoirs alternative to geostatistics. The approach is based of semisupervised learning, which handles both ?labelled? observed data and ?unlabelled? data, which have no measured value but describe prior knowledge and other relevant data in forms of manifolds in the input space where the modelled property is continuous. Proposed semi-supervised Support Vector Regression (SVR) model has demonstrated its capability to represent realistic geological features and describe stochastic variability and non-uniqueness of spatial properties. On the other hand, it is able to capture and preserve key spatial dependencies such as connectivity of high permeability geo-bodies, which is often difficult in contemporary petroleum reservoir studies. Semi-supervised SVR as a data driven algorithm is designed to integrate various kind of conditioning information and learn dependences from it. The semi-supervised SVR model is able to balance signal/noise levels and control the prior belief in available data. In this work, stochastic semi-supervised SVR geomodel is integrated into Bayesian framework to quantify uncertainty of reservoir production with multiple models fitted to past dynamic observations (production history). Multiple history matched models are obtained using stochastic sampling and/or MCMC-based inference algorithms, which evaluate posterior probability distribution. Uncertainty of the model is described by posterior probability of the model parameters that represent key geological properties: spatial correlation size, continuity strength, smoothness/variability of spatial property distribution. The developed approach is illustrated with a fluvial reservoir case. The resulting probabilistic production forecasts are described by uncertainty envelopes. The paper compares the performance of the models with different combinations of unknown parameters and discusses sensitivity issues.

Desenvolupament d'algorismes biològicament inspirats basats en Support Vector Machines

L'objectiu d'aquest projecte ha estat el desenvolupament d'algorismes biològicament inspirats per a l'olfacció artificial. Per a assolir-lo ens hem basat en el paradigma de les màquines amb suport vectorial. Hem construit algoritmes que imitaven els processos computacionals dels diferents sistemes que formen el sistema olfactiu dels insectes, especialment de la llagosta Schistocerca gregaria. Ens hem centrat en el lòbuls de les antenes, i en el cos fungiforme. El primer està considerat un dispositiu de codificació de les olors, que a partir de la resposta temporal dels receptors olfactius a les antenes genera un patró d'activació espaial i temporal. Quant al cos fungiforme es considera que la seva funció és la d'una memòria per als olors, així com un centre per a la integració multi-sensorial. El primer pas ha estat la construcció de models detallats dels dos sistemes. A continuació, hem utilitzat aquests models per a processar diferents tipus de senyals amb l'objectiu de abstraure els principis computacionals subjacents. Finalment, hem avaluat les capacitats d'aquests models abstractes, i els hem utilitzat per al processat de dades provinents de sensors de gasos. Els resultats mostren que el models abstractes tenen millor comportament front el soroll i més capacitat d'emmagatzematge de records que altres models més clàssics, com ara les memòries associatives de Hopfield o fins i tot en determinades circumstàncies que les mateixes Support Vector Machines.

The EPAGE internet guideline as a decision support tool for determining the appropriateness of colonoscopy

RÉSUMÉ Contexte : Peu d'études ont examiné la façon dont les médecins appréhendent les guidelines, et encore moins celle dont ils perçoivent de tels guidelines disponibles sur Internet. Cette étude évalue l'acceptation par les médecins d'un guideline électronique portant sur l'adéquation de la colonoscopie. Méthode : Des gastroentérologues participant à une étude observationnelle internationale ont consulté un guideline électronique pour une série consécutive de patients adressés pour une colonoscopie. Le guideline a été élaboré par le Panel Européen sur l'Adéquation de l'Endoscopie Gastro-intestinale (EPAGE en version anglaise), utilisant une méthode validée (RAND). Les opinions des médecins sur le guideline, sur le site Internet et sur les perspectives d'utilisation ont été recueillies au moyen de questionnaires. Résultats : 289 patients ont été inclus dans l'étude. Le temps moyen pour consulter le site Internet a été de 1.8 min et 86% des médecins l'ont considéré comme simple à utiliser. Les recommandations ont été facilement localisées pour 82% des patients et les médecins étaient d'accord avec l'adéquation de la colonoscopie dans 86% des cas. Selon les critères EPAGE, la colonoscopie était appropriée, incertaine et inappropriée, respectivement chez 59, 28 et 13% des patients. Conclusions : Le guideline EPAGE a été considéré comme acceptable et simple à utiliser. L'utilisation, l'utilité et la pertinence du site Internet a été jugée comme acceptable. Son utilisation effective dépendra cependant de la levée de certains obstacles au niveau organisationnel et culturel.

Statistical discrimination of steroid profiles in doping control with support vector machines.

Due to their performance enhancing properties, use of anabolic steroids (e.g. testosterone, nandrolone, etc.) is banned in elite sports. Therefore, doping control laboratories accredited by the World Anti-Doping Agency (WADA) screen among others for these prohibited substances in urine. It is particularly challenging to detect misuse with naturally occurring anabolic steroids such as testosterone (T), which is a popular ergogenic agent in sports and society. To screen for misuse with these compounds, drug testing laboratories monitor the urinary concentrations of endogenous steroid metabolites and their ratios, which constitute the steroid profile and compare them with reference ranges to detect unnaturally high values. However, the interpretation of the steroid profile is difficult due to large inter-individual variances, various confounding factors and different endogenous steroids marketed that influence the steroid profile in various ways. A support vector machine (SVM) algorithm was developed to statistically evaluate urinary steroid profiles composed of an extended range of steroid profile metabolites. This model makes the interpretation of the analytical data in the quest for deviating steroid profiles feasible and shows its versatility towards different kinds of misused endogenous steroids. The SVM model outperforms the current biomarkers with respect to detection sensitivity and accuracy, particularly when it is coupled to individual data as stored in the Athlete Biological Passport.

Infornut® Process; improves accessibility to diagnosis and nutritional support for the malnourished hospitalized patient; impact on management indicators; two-year assessment.

Introduction: The high prevalence of disease-related hospital malnutrition justifies the need for screening tools and early detection in patients at risk for malnutrition, followed by an assessment targeted towards diagnosis and treatment. At the same time there is clear undercoding of malnutrition diagnoses and the procedures to correct it Objectives: To describe the INFORNUT program/ process and its development as an information system. To quantify performance in its different phases. To cite other tools used as a coding source. To calculate the coding rates for malnutrition diagnoses and related procedures. To show the relationship to Mean Stay, Mortality Rate and Urgent Readmission; as well as to quantify its impact on the hospital Complexity Index and its effect on the justification of Hospitalization Costs. Material and methods: The INFORNUT® process is based on an automated screening program of systematic detection and early identification of malnourished patients on hospital admission, as well as their assessment, diagnoses, documentation and reporting. Of total readmissions with stays longer than three days incurred in 2008 and 2010, we recorded patients who underwent analytical screening with an alert for a medium or high risk of malnutrition, as well as the subgroup of patients in whom we were able to administer the complete INFORNUT® process, generating a report for each.

Maladies vasculaires: quels dépistages au cabinet du praticien [Vascular diseases: what screening could be done in the office?]

The screening of vascular pathologies in physician offices starts with precise medical history and clinical exam. Tools like the Edinburgh Claudication Questionnaire for the peripheral artery disease or the Wells score for the probability of a thromboembolic event are useful. The measure of the ankle brachial index, the D-dimers or any other biological screening are complementary. In the presence of pathological features, it is recommended to organise a specialised consultation in order to precise diagnosis, treatment and follow-up. The screening of a vascular disease is interesting not only for the management of local symptoms, but also for the associated systemic pathologies to provide a preventive medicine of good quality.

Soins palliatifs et soins de support: à la frontière de la toute-puissance médicale [Palliative and supportive care: at the frontiers of medical omnipotence].

Cancer patients have physical, social, spiritual and emotional needs. They may suffer from severe physical symptoms, from social isolation, spiritual abandonment, and emotions such as sadness and anxiety, or feelings of deception, helplessness, anger and guilt. In some of them, the disease is rapidly progressing and ultimately they die. Their demanding care evokes intense feelings in health care providers, the more since these incurable patients represent a challenge, which could be condensed under the heading "the challenge of medical omnipotence". We suppose that the way health care providers cope with these circumstances has a profound influence on the way these patients are cared for. The attitudes towards the emerging heterogeneous movement of palliative and supportive care and towards its different models of implementation can be viewed from this point of view. We try to demonstrate these interrelations and to discuss the danger that may arise if they remain obscure and unreflected.

A Clinical Decision Support using a Terminology Server to improve Patient Safety.

Clinical Decision Support Systems (CDSS) are software applications that support clinicians in making healthcare decisions providing relevant information for individual patients about their specific conditions. The lack of integration between CDSS and Electronic Health Record (EHR) has been identified as a significant barrier to CDSS development and adoption. Andalusia Healthcare Public System (AHPS) provides an interoperable health information infrastructure based on a Service Oriented Architecture (SOA) that eases CDSS implementation. This paper details the deployment of a CDSS jointly with the deployment of a Terminology Server (TS) within the AHPS infrastructure. It also explains a case study about the application of decision support to thromboembolism patients and its potential impact on improving patient safety. We will apply the inSPECt tool proposal to evaluate the appropriateness of alerts in this scenario.

Decision Support Methods for Global Optomization

Immobile location-allocation (LA) problems is a type of LA problem that consists in determining the service each facility should offer in order to optimize some criterion (like the global demand), given the positions of the facilities and the customers. Due to the complexity of the problem, i.e. it is a combinatorial problem (where is the number of possible services and the number of facilities) with a non-convex search space with several sub-optimums, traditional methods cannot be applied directly to optimize this problem. Thus we proposed the use of clustering analysis to convert the initial problem into several smaller sub-problems. By this way, we presented and analyzed the suitability of some clustering methods to partition the commented LA problem. Then we explored the use of some metaheuristic techniques such as genetic algorithms, simulated annealing or cuckoo search in order to solve the sub-problems after the clustering analysis

Four main virotypes among extended-spectrum-β-lactamase-producing isolates of Escherichia coli O25b:H4-B2-ST131: bacterial, epidemiological, and clinical characteristics.

A total of 1,021 extended-spectrum-β-lactamase-producing Escherichia coli (ESBLEC) isolates obtained in 2006 during a Spanish national survey conducted in 44 hospitals were analyzed for the presence of the O25b:H4-B2-ST131 (sequence type 131) clonal group. Overall, 195 (19%) O25b-ST131 isolates were detected, with prevalence rates ranging from 0% to 52% per hospital. Molecular characterization of 130 representative O25b-ST131 isolates showed that 96 (74%) were positive for CTX-M-15, 15 (12%) for CTX-M-14, 9 (7%) for SHV-12, 6 (5%) for CTX-M-9, 5 (4%) for CTX-M-32, and 1 (0.7%) each for CTX-M-3 and the new ESBL enzyme CTX-M-103. The 130 O25b-ST131 isolates exhibited relatively high virulence scores (mean, 14.4 virulence genes). Although the virulence profiles of the O25b-ST131 isolates were fairly homogeneous, they could be classified into four main virotypes based on the presence or absence of four distinctive virulence genes: virotypes A (22%) (afa FM955459 positive, iroN negative, ibeA negative, sat positive or negative), B (31%) (afa FM955459 negative, iroN positive, ibeA negative, sat positive or negative), C (32%) (afa FM955459 negative, iroN negative, ibeA negative, sat positive), and D (13%) (afa FM955459 negative, iroN positive or negative, ibeA positive, sat positive or negative). The four virotypes were also identified in other countries, with virotype C being overrepresented internationally. Correspondingly, an analysis of XbaI macrorestriction profiles revealed four major clusters, which were largely virotype specific. Certain epidemiological and clinical features corresponded with the virotype. Statistically significant virotype-specific associations included, for virotype B, older age and a lower frequency of infection (versus colonization), for virotype C, a higher frequency of infection, and for virotype D, younger age and community-acquired infections. In isolates of the O25b:H4-B2-ST131 clonal group, these findings uniquely define four main virotypes, which are internationally distributed, correspond with pulsed-field gel electrophoresis (PFGE) profiles, and exhibit distinctive clinical-epidemiological associations.