Joint Analysis of Multiple Algorithms and Performance Measures

There has been an increasing interest in the development of new methods using Pareto optimality to deal with multi-objective criteria (for example, accuracy and time complexity). Once one has developed an approach to a problem of interest, the problem is then how to compare it with the state of art. In machine learning, algorithms are typically evaluated by comparing their performance on different data sets by means of statistical tests. Standard tests used for this purpose are able to consider jointly neither performance measures nor multiple competitors at once. The aim of this paper is to resolve these issues by developing statistical procedures that are able to account for multiple competing measures at the same time and to compare multiple algorithms altogether. In particular, we develop two tests: a frequentist procedure based on the generalized likelihood-ratio test and a Bayesian procedure based on a multinomial-Dirichlet conjugate model. We further extend them by discovering conditional independences among measures to reduce the number of parameters of such models, as usually the number of studied cases is very reduced in such comparisons. Data from a comparison among general purpose classifiers is used to show a practical application of our tests.

MAGIC observations of the February 2014 flare of 1ES 1011+496 and ensuing constraint of the EBL density

Context. In February-March 2014, the MAGIC telescopes observed the high-frequency peaked BL Lac 1ES 1011+496 (z=0.212) in flaring state at very-high energy (VHE, E>100GeV). The flux reached a level more than 10 times higher than any previously recorded flaring state of the source. Aims. Description of the characteristics of the flare presenting the light curve and the spectral parameters of the night-wise spectra and the average spectrum of the whole period. From these data we aim at detecting the imprint of the Extragalactic Background Light (EBL) in the VHE spectrum of the source, in order to constrain its intensity in the optical band. Methods. We analyzed the gamma-ray data from the MAGIC telescopes using the standard MAGIC software for the production of the light curve and the spectra. For the constraining of the EBL we implement the method developed by the H.E.S.S. collaboration in which the intrinsic energy spectrum of the source is modeled with a simple function (< 4 parameters), and the EBL-induced optical depth is calculated using a template EBL model. The likelihood of the observed spectrum is then maximized, including a normalization factor for the EBL opacity among the free parameters. Results. The collected data allowed us to describe the flux changes night by night and also to produce di_erential energy spectra for all nights of the observed period. The estimated intrinsic spectra of all the nights could be fitted by power-law functions. Evaluating the changes in the fit parameters we conclude that the spectral shape for most of the nights were compatible, regardless of the flux level, which enabled us to produce an average spectrum from which the EBL imprint could be constrained. The likelihood ratio test shows that the model with an EBL density 1:07 (-0.20,+0.24)stat+sys, relative to the one in the tested EBL template (Domínguez et al. 2011), is preferred at the 4:6 σ level to the no-EBL hypothesis, with the assumption that the intrinsic source spectrum can be modeled as a log-parabola. This would translate into a constraint of the EBL density in the wavelength range [0.24 μm,4.25 μm], with a peak value at 1.4 μm of λF_ = 12:27^(+2:75)_ (-2:29) nW m^(-2) sr^(-1), including systematics.

Association of SNPs in LCP1 and CTIF with hearing in 11 year old children:findings from the Avon Longitudinal Study of Parents and Children (ALSPAC) birth cohort and the G-EAR consortium

BACKGROUND: The genetic basis of hearing loss in humans is relatively poorly understood. In recent years, experimental approaches including laboratory studies of early onset hearing loss in inbred mouse strains, or proteomic analyses of hair cells or hair bundles, have suggested new candidate molecules involved in hearing function. However, the relevance of these genes/gene products to hearing function in humans remains unknown. We investigated whether single nucleotide polymorphisms (SNPs) in the human orthologues of genes of interest arising from the above-mentioned studies correlate with hearing function in children. METHODS: 577 SNPs from 13 genes were each analysed by linear regression against averaged high (3, 4 and 8 kHz) or low frequency (0.5, 1 and 2 kHz) audiometry data from 4970 children in the Avon Longitudinal Study of Parents and Children (ALSPAC) birth-cohort at age eleven years. Genes found to contain SNPs with low p-values were then investigated in 3417 adults in the G-EAR study of hearing. RESULTS: Genotypic data were available in ALSPAC for a total of 577 SNPs from 13 genes of interest. Two SNPs approached sample-wide significance (pre-specified at p = 0.00014): rs12959910 in CBP80/20-dependent translation initiation factor (CTIF) for averaged high frequency hearing (p = 0.00079, β = 0.61 dB per minor allele); and rs10492452 in L-plastin (LCP1) for averaged low frequency hearing (p = 0.00056, β = 0.45 dB). For low frequencies, rs9567638 in LCP1 also enhanced hearing in females (p = 0.0011, β = -1.76 dB; males p = 0.23, β = 0.61 dB, likelihood-ratio test p = 0.006). SNPs in LCP1 and CTIF were then examined against low and high frequency hearing data for adults in G-EAR. Although the ALSPAC results were not replicated, a SNP in LCP1, rs17601960, is in strong LD with rs9967638, and was associated with enhanced low frequency hearing in adult females in G-EAR (p = 0.00084). CONCLUSIONS: There was evidence to suggest that multiple SNPs in CTIF may contribute a small detrimental effect to hearing, and that a sex-specific locus in LCP1 is protective of hearing. No individual SNPs reached sample-wide significance in both ALSPAC and G-EAR. This is the first report of a possible association between LCP1 and hearing function.

Wild carrot differentiation in Europe and positive selection at DcAOX1 gene

By definition, the domestication process leads to an overall reduction of crop genetic diversity. This lead to the current search of genomic regions in wild crop relatives (CWR), an important task for modern carrot breeding. Nowadays massive sequencing possibilities can allow for discovery of novel genetic resources in wild populations, but this quest could be aided by the use of a surrogate gene (to first identify and prioritize novel wild populations for increased sequencing effort). Alternative oxidase (AOX) gene family seems to be linked to all kinds of abiotic and biotic stress reactions in various organisms and thus have the potential to be used in the identification of CWR hotspots of environment-adapted diversity. High variability of DcAOX1 was found in populations of wild carrot sampled across a West-European environmental gradient. Even though no direct relation was found with the analyzed climatic conditions or with physical distance, population differentiation exists and results mainly from the polymorphisms associated with DcAOX1 exon 1 and intron 1. The relatively high number of amino acid changes and the identification of several unusually variable positions (through a likelihood ratio test), suggests that DcAOX1 gene might be under positive selection. However, if positive selection is considered, it only acts on some specific populations (i.e. is in the form of adaptive differences in different population locations) given the observed high genetic diversity. We were able to identify two populations with higher levels of differentiation which are promising as hot spots of specific functional diversity.

Assessing the fit of unidimensional IRT models for binary data under model misspecification

Model misspecification affects the classical test statistics used to assess the fit of the Item Response Theory (IRT) models. Robust tests have been derived under model misspecification, as the Generalized Lagrange Multiplier and Hausman tests, but their use has not been largely explored in the IRT framework. In the first part of the thesis, we introduce the Generalized Lagrange Multiplier test to detect differential item response functioning in IRT models for binary data under model misspecification. By means of a simulation study and a real data analysis, we compare its performance with the classical Lagrange Multiplier test, computed using the Hessian and the cross-product matrix, and the Generalized Jackknife Score test. The power of these tests is computed empirically and asymptotically. The misspecifications considered are local dependence among items and non-normal distribution of the latent variable. The results highlight that, under mild model misspecification, all tests have good performance while, under strong model misspecification, the performance of the tests deteriorates. None of the tests considered show an overall superior performance than the others. In the second part of the thesis, we extend the Generalized Hausman test to detect non-normality of the latent variable distribution. To build the test, we consider a seminonparametric-IRT model, that assumes a more flexible latent variable distribution. By means of a simulation study and two real applications, we compare the performance of the Generalized Hausman test with the M2 limited information goodness-of-fit test and the Likelihood-Ratio test. Additionally, the information criteria are computed. The Generalized Hausman test has a better performance than the Likelihood-Ratio test in terms of Type I error rates and the M2 test in terms of power. The performance of the Generalized Hausman test and the information criteria deteriorates when the sample size is small and with a few items.

Validación diagnóstica del test neural 1 para el síndrome de túnel del carpo

El objetivo de este estudio fue realizar una prueba de validez diagnostica del test neural 1 para el diagnóstico del Síndrome de Túnel del Carpo (STC) utilizando como prueba de referencia o de oro el test de conducción nerviosa. En este estudio participaron 115 sujetos, 230 manos con sospecha clínica de STC quienes fueron evaluados con el test de conducción nerviosa y el test neural 1. Se encontró una sensibilidad del 93.0% (IC 95%:88,21-96,79) y una especificidad del 6,67% (IC 95%:0,0-33,59), razón de verosimilitud positiva fue de 1,00 y razón de verosimilitud negativa de 1,05. Valor predictivo positivo de 86,9% y un valor predictivo negativo de 12,5%. Se concluye que el test neural 1 es una prueba clínica de alta sensibilidad y baja especificidad de gran utilidad para el monitoreo e identificación del STC. Es un procedimiento para el diagnóstico clínico de bajo costo que puede incluirse en los exámenes de rutina de los trabajadores como complemento a las pruebas clínicas sugeridas por las Gatiso para dar mayor precisión a la identificación temprana del STC. Se sugiere combinarla con otros test de mayor especificidad para ser aplicada en trabajadores en condiciones de riesgo o que presenten síntomas en miembros superiores y realizar otros estudios en donde participen sujetos sin diagnóstico clínico del STC.

Active-control trials with binary data: a comparison of methods for testing superiority or non-inferiority using the odds ratio

This paper considers methods for testing for superiority or non-inferiority in active-control trials with binary data, when the relative treatment effect is expressed as an odds ratio. Three asymptotic tests for the log-odds ratio based on the unconditional binary likelihood are presented, namely the likelihood ratio, Wald and score tests. All three tests can be implemented straightforwardly in standard statistical software packages, as can the corresponding confidence intervals. Simulations indicate that the three alternatives are similar in terms of the Type I error, with values close to the nominal level. However, when the non-inferiority margin becomes large, the score test slightly exceeds the nominal level. In general, the highest power is obtained from the score test, although all three tests are similar and the observed differences in power are not of practical importance. Copyright (C) 2007 John Wiley & Sons, Ltd.

Comparison of the psychological evaluation test and classical psychoanalysis in infertile women

This study aims to compare a psychological evaluation test to classical psychoanalysis in infertile women. Two hundred women were submitted to the Psychological Evaluation Test (PET). The sum of the scores for the responses ranged from 15 to 60 points, with scores 30 points being defined as 'psycho-emotional maladjustment' (cut-off point: median + 25%). For comparison, the patients were simultaneously submitted to a psychological examination by a psychologist, who was unaware of the PET results. of the 200 patients, 66 (33%) presented a test with greater than or equal to30 points ('psycho-emotional maladjustment') and 134 (67%) a test with <30 points (normal). Upon psychological examination, 105 (52.5%) presented an abnormal evaluation and 95 (47.5%) a normal evaluation. For the PET, statistical analysis showed 82% efficiency, 62% sensitivity, 98% positive predictive value, 99% specificity, 70% negative predictive value, likelihood ratio for a positive test result 62, and likelihood ratio for negative test result 0.38. The PET proved to be a useful clinical instrument, being of help in the selection of patients with psychological needs induced by infertility.

Evaluation of the double agar gel immunodiffusion test and of the enzyme-linked immunosorbent assay in the diagnosis and follow-up of patients with chronic pulmonary aspergillosis

Coordenação de Aperfeiçoamento do Pessoal de Nível Superior (CAPES)

Accuracy of B-type natriuretic peptide tests to exclude congestive heart failure: systematic review of test accuracy studies

BACKGROUND: Congestive heart failure (CHF) is a major public health problem. The use of B-type natriuretic peptide (BNP) tests shows promising diagnostic accuracy. Herein, we summarize the evidence on the accuracy of BNP tests in the diagnosis of CHF and compare the performance of rapid enzyme-linked immunosorbent assay (ELISA) and standard radioimmunosorbent assay (RIA) tests. METHODS: We searched electronic databases and the reference lists of included studies, and we contacted experts. Data were extracted on the study population, the type of test used, and methods. Receiver operating characteristic (ROC) plots and summary ROC curves were produced and negative likelihood ratios pooled. Random-effect meta-analysis and metaregression were used to combine data and explore sources of between-study heterogeneity. RESULTS: Nineteen studies describing 22 patient populations (9 ELISA and 13 RIA) and 9093 patients were included. The diagnosis of CHF was verified by echocardiography, radionuclide scan, or echocardiography combined with clinical criteria. The pooled negative likelihood ratio overall from random-effect meta-analysis was 0.18 (95% confidence interval [CI], 0.13-0.23). It was lower for the ELISA test (0.12; 95% CI, 0.09-0.16) than for the RIA test (0.23; 95% CI, 0.16-0.32). For a pretest probability of 20%, which is typical for patients with suspected CHF in primary care, a negative result of the ELISA test would produce a posttest probability of 2.9%; a negative RIA test, a posttest probability of 5.4%. CONCLUSIONS: The use of BNP tests to rule out CHF in primary care settings could reduce demand for echocardiography. The advantages of rapid ELISA tests need to be balanced against their higher cost.

Is perioperative point-of-care prothrombin time testing accurate compared to the standard laboratory test?

There is no accepted way of measuring prothrombin time without time loss for patients undergoing major surgery who are at risk of intraoperative dilution and consumption coagulopathy due to bleeding and volume replacement with crystalloids or colloids. Decisions to transfuse fresh frozen plasma and procoagulatory drugs have to rely on clinical judgment in these situations. Point-of-care devices are considerably faster than the standard laboratory methods. In this study we assessed the accuracy of a Point-of-care (PoC) device measuring prothrombin time compared to the standard laboratory method. Patients undergoing major surgery and intensive care unit patients were included. PoC prothrombin time was measured by CoaguChek XS Plus (Roche Diagnostics, Switzerland). PoC and reference tests were performed independently and interpreted under blinded conditions. Using a cut-off prothrombin time of 50%, we calculated diagnostic accuracy measures, plotted a receiver operating characteristic (ROC) curve and tested for equivalence between the two methods. PoC sensitivity and specificity were 95% (95% CI 77%, 100%) and 95% (95% CI 91%, 98%) respectively. The negative likelihood ratio was 0.05 (95% CI 0.01, 0.32). The positive likelihood ratio was 19.57 (95% CI 10.62, 36.06). The area under the ROC curve was 0.988. Equivalence between the two methods was confirmed. CoaguChek XS Plus is a rapid and highly accurate test compared with the reference test. These findings suggest that PoC testing will be useful for monitoring intraoperative prothrombin time when coagulopathy is suspected. It could lead to a more rational use of expensive and limited blood bank resources.

Clinicopathological And Immunohistochemical Evaluation Of Oral And Oropharyngeal Squamous Cell Carcinoma In Chilean Population.

In oral and oropharyngeal squamous cell carcinoma (OCSCC and OPSCC) exist an association between clinical and histopathological parameters with cell proliferation, basal lamina, connective tissue degradation and surrounding stroma markers. We evaluated these associations in Chilean patients. A convenience sample of 37 cases of OCSCC (n=16) and OPSCC (n=21) was analyzed clinically (TNM, clinical stage) and histologically (WHO grade of differentiation, pattern of tumor invasion). We assessed the expression of p53, Ki67, HOXA1, HOXB7, type IV collagen (ColIV) and carcinoma-associated fibroblast (α-SMA-positive cells). Additionally we conducted a univariate/bivariate analysis to assess the relationship of these variables with survival rates. Males were mostly affected (56.2% OCSCC, 76.2% OPSCC). Patients were mainly diagnosed at III/IV clinical stages (68.8% OCSCC, 90.5% OPSCC) with a predominantly infiltrative pattern invasion (62.9% OCSCC, 57.1% OPSCC). Significant association between regional lymph nodes (N) and clinical stage with OCSCC-HOXB7 expression (Chi-Square test P < 0.05) was observed. In OPSCC a statistically significant association exists between p53, Ki67 with gender (Chi-Square test P < 0.05). In OCSCC and OPSCC was statistically significant association between ki67 with HOXA1, HOXB7, and between these last two antigens (Pearson's Correlation test P < 0.05). Furthermore OPSCC-p53 showed significant correlation when it was compared with α-SMA (Kendall's Tau-c test P < 0.05). Only OCSCC-pattern invasion and OPSCC-primary tumor (T) pattern resulted associated with survival at the end of the follow up period (Chi-Square Likelihood Ratio, P < 0.05). Clinical, histological and immunohistochemical features are similar to seen in other countries. Cancer proliferation markers were associated strongly from each other. Our sample highlights prognostic value of T and pattern of invasion, but the conclusions may be limited and should be considered with caution (small sample). Many cases were diagnosed in the advanced stages of the disease, which suggests that the diagnosis of OCSCC and OPSCC is made late.

Prevalência do HPV em mulheres rastreadas para o câncer cervical

OBJETIVO: Analisar a prevalência da infecção genital por papilomavírus humano (HPV) de alto risco por faixa etária e fatores associados. MÉTODOS: Estudo transversal com amostra de 2.300 mulheres (15-65 anos) que buscaram rastreamento para o câncer cervical entre fevereiro de 2002 e março de 2003 em São Paulo e Campinas, estado de São Paulo. Aplicou-se questionário epidemiológico e realizou-se coleta cervical para citologia oncológica e teste de captura híbrida II. As análises estatísticas empregadas foram teste de qui-quadrado de Pearson e análise multivariada pelo método forward likelihood ratio. RESULTADOS: A prevalência total da infecção genital por HPV de alto risco foi de 17,8%, distribuída nas faixas etárias: 27,1% (<25 anos), 21,3% (25-34 anos), 12,1% (35-44 anos), 12,0% (45-54 anos) e de 13,9% (55-65 anos). Participantes com maior número de parceiros sexuais durante a vida apresentaram maior freqüência da infecção. Relacionamento estável, idade de 35 a 44 anos e ex-fumantes foram associados à proteção da infecção. A infecção genital por HPV de alto risco ocorreu em 14,3% das citologias normais, em 77,8% das lesões escamosas de alto grau e nos dois (100%) casos de carcinoma. CONCLUSÕES: A prevalência da infecção genital por HPV de alto risco na amostra estudada foi alta. Houve predomínio de casos abaixo dos 25 anos e tendência a um novo aumento após os 55 anos, com maior freqüência naqueles com maior número de parceiros sexuais durante a vida

A comparison of lipid variables as predictors of cardiovascular disease in the Asia Pacific Region

PURPOSE: Many guidelines advocate measurement of total or low density lipoprotein cholesterol (LDL), high density lipoprotein cholesterol (HDL), and triglycerides (TG) to determine treatment recommendations for preventing coronary heart disease (CHD) and cardiovascular disease (CVD). This analysis is a comparison of lipid variables as predictors of cardiovascular disease. METHODS: Hazard ratios for coronary and cardiovascular deaths by fourths of total cholesterol (TC), LDL, HDL, TG, non-HDL, TC/HDL, and TG/HDL values, and for a one standard deviation change in these variables, were derived in an individual participant data meta-analysis of 32 cohort studies conducted in the Asia-Pacific region. The predictive value of each lipid variable was assessed using the likelihood ratio statistic. RESULTS: Adjusting for confounders and regression dilution, each lipid variable had a positive (negative for HDL) log-linear association with fatal CHD and CVD. Individuals in the highest fourth of each lipid variable had approximately twice the risk of CHD compared with those with lowest levels. TG and HDL were each better predictors of CHD and CVD risk compared with TC alone, with test statistics similar to TC/HDL and TG/HDL ratios. Calculated LDL was a relatively poor predictor. CONCLUSIONS: While LDL reduction remains the main target of intervention for lipid-lowering, these data support the potential use of TG or lipid ratios for CHD risk prediction. (c) 2005 Elsevier Inc. All rights reserved.

Predictive value for malignancy in the finding of microcalcifications on ultrasonography of thyroid nodules

Background. The am of this study was to determine the predictive value for malignancy of microcalcifications determined by ultrasonography in thyroid nodules. Methods. One hundred seventy-seven nodules were prospectively studied by ultrasonography and compared with their fine-needle aspirative biopsy. The association between the presence and type of calcification and cytologic findings was verified through the chi-square test or likelihood ratio. Results. Thirty nodules showed calcification, of which 17 had fine calcifications, 3 had fine and gross calcifications, and 10 had only coarse calcification. Seven (41.18%) of 17 fine calcified nodules were malignant on cytology, 8 (47.06%) were benign, 1 (5,88%) was indeterminate, and 1 was suspect for malignancy. We found statistical significance between the presence of fine calcifications and malignancy (p =.001) and, in the 13 malignant nodule group, 8 (61.50%) had fine calcifications. Conclusion. This study suggests that microcalcifications were highly specific for malignancy and were present in 61% of the malignant nodules. (c) 2008 Wiley Periodicals, Inc.