The use of classification and regression trees to predict the likelihood of seasonal influenza.

Background Individual signs and symptoms are of limited value for the diagnosis of influenza. Objective To develop a decision tree for the diagnosis of influenza based on a classification and regression tree (CART) analysis. Methods Data from two previous similar cohort studies were assembled into a single dataset. The data were randomly divided into a development set (70%) and a validation set (30%). We used CART analysis to develop three models that maximize the number of patients who do not require diagnostic testing prior to treatment decisions. The validation set was used to evaluate overfitting of the model to the training set. Results Model 1 has seven terminal nodes based on temperature, the onset of symptoms and the presence of chills, cough and myalgia. Model 2 was a simpler tree with only two splits based on temperature and the presence of chills. Model 3 was developed with temperature as a dichotomous variable (≥38°C) and had only two splits based on the presence of fever and myalgia. The area under the receiver operating characteristic curves (AUROCC) for the development and validation sets, respectively, were 0.82 and 0.80 for Model 1, 0.75 and 0.76 for Model 2 and 0.76 and 0.77 for Model 3. Model 2 classified 67% of patients in the validation group into a high- or low-risk group compared with only 38% for Model 1 and 54% for Model 3. Conclusions A simple decision tree (Model 2) classified two-thirds of patients as low or high risk and had an AUROCC of 0.76. After further validation in an independent population, this CART model could support clinical decision making regarding influenza, with low-risk patients requiring no further evaluation for influenza and high-risk patients being candidates for empiric symptomatic or drug therapy.

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.

Generalized descriptive set theory and classification theory

Descriptive set theory is mainly concerned with studying subsets of the space of all countable binary sequences. In this paper we study the generalization where countable is replaced by uncountable. We explore properties of generalized Baire and Cantor spaces, equivalence relations and their Borel reducibility. The study shows that the descriptive set theory looks very different in this generalized setting compared to the classical, countable case. We also draw the connection between the stability theoretic complexity of first-order theories and the descriptive set theoretic complexity of their isomorphism relations. Our results suggest that Borel reducibility on uncountable structures is a model theoretically natural way to compare the complexity of isomorphism relations.

The separate and combined effects of MHC genotype, parasite clone, and host gender on the course of malaria in mice.

BACKGROUND: The link between host MHC (major histocompatibility complex) genotype and malaria is largely based on correlative data with little or no experimental control of potential confounding factors. We used an experimental mouse model to test for main effects of MHC-haplotypes, MHC heterozygosity, and MHC x parasite clone interactions. We experimentally infected MHC-congenic mice (F2 segregants, homo- and heterozygotes, males and females) with one of two clones of Plasmodium chabaudi and recorded disease progression. RESULTS: We found that MHC haplotype and parasite clone each have a significant influence on the course of the disease, but there was no significant host genotype by parasite genotype interaction. We found no evidence for overdominance nor any other sort of heterozygote advantage or disadvantage. CONCLUSION: When tested under experimental conditions, variation in the MHC can significantly influence the course of malaria. However, MHC heterozygote advantage through overdominance or dominance of resistance cannot be assumed in the case of single-strain infections. Future studies might focus on the interaction between MHC heterozygosity and multiple-clone infections.

Triatominae as a model of morphological plasticity under ecological pressure

The use of biochemical and genetic characters to explore species or population relationships has been applied to taxonomic questions since the 60s. In responding to the central question of the evolutionary history of Triatominae, i.e. their monophyletic or polyphyletic origin, two important questions arise (i) to what extent is the morphologically-based classification valid for assessing phylogenetic relationships? and (ii) what are the main mechanisms underlying speciation in Triatominae? Phenetic and genetic studies so far developed suggest that speciation in Triatominae may be a rapid process mainly driven by ecological factors.

Validation of the French version of the Hierarchical Personality Inventory for Children (HiPIC): Influence of gender and age on personality traits from 8 to 12 years

The study was designed to investigate the psychometric properties of the French version and the cross-language replicability of the Hierarchical Personality Inventory for Children (HiPIC). The HiPIC is an instrument aimed at assessing the five dimensions of the Five-Factor Model for Children. Subjects were 552 children aged between 8 and 12 years, rated by one or both parents. At the domain level, reliability ranged from .83 to .93 and at the facet level, reliability ranged from .69 to .89. Differences between genders were congruent with those found in the Dutch sample. Girls scored higher on Benevolence and Conscientiousness. Age was negatively correlated with Extraversion and Imagination. For girls, we also observed a decrease of Emotional Stability. A series of exploratory factor analyses confirmed the overall five-factor structure for girls and boys. Targeted factor analyses and congruence coefficients revealed high cross-language replicability at the domain and at the facet levels. The results showed that the French version of the HiPIC is a reliable and valid instrument for assessing personality with children and has a particularly high cross-language replicability.

IEF pattern classification-derived criteria for the identification of epoetin-delta in urine.

Epoetin-delta (Dynepo Shire Pharmaceuticals, Basing stoke, UK) is a synthetic form of erythropoietin (EPO) whose resemblance with endogenous EPO makes it hard to identify using the classical identification criteria. Urine samples collected from six healthy volunteers treated with epoetin-delta injections and from a control population were immuno-purified and analyzed with the usual IEF method. On the basis of the EPO profiles integration, a linear multivariate model was computed for discriminant analysis. For each sample, a pattern classification algorithm returned a bands distribution and intensity score (bands intensity score) saying how representative this sample is of one of the two classes, positive or negative. Effort profiles were also integrated in the model. The method yielded a good sensitivity versus specificity relation and was used to determine the detection window of the molecule following multiple injections. The bands intensity score, which can be generalized to epoetin-alpha and epoetin-beta, is proposed as an alternative criterion and a supplementary evidence for the identification of EPO abuse.

Binary classification of dyslipidemia from the waist-to-hip ratio and body mass index: a comparison of linear, logistic, and CART models.

BACKGROUND: We sought to improve upon previously published statistical modeling strategies for binary classification of dyslipidemia for general population screening purposes based on the waist-to-hip circumference ratio and body mass index anthropometric measurements. METHODS: Study subjects were participants in WHO-MONICA population-based surveys conducted in two Swiss regions. Outcome variables were based on the total serum cholesterol to high density lipoprotein cholesterol ratio. The other potential predictor variables were gender, age, current cigarette smoking, and hypertension. The models investigated were: (i) linear regression; (ii) logistic classification; (iii) regression trees; (iv) classification trees (iii and iv are collectively known as "CART"). Binary classification performance of the region-specific models was externally validated by classifying the subjects from the other region. RESULTS: Waist-to-hip circumference ratio and body mass index remained modest predictors of dyslipidemia. Correct classification rates for all models were 60-80%, with marked gender differences. Gender-specific models provided only small gains in classification. The external validations provided assurance about the stability of the models. CONCLUSIONS: There were no striking differences between either the algebraic (i, ii) vs. non-algebraic (iii, iv), or the regression (i, iii) vs. classification (ii, iv) modeling approaches. Anticipated advantages of the CART vs. simple additive linear and logistic models were less than expected in this particular application with a relatively small set of predictor variables. CART models may be more useful when considering main effects and interactions between larger sets of predictor variables.

Gender differences in whole-body fat oxidation kinetics during exercise.

Discrepancies appear in studies comparing fat oxidation between men and women. Therefore, this study aimed to quantitatively describe and compare whole-body fat oxidation kinetics between genders during exercise, using a sinusoidal (SIN) model. Twelve men and 11 women matched for age, body mass index, and aerobic fitness (maximal oxygen uptake and maximal power output per kilogram of fat-free mass (FFM)) performed submaximal incremental tests (Incr) with 5-min stages and a 7.5% maximal power output increment on a cycle ergometer. Fat oxidation rates were determined using indirect calorimetry, and plotted as a function of exercise intensity. The SIN model, which includes 3 independent variables (dilatation, symmetry, translation) that account for the main quantitative characteristics of kinetics, was used to mathematically describe fat oxidation kinetics and to determine the intensity (Fatmax) eliciting the maximal fat oxidation (MFO). During Incr, women exhibited greater fat oxidation rates from 35% to 85% maximal oxygen uptake, MFO (6.6 ± 0.9 vs. 4.5 ± 0.3 mg·kg FFM-1·min-1), and Fatmax (58.1% ± 1.9% vs. 50.0% ± 2.7% maximal oxygen uptake) than men (p < 0.05). While men and women showed similar global shapes of fat oxidation kinetics in terms of dilatation and symmetry (p > 0.05), the fat oxidation curve tended to be shifted toward higher exercise intensities in women (rightward translation, p = 0.08). These results support the idea that women have a greater reliance on fat oxidation than men during submaximal exercise, but also indicate that this greater fat oxidation is shifted toward higher exercise intensities in women than in men.

Key competences in the new ventures: a model for evaluating

This research studies from an internal view based on the Competency-Based Perspective (CBP), key organizational competencies developed for small new business. CBP is chosen in an attempt to explain the differences characterizing the closed companies from the consolidated ones. The main contribution of this paper is the definition of a set of key organizational competencies for new ventures from services and low technology based sectors. Using the classification proposed by [1] and a review of the entrepreneurship literature, the main competencies were defined and classified as: managerial, input-based, transformation-based, and output-based competencies. The proposed model for evaluating new ventures organizational competence is tested by means of Structural Equation

Health problems among detainees in Switzerland: a study using the ICPC-2 classification.

BACKGROUND: Little is known about the health status of prisoners in Switzerland. The aim of this study was to provide a detailed description of the health problems presented by detainees in Switzerland's largest remand prison. METHODS: In this retrospective cross-sectional study we reviewed the health records of all detainees leaving Switzerland's largest remand prison in 2007. The health problems were coded using the International Classification for Primary Care (ICPC-2). Analyses were descriptive, stratified by gender. RESULTS: A total of 2195 health records were reviewed. Mean age was 29.5 years (SD 9.5); 95% were male; 87.8% were migrants. Mean length of stay was 80 days (SD 160). Illicit drug use (40.2%) and mental health problems (32.6%) were frequent, but most of these detainees (57.6%) had more generic primary care problems, such as skin (27.0%), infectious diseases (23.5%), musculoskeletal (19.2%), injury related (18.3%), digestive (15.0%) or respiratory problems (14.0%). Furthermore, 7.9% reported exposure to violence during arrest by the police. CONCLUSION: Morbidity is high in this young, predominantly male population of detainees, in particular in relation to substance abuse. Other health problems more commonly seen in general practice are also frequent. These findings support the further development of coordinated primary care and mental health services within detention centers.

Scene Classification Using a Hybrid Generative/Discriminative Approach

We investigate whether dimensionality reduction using a latent generative model is beneficial for the task of weakly supervised scene classification. In detail, we are given a set of labeled images of scenes (for example, coast, forest, city, river, etc.), and our objective is to classify a new image into one of these categories. Our approach consists of first discovering latent ";topics"; using probabilistic Latent Semantic Analysis (pLSA), a generative model from the statistical text literature here applied to a bag of visual words representation for each image, and subsequently, training a multiway classifier on the topic distribution vector for each image. We compare this approach to that of representing each image by a bag of visual words vector directly and training a multiway classifier on these vectors. To this end, we introduce a novel vocabulary using dense color SIFT descriptors and then investigate the classification performance under changes in the size of the visual vocabulary, the number of latent topics learned, and the type of discriminative classifier used (k-nearest neighbor or SVM). We achieve superior classification performance to recent publications that have used a bag of visual word representation, in all cases, using the authors' own data sets and testing protocols. We also investigate the gain in adding spatial information. We show applications to image retrieval with relevance feedback and to scene classification in videos

Leprosy reactions: the effect of gender and household contacts

Various host-related factors have been reported as relevant risk factors for leprosy reactions. To support a new hypothesis that an antigenic load in local tissues that is sufficient to trigger the immune response may come from an external supply of Mycobacterium leprae organisms, the prevalence of reactional leprosy was assessed against the number of household contacts. The number of contacts was ascertained at diagnosis in leprosy patients coming from an endemic area of Brazil. The prevalence of reactions (patients with reactions/total patients) was fitted by binomial regression and the risk difference (RD) was estimated with a semi-robust estimation of variance as a measure of effect. Five regression models were fitted. Model 1 included only the main exposure variable "number of household contacts"; model 2 included all four explanatory variables ("contacts", "fertile age", "number of skin lesions" and "bacillary index") that were found to be associated with the outcome upon univariate analysis; models 3-5 contained various combinations of three predictors. Male and female patients were analyzed separately. In females, household contacts were a significant predictor for leprosy reactions in model 1 [crude RD = 0.06; 95% confidence interval (CI) = 0.01; 0.12] and model 5 (RD = 0.05; CI = 0.02; 0.09), which included contacts, bacillary index and skin lesions as predictors. Other models were unsatisfactory because the joint presence of fertile age and bacillary index was a likely source of multicollinearity. No significant results were obtained for males. The likely interpretation of our findings might suggest that in female patients, leprosy reactions may be triggered by an external spreading of M. leprae by healthy carrier family members. The small number of observations is an obvious limitation of our study which requires larger confirmatory studies.

Gender-Specific Effects of Genetic Variants within Th1 and Th17 Cell-Mediated Immune Response Genes on the Risk of Developing Rheumatoid Arthritis.

The present study was conducted to explore whether single nucleotide polymorphisms (SNPs) in Th1 and Th17 cell-mediated immune response genes differentially influence the risk of rheumatoid arthritis (RA) in women and men. In phase one, 27 functional/tagging polymorphisms in C-type lectins and MCP-1/CCR2 axis were genotyped in 458 RA patients and 512 controls. Carriers of Dectin-2 rs4264222T allele had an increased risk of RA (OR = 1.47, 95%CI 1.10-1.96) whereas patients harboring the DC-SIGN rs4804803G, MCP-1 rs1024611G, MCP-1 rs13900T and MCP-1 rs4586C alleles had a decreased risk of developing the disease (OR = 0.66, 95%CI 0.49-0.88; OR = 0.66, 95%CI 0.50-0.89; OR = 0.73, 95%CI 0.55-0.97 and OR = 0.68, 95%CI 0.51-0.91). Interestingly, significant gender-specific differences were observed for Dectin-2 rs4264222 and Dectin-2 rs7134303: women carrying the Dectin-2 rs4264222T and Dectin-2 rs7134303G alleles had an increased risk of RA (OR = 1.93, 95%CI 1.34-2.79 and OR = 1.90, 95%CI 1.29-2.80). Also five other SNPs showed significant associations only with one gender: women carrying the MCP-1 rs1024611G, MCP-1 rs13900T and MCP-1 rs4586C alleles had a decreased risk of RA (OR = 0.61, 95%CI 0.43-0.87; OR = 0.67, 95%CI 0.47-0.95 and OR = 0.60, 95%CI 0.42-0.86). In men, carriers of the DC-SIGN rs2287886A allele had an increased risk of RA (OR = 1.70, 95%CI 1.03-2.78), whereas carriers of the DC-SIGN rs4804803G had a decreased risk of developing the disease (OR = 0.53, 95%CI 0.32-0.89). In phase 2, we genotyped these SNPs in 754 RA patients and 519 controls, leading to consistent gender-specific associations for Dectin-2 rs4264222, MCP-1 rs1024611, MCP-1 rs13900 and DC-SIGN rs4804803 polymorphisms in the pooled sample (OR = 1.38, 95%CI 1.08-1.77; OR = 0.74, 95%CI 0.58-0.94; OR = 0.76, 95%CI 0.59-0.97 and OR = 0.56, 95%CI 0.34-0.93). SNP-SNP interaction analysis of significant SNPs also showed a significant two-locus interaction model in women that was not seen in men. This model consisted of Dectin-2 rs4264222 and Dectin-2 rs7134303 SNPs and suggested a synergistic effect between the variants. These findings suggest that Dectin-2, MCP-1 and DC-SIGN polymorphisms may, at least in part, account for gender-associated differences in susceptibility to RA.

Relevant gender differences in epidemiological profile, exposure to first antiretroviral regimen and survival in the Spanish AIDS Research Network Cohort.

BACKGROUND The possible differences in the disease spectrum and prognosis of HIV infection in women and men is a major point of concern. Women are under-represented in randomized clinical trials and in some cohorts. Discordant results have often been obtained depending on the setting. METHODS We assessed gender differences in clinical and epidemiological features, antiretroviral treatment (ART) exposure and survival in two multicentre cohorts of HIV-positive subjects in Spain: CoRIS-MD and CoRIS. Competing risk regression models were used to assess gender effect on time to start ART and time to first ART change, and a Cox regression model to estimate gender effect on time to death. RESULTS Between January 1996 and December 2008, 1,953 women and 6,072 men naive to ART at study entry were included. The trend analysis over time showed the percentage of women in the younger (<20 years) and older (>50 years) strata increased significantly (P<0.001) from 0.5% and 1.8% in 1996 to 4.9% and 4.2% in 2008, respectively. By competing risk analysis women started ART earlier than men (adjusted subhazard ratio [ASHR] 1.21, 95% CI 1.11, 1.31) in CoRIS cohort, while in CoRIS-MD none of these differences were observed. In both cohorts women showed a shorter time to the first ART change (ASHR 1.10, 95% CI 1.01, 1.19). Pregnancy and patient's/physician's decisions as reasons for changing were more frequent in women than in men in CoRIS. In the Cox regression model, gender was not associated with differences in survival. CONCLUSIONS In two large cohorts in Spain, we observed relevant gender differences in epidemiological characteristics and antiretroviral exposure outcomes, while survival differences were not attributable to gender.