Multitask learning of environmental spatial data

The present research deals with an application of artificial neural networks for multitask learning from spatial environmental data. The real case study (sediments contamination of Geneva Lake) consists of 8 pollutants. There are different relationships between these variables, from linear correlations to strong nonlinear dependencies. The main idea is to construct a subsets of pollutants which can be efficiently modeled together within the multitask framework. The proposed two-step approach is based on: 1) the criterion of nonlinear predictability of each variable ?k? by analyzing all possible models composed from the rest of the variables by using a General Regression Neural Network (GRNN) as a model; 2) a multitask learning of the best model using multilayer perceptron and spatial predictions. The results of the study are analyzed using both machine learning and geostatistical tools.

Influence of compliance to respiratory weaning consensus conference criteria on respiratory outcome within 48h after planned extubation

INTRODUCTION. A two-step assessment (readiness to wean (RW) followed by spontaneousbreathing trial (SBT)) of predefined criteria is recommended before planned extubation(PE)1.OBJECTIVES. We aimed to evaluate if compliance to all guideline criteria was associatedwith better respiratory outcome within 48 h after PE.METHODS. The data (extracted from our clinical information system) of 458 consecutivepatients who underwent PE after C48 h of invasive ventilation in our medico-surgical ICUwere analyzed. We evaluated compliance with guidelines [1] regarding respiratory rate, tidalvolume, PaO2, FiO2, PEEP, PaCO2, pH, heart rate, systolic arterial pressure and arrhythmiaduringRWand SBT assessment (RW and SBT within 2 h). A patient was classified as RW+ ifallRWcriteria were fulfilled andRW-if at least 1 criterion was violated. The same approachwas used to define SBT+ and SBT- patients. During the 48 h following PE, we assessed theoccurrence of post-PE respiratory failure (PRF) (defined as the presence of at least 1 consensuscriterion of respiratory failure [1]), reintubation (after NIV failure or because of immediateintubation criteria) and cumulative duration of post-PE ventilation (PPEV = Post-PE invasive+ non-invasive ventilation). ICU mortality was recorded. Comparisons for variousoutcomes were performed by Chi-square and t tests.RESULTS. All consensus criteria were fulfilled in 77.3% of the patients during RW and in68.1% of the patients during SBT.[Compliance to weaning criteria and outcome]N = 458 PRF (%) Reintubation (%) PPEV (min) ICU mortality (%)All patients 53.5 10.0 542 ± 664 6.1RW+ 50.0 9.3 490 ± 626 5.4RW- 65.4* 12.5 718 ± 757** 8.7SBT+ 52.6 8.0 498 ± 594 6.7SBT- 55.5 14.4*** 637 ± 788**** 4.8Occurrence of PRF only was not associated with increased ICU mortality: 4.2 versus 7.8%,p = 0.11. By contrast, ICU mortality was significantly increased in patients requiring reintubation:21.7 versus 4.4%. p\0.001; * p = 0.006 RW+ versus RW-; ** p = 0.003RW+ versus RW-; *** p = 0.035 SBT+ versus SBT-; **** p = 0.030 SBT+ versusSBTCONCLUSIONS.In our ICU, compliance to all criteria of the two-step published approach ofrespiratory weaning was not optimal but reintubation rate was comparable to published data.Compliance with consensus conference guidelines was associated with lower reintubation rateand shorter PPEV but not with ICU mortality. As mortality was increased by reintubation,more sensitive and specific criteria to predict the risk of reintubation are probably needed.REFERENCE. Boles JM, et al. Eur Respir J 2007;29:1033-56.

International periodic fever, aphthous stomatitis, pharyngitis, cervical adenitis syndrome cohort: description of distinct phenotypes in 301 patients.

OBJECTIVES: The aims of this study were to describe the clinical features of periodic fever, aphthous stomatitis, pharyngitis and cervical adenitis (PFAPA) and identify distinct phenotypes in a large cohort of patients from different countries. METHODS: We established a web-based multicentre cohort through an international collaboration within the periodic fevers working party of the Pediatric Rheumatology European Society (PReS). The inclusion criterion was a diagnosis of PFAPA given by an experienced paediatric rheumatologist participating in an international working group on periodic fever syndromes. RESULTS: Of the 301 patients included from the 15 centres, 271 had pharyngitis, 236 cervical adenitis, 171 oral aphthosis and 132 with all three clinical features. A total of 228 patients presented with additional symptoms (131 gastrointestinal symptoms, 86 arthralgias and/or myalgias, 36 skin rashes, 8 neurological symptoms). Thirty-one patients had disease onset after 5 years and they reported more additional symptoms. A positive family history for recurrent fever or recurrent tonsillitis was found in 81 patients (26.9%). Genetic testing for monogenic periodic fever syndromes was performed on 111 patients, who reported fewer occurrences of oral aphthosis or additional symptoms. Twenty-four patients reported symptoms (oral aphthosis and malaise) outside the flares. The CRP was >50 mg/l in the majority (131/190) of the patients tested during the fever. CONCLUSION: We describe the largest cohort of PFAPA patients presented so far. We confirm that PFAPA may present with varied clinical manifestations and we show the limitations of the commonly used diagnostic criteria. Based on detailed analysis of this cohort, a consensus definition of PFAPA with better-defined criteria should be proposed.

Estimating biophysical variable dependences with kernels

This paper introduces a nonlinear measure of dependence between random variables in the context of remote sensing data analysis. The Hilbert-Schmidt Independence Criterion (HSIC) is a kernel method for evaluating statistical dependence. HSIC is based on computing the Hilbert-Schmidt norm of the cross-covariance operator of mapped samples in the corresponding Hilbert spaces. The HSIC empirical estimator is very easy to compute and has good theoretical and practical properties. We exploit the capabilities of HSIC to explain nonlinear dependences in two remote sensing problems: temperature estimation and chlorophyll concentration prediction from spectra. Results show that, when the relationship between random variables is nonlinear or when few data are available, the HSIC criterion outperforms other standard methods, such as the linear correlation or mutual information.

Benchmarking therapeutic drug monitoring software: A systematic evaluation of available computer tools

Introduction: Therapeutic drug monitoring (TDM) aims at optimizing treatment by individualizing dosage regimen based on measurement of blood concentrations. Maintaining concentrations within a target range requires pharmacokinetic and clinical capabilities. Bayesian calculation represents a gold standard in TDM approach but requires computing assistance. In the last decades computer programs have been developed to assist clinicians in this assignment. The aim of this benchmarking was to assess and compare computer tools designed to support TDM clinical activities.¦Method: Literature and Internet search was performed to identify software. All programs were tested on common personal computer. Each program was scored against a standardized grid covering pharmacokinetic relevance, user-friendliness, computing aspects, interfacing, and storage. A weighting factor was applied to each criterion of the grid to consider its relative importance. To assess the robustness of the software, six representative clinical vignettes were also processed through all of them.¦Results: 12 software tools were identified, tested and ranked. It represents a comprehensive review of the available software's characteristics. Numbers of drugs handled vary widely and 8 programs offer the ability to the user to add its own drug model. 10 computer programs are able to compute Bayesian dosage adaptation based on a blood concentration (a posteriori adjustment) while 9 are also able to suggest a priori dosage regimen (prior to any blood concentration measurement), based on individual patient covariates, such as age, gender, weight. Among those applying Bayesian analysis, one uses the non-parametric approach. The top 2 software emerging from this benchmark are MwPharm and TCIWorks. Other programs evaluated have also a good potential but are less sophisticated (e.g. in terms of storage or report generation) or less user-friendly.¦Conclusion: Whereas 2 integrated programs are at the top of the ranked listed, such complex tools would possibly not fit all institutions, and each software tool must be regarded with respect to individual needs of hospitals or clinicians. Interest in computing tool to support therapeutic monitoring is still growing. Although developers put efforts into it the last years, there is still room for improvement, especially in terms of institutional information system interfacing, user-friendliness, capacity of data storage and report generation.

Benchmarking therapeutic drug monitoring software: A systematic evaluation of available computer tools

Objectives: Therapeutic drug monitoring (TDM) aims at optimizing treatment by individualizing dosage regimen based on blood concentrations measurement. Maintaining concentrations within a target range requires pharmacokinetic (PK) and clinical capabilities. Bayesian calculation represents a gold standard in TDM approach but requires computing assistance. The aim of this benchmarking was to assess and compare computer tools designed to support TDM clinical activities.¦Methods: Literature and Internet were searched to identify software. Each program was scored against a standardized grid covering pharmacokinetic relevance, user-friendliness, computing aspects, interfacing, and storage. A weighting factor was applied to each criterion of the grid to consider its relative importance. To assess the robustness of the software, six representative clinical vignettes were also processed through all of them.¦Results: 12 software tools were identified, tested and ranked. It represents a comprehensive review of the available software characteristics. Numbers of drugs handled vary from 2 to more than 180, and integration of different population types is available for some programs. Nevertheless, 8 programs offer the ability to add new drug models based on population PK data. 10 computer tools incorporate Bayesian computation to predict dosage regimen (individual parameters are calculated based on population PK models). All of them are able to compute Bayesian a posteriori dosage adaptation based on a blood concentration while 9 are also able to suggest a priori dosage regimen, only based on individual patient covariates. Among those applying Bayesian analysis, MM-USC*PACK uses a non-parametric approach. The top 2 programs emerging from this benchmark are MwPharm and TCIWorks. Others programs evaluated have also a good potential but are less sophisticated or less user-friendly.¦Conclusions: Whereas 2 software packages are ranked at the top of the list, such complex tools would possibly not fit all institutions, and each program must be regarded with respect to individual needs of hospitals or clinicians. Programs should be easy and fast for routine activities, including for non-experienced users. Although interest in TDM tools is growing and efforts were put into it in the last years, there is still room for improvement, especially in terms of institutional information system interfacing, user-friendliness, capability of data storage and automated report generation.

Instrumental leadership: An extension of the Full-Range Leadership Theory

From a theoretical perspective, an extension to the Full Range leadership Theory (FRLT) seems needed. In this paper, we explain why instrumental leadership--a class of leadership includes leader behaviors focusing on task and strategic aspects that are neither values nor exchange oriented--can fulfill this extension. Instrument leadership is composed of four factors: environmental monitoring, strategy formulation and implementation, path-goal facilitation and outcome monitoring; these aspects of leadership are currently not included in any of the FRLT's nine leadership scales (as measured by the MLQ--Multifactor Leadership Questionnaire). We present results from two empirical studies using very large samples from a wide array of countries (N > 3,000) to examine the factorial, discriminant and criterion-related validity of the instrumental leadership scales. We find support for a four-factor instrumental leadership model, which explains incremental variance in leader outcomes in over and above transactional and transformational leadership.

International collaborative proficiency study of Human Papillomavirus type 16 serology.

We performed an international proficiency study of Human Papillomavirus (HPV) type 16 serology. A common methodology for serology based on virus-like particle (VLP) ELISA was used by 10 laboratories in 6 continents. The laboratories used the same VLP reference reagent, which was selected as the most stable, sensitive and specific VLP preparation out of VLPs donated from 5 different sources. A blinded proficiency panel consisting of 52 serum samples from women with PCR-verified HPV 16-infection, 11 control serum samples from virginal women and the WHO HPV 16 International Standard (IS) serum were distributed. The mean plus 3 standard deviations of the negative control serum samples was the most generally useful "cut-off" criterion for distinguishing positive and negative samples. Using sensitivity of at least 50% and a specificity of 100% as proficiency criteria, 6/10 laboratories were proficient. In conclusion, an international Standard Operating Procedure for HPV serology, an international reporting system in International Units (IU) and a common "cut-off" criterion have been evaluated in an international HPV serology proficiency study.

Accuracy of four cardiovascular risk scores in a Swiss population-based cohort

Purpose: To assess the global cardiovascular (CV) risk of an individual, several scores have been developed. However, their accuracy and comparability need to be evaluated in populations others from which they were derived. The aim of this study was to compare the predictive accuracy of 4 CV risk scores using data of a large population-based cohort. Methods: Prospective cohort study including 4980 participants (2698 women, mean age± SD: 52.7±10.8 years) in Lausanne, Switzerland followed for an average of 5.5 years (range 0.2 - 8.5). Two end points were assessed: 1) coronary heart disease (CHD), and 2) CV diseases (CVD). Four risk scores were compared: original and recalibrated Framingham coronary heart disease scores (1998 and 2001); original PROCAM score (2002) and its recalibrated version for Switzerland (IAS-AGLA); Reynolds risk score. Discrimination was assessed using Harrell's C statistics, model fitness using Akaike's information criterion (AIC) and calibration using pseudo Hosmer-Lemeshow test. The sensitivity, specificity and corresponding 95% confidence intervals were assessed for each risk score using the highest risk category ([20+ % at 10 years) as the "positive" test. Results: Recalibrated and original 1998 and original 2001 Framingham scores show better discrimination (>0.720) and model fitness (low AIC) for CHD and CVD. All 4 scores are correctly calibrated (Chi2<20). The recalibrated Framingham 1998 score has the best sensitivities, 37.8% and 40.4%, for CHD and CVD, respectively. All scores present specificities >90%. Framingham 1998, PROCAM and IAS-AGLA scores include the greatest proportion of subjects (>200) in the high risk category whereas recalibrated Framingham 2001 and Reynolds include <=44 subjects. Conclusion: In this cohort, we see variations of accuracy between risk scores, the original Framingham 2001 score demonstrating the best compromise between its accuracy and its limited selection of subjects in the highest risk category. We advocate that national guidelines, based on independently validated data, take into account calibrated CV risk scores for their respective countries.

Jevons et Walras : entre philosophie morale et économie sociale, un jalon dans la compréhension de la décision publique

This PhD dissertation deals with the question of evaluating social welfare and public policy making through the works of William Stanley Jevons (1835-1882) and Léon Walras (1834-1910), two economists who lived in the 19th century. These authors, well-known for their analyses on pure economics, were also deeply interested in the social problems of their time and proposed solutions to remedy them. In accordance with utilitarianism, Jevons was convinced that the reforms implemented by public authorities should improve social welfare (viewed as dependant on individual utilities). As for Walras, he defined a justice criterion based on a particular definition of natural law, and the reforms he proposed had the objective of restoring the rights following from this definition. We have drawn two principal conclusions from our work. First, the analyses of Jevons and Walras, who are often considered marginalise, are different not only in pure economics (as suggested by Jaffé in the seventies), but also from the point of view of welfare economics (defined as a science dealing with the evaluation of different social states). Secondly, these authors propose two different ways of justifying social reforms which have similarities with modern theory even though neither Jevons nor Walras are considered pioneers of welfare economics or public economics. Based on these two conclusions, we claim that studying these authors' theories might be of interest for the evaluation of public projects by the economists of today. Not only were their problems similar to ours, but also the original ideas present in their analyses may lead to the refinement of modern methods.Résumé en françaisCette these de doctorat porte sur les manières dont deux économistes du XIXe siècle, William Stanley Jevons (1835-1882) et Léon Walras (1834-1910), abordent la question du bien- etre collectif et la prise de décision publique. Connus principalement pour leurs travaux en économie pure, ces auteurs s'intéressent également à la question sociale et proposent des solutions pour y remédier. Jevons, conformément à son adhésion à la tradition utilitariste, estime que les interventions des autorités publiques doivent avoir pour objectif l'amélioration du bien- être collectif. Quant à Walras, il adopte un critère de justice qui relève du « droit naturel » et les propositions sociales qu'il préconise sont justifiées sur la base de leur conformité avec ce dernier. A l'issue du travail effectué dans le cadre de cette thèse, nous avons abouti à deux résultats principaux : (1) Les analyses de Jevons et Walras, deux économistes qui sont souvent considérés comme des « marginalistes » diffèrent fondamentalement non seulement dans leur analyse de l'économie pure, comme Jaffé le soulignait (1976), mais également en termes d'économie du bien-être (comprise comme une science ayant pour objectif l'évaluation des différents états sociaux). (2) Il existe deux voies originales pour justifier les réformes sociales dans les oeuvres de Jevons et Walras qui partagent des similarités importantes avec les théories modernes bien que ces auteurs ne fassent pas partie des jalons généralement retenus de l'histoire de l'économie du bien-être et de l'économie publique. Ces deux résultats nous conduisent à avancer que l'étude des approches de Jevons et de Walras peut contribuer à la formulation de nouvelles solutions à des problèmes rencontrés dans l'évaluation des différentes politiques publiques, car non seulement les préoccupations de ces auteurs était similaires aux nôtres, mais vu les éléments originaux que leurs propositions comportent, elles peuvent permettre de nuancer certaines aspects des méthodes modernes.

Evaluation of the toxicity of different metal compounds in the developing brain using aggregating cell cultures as a model.

To evaluate their toxicity in the developing brain, eight metal compounds, [bismuth sodium tartrate (BiNA-tartrate), CdCl(2), CoCl(2), HgCl(2), dimethyl mercury, NiCl(2), TlCl and triethyltin chloride (TET)] were tested in aggregating cell cultures of foetal rat telencephalon. The compounds were applied to the cultures continuously, either during an early developmental stage (between days 5 and 14) or during and advanced stage of maturation (between days 24 and 34). Changes in the activities of cell type-specific enzymes were used as a criterion for toxicity. A general cytotoxic effect was observed after treatment with either CdCl(2), HgCl(2) or TET at 10(-6)m, and with TlCl at 10(-5)m. Selective effects were found with BiNa-tartrate and dimethylmercury. CoCl(2) did not modify the parameters tested, whereas a stimulant effect was found with NiCl(2). The effects of several compounds were development dependent: HgCl(2), TET and TlCl were more toxic in immature cultures, whereas BiNa-tartrate, dimethylmercury and NiCl(2) were more effective in differentiated cultures.

Detecting and measuring deprivation in primary care: development, validity and reliability of a self-reported questionnaire - the DiPCare-Q

Background: General practitioners play a central role in taking deprivation into consideration when caring for patients in primary care. Validated questions to identify deprivation in primary-care practices are still lacking. For both clinical and research purposes, this study therefore aims to develop and validate a standardized instrument measuring both material and social deprivation at an individual level. Methods: The Deprivation in Primary Care Questionnaire (DiPCare-Q) was developed using qualitative and quantitative approaches between 2008 and 2011. A systematic review identified 199 questions related to deprivation. Using judgmental item quality, these were reduced to 38 questions. Two focus groups (primary-care physicians, and primary-care researchers), structured interviews (10 laymen), and think aloud interviews (eight cleaning staff) assured face validity. Item response theory analysis was then used to derive the DiPCare-Q index using data obtained from a random sample of 200 patients who were to complete the questionnaire a second time over the phone. For construct and criterion validity, the final 16 questions were administered to a random sample of 1,898 patients attending one of 47 different private primary-care practices in western Switzerland (validation set) along with questions on subjective social status (subjective SES ladder), education, source of income, welfare status, and subjective poverty. Results: Deprivation was defined in three distinct dimensions (table); material deprivation (eight items), social deprivation (five items) and health deprivation (three items). Item consistency was high in both the derivation (KR20 = 0.827) and the validation set (KR20 = 0.778). The DiPCare-Q index was reliable (ICC = 0.847). For construct validity, we showed the DiPCare-Q index to be correlated to patients' estimation of their position on the subjective SES ladder (rs = 0.539). This position was correlated to both material and social deprivation independently suggesting two separate mechanisms enhancing the feeling of deprivation. Conclusion: The DiPCare-Q is a rapid, reliable and validated instrument useful for measuring both material and social deprivation in primary care. Questions from the DiPCare-Q are easy to use when investigating patients' social history and could improve clinicians' ability to detect underlying social distress related to deprivation.

Autophagic cell death exists.

The term autophagic cell death (ACD) initially referred to cell death with greatly enhanced autophagy, but is increasingly used to imply a death-mediating role of autophagy, as shown by a protective effect of autophagy inhibition. In addition, many authors require that autophagic cell death must not involve apoptosis or necrosis. Adopting these new and restrictive criteria, and emphasizing their own failure to protect human osteosarcoma cells by autophagy inhibition, the authors of a recent Editor's Corner article in this journal argued for the extreme rarity or nonexistence of autophagic cell death. We here maintain that, even with the more stringent recent criteria, autophagic cell death exists in several situations, some of which were ignored by the Editor's Corner authors. We reject their additional criterion that the autophagy in ACD must be the agent of ultimate cell dismantlement. And we argue that rapidly dividing mammalian cells such as cancer cells are not the most likely situation for finding pure ACD.

Why is the cladoceran Simocephalus vetulus (Müller) not a 'bang-bang strategist '? A critique of the optimal-body-size model

Lynch's (1980a) optimal-body-size model is designed to explain some major trends in cladoceran life histories; in particular the fact that large and littoral species seem to be bang-bang strategists (they grow first and the reproduce) whereas smaller planktonic species seem to be intermediate strategists (they grow and reproduce simultaneously). Predation is assumed to be an important selective pressure for these trends. Simocephalus vetulus (Müller) does not fit this pattern; being a littoral and relatively large species but an intermediate strategist. As shown by computer simulations, this species would reduce its per capita rate of increase by adopting the strategy predicted by the optimal-body-size model. Two aspects of the model are criticized: (1) the optimization criterion is shown to be incorrect and (2) the prediction of an intermediate strategy is not justified. Structural constraints are suggested to be responsible for the intermediate strategy of S.vetulus. Biotic interactions seem to have little effect on the observed life-history patterns of this species.

Time invested in the global respiratory care of cystic fibrosis paediatrics patients.

INTRODUCTION: Respiratory therapy is a keystone of the treatment for cystic fibrosis (CF) lung disease, but it is time consuming. OBJECTIVES: We aimed to assess the total time spent on respiratory therapy, including chest physiotherapy (CPT) and physical activity (PA), as well as inhalation therapy (IT) and maintenance of materials (MM) to rationalise and optimise treatment. METHODS: A cross-sectional prospective study in a paediatric CF cohort. A questionnaire was developed to look at the time spent on respiratory care over 3 months. Enrolled in this study are all CF patients aged from 6 to 16 years (the exclusion criterion was lung transplantation). RESULTS: Of the 40 enrolled patients, 22 participated (13 boys and 9 girls), with a mean age of 11 years. The patients spent approximately 19.46 h per week (standard deviation ± 7.53, 8.00-35.25 h) on therapy: CPT (30.58%), IT (15.11%), PA (50%) and MM (4.32%), without statistical significance between sexes. CONCLUSION: In our cohort, CF patients spent an average of nearly 20 h a week in respiratory therapy, within a wide range of between 8 h to almost 36 h a week. PA consumes almost half of the time. Physicians have to take into consideration the burden of the treatment, to optimise the therapy.