Content Swapping: A New Image Synthesis for Construction Sign Detection in Autonomous Vehicles.

Construction signs alert drivers to the dangers of abnormally blocked roads. In the case of autonomous vehicles, construction signs should be detected automatically to prevent accidents. One might think that we can accomplish the goal easily using the popular deep-learning-based detectors, but it is not the case. To train the deep learning detectors to detect construction signs, we need a large amount of training images which contain construction signs. However, collecting training images including construction signs is very difficult in the real world because construction events do not occur frequently. To make matters worse, the construction signs might have dozens of different construction signs (i.e., contents). To address this problem, we propose a new method named content swapping. Our content swapping divides a construction sign into two parts: the board and the frame. Content swapping generates numerous synthetic construction signs by combining the board images (i.e., contents) taken from the in-domain images and the frames (i.e., geometric shapes) taken from the out-domain images. The generated synthetic construction signs are then added to the background road images via the cut-and-paste mechanism, increasing the number of training images. Furthermore, three fine-tuning methods regarding the region, size, and color of the construction signs are developed to make the generated training images look more realistic. To validate our approach, we applied our method to real-world images captured in South Korea. Finally, we achieve an average precision (AP50) score of 84.98%, which surpasses that of the off-the-shelf method by 9.15%. Full experimental results are available online as a supplemental video. The images used in the experiments are also released as a new dataset CSS138 for the benefit of the autonomous driving community.

Penalized variable selection for cause-specific hazard frailty models with clustered competing-risks data.

Competing risks data usually arise when an occurrence of an event precludes other types of events from being observed. Such data are often encountered in a clustered clinical study such as a multi-center clinical trial. For the clustered competing-risks data which are correlated within a cluster, competing-risks models allowing for frailty terms have been recently studied. To the best of our knowledge, however, there is no literature on variable selection methods for cause-specific hazard frailty models. In this article, we propose a variable selection procedure for fixed effects in cause-specific competing risks frailty models using a penalized h-likelihood (HL). Here, we study three penalty functions, LASSO, SCAD, and HL. Simulation studies demonstrate that the proposed procedure using the HL penalty works well, providing a higher probability of choosing the true model than LASSO and SCAD methods without losing prediction accuracy. The proposed method is illustrated by using two kinds of clustered competing-risks cancer data sets.

A Selection Operator for Summary Association Statistics Reveals Allelic Heterogeneity of Complex Traits.

In recent years, as a secondary analysis in genome-wide association studies (GWASs), conditional and joint multiple-SNP analysis (GCTA-COJO) has been successful in allowing the discovery of additional association signals within detected loci. This suggests that many loci mapped in GWASs harbor more than a single causal variant. In order to interpret the underlying mechanism regulating a complex trait of interest in each discovered locus, researchers must assess the magnitude of allelic heterogeneity within the locus. We developed a penalized selection operator for jointly analyzing multiple variants (SOJO) within each mapped locus on the basis of LASSO (least absolute shrinkage and selection operator) regression derived from summary association statistics. We found that, compared to stepwise conditional multiple-SNP analysis, SOJO provided better sensitivity and specificity in predicting the number of alleles associated with complex traits in each locus. SOJO suggested causal variants potentially missed by GCTA-COJO. Compared to using top variants from genome-wide significant loci in GWAS, using SOJO increased the proportion of variance prediction for height by 65% without additional discovery samples or additional loci in the genome. Our empirical results indicate that human height is not only a highly polygenic trait, but also has high allelic heterogeneity within its established hundreds of loci.

Frailty modelling approaches for semi-competing risks data.

In the semi-competing risks situation where only a terminal event censors a non-terminal event, observed event times can be correlated. Recently, frailty models with an arbitrary baseline hazard have been studied for the analysis of such semi-competing risks data. However, their maximum likelihood estimator can be substantially biased in the finite samples. In this paper, we propose effective modifications to reduce such bias using the hierarchical likelihood. We also investigate the relationship between marginal and hierarchical likelihood approaches. Simulation results are provided to validate performance of the proposed method. The proposed method is illustrated through analysis of semi-competing risks data from a breast cancer study.

Automatic detection of significant areas for functional data with directional error control.

In this paper, we propose a large-scale multiple testing procedure to find the significant sub-areas between two samples of curves automatically. The procedure is optimal in that it controls the directional false discovery rate at any specified level on a continuum asymptotically. By introducing a nonparametric Gaussian process regression model for the two-sided multiple test, the procedure is computationally inexpensive. It can cope with problems with multidimensional covariates and accommodate different sampling designs across the samples. We further propose the significant curve/surface, giving an insight on dynamic significant differences between two curves. Simulation studies demonstrate that the proposed procedure enjoys superior performance with strong power and good directional error control. The procedure is also illustrated with the application to two executive function studies in hemiplegia.

The Impact of Internet Gaming on Alcohol Consumption.

BACKGROUND: Despite growing evidence of the adverse effects of internet gaming, it has emerged as a popular leisure activity in South Korea and Asia. This is the first study that examines the causal effect of internet gaming on alcohol consumption. AIMS OF THE STUDY: The primary goal of this study is to evaluate the effect of internet gaming on alcohol consumption while controlling for unobserved individual attributes that are omitted in the alcohol consumption regression but are correlated with internet game usage. METHODS: We use data from a survey of 5,003 men and women who lived in Seoul and the surrounding metropolitan area of South Korea during the year 2014. We use the instrumental variable regressions and partially linear regressions. RESULTS: We first find that the age at which an individual starts internet gaming and being a member of an internet gaming club are significantly associated with the average hours spent internet gaming in adulthood. Using these two instrumental variables, we show that longer hours of internet gaming is associated with less consumption of alcohol among men, but more consumption of alcohol among women. The opposite effects of internet gaming on alcohol consumption for male and female users are robust to alternative specifications and estimation methods. DISCUSSION: We investigate potential channels through which men and women are differently affected by internet gaming on alcohol consumption. We find large disparities in types of gaming devices and playing partners between men and women and that these factors account for part of the different gaming effects by gender. Other gaming preferences contributing to the heterogeneous game effects are not examined due to lack of data, which is the limitation of this study. IMPLICATIONS FOR HEALTH POLICIES: The empirical findings suggest that female users of internet games, in particular those who are vulnerable to social isolation, can reap the most benefit toward reducing the risk of developing Alcohol Use Disorder (AUD) from health interventions that aim to monitor unhealthy use of internet games. IMPLICATIONS FOR HEALTH CARE PROVISION AND USE: Understanding the impact of internet gaming on other substance use such as alcohol will be useful for the design of effective clinical treatments and preventative health care provision. IMPLICATIONS FOR FURTHER RESEARCH: Based on the finding that men are likely to sit for longer periods of time indulging in games, further research may examine how the prolonged sedentary leisure activity of internet gaming affect obesity and other physical health problems.

Association between fever pattern and clinical manifestations of adult-onset Still's disease: unbiased analysis using hierarchical clustering.

OBJECTIVES: To perform unbiased analysis of fever patterns and to investigate their association with clinical manifestations and outcome of patients with adult-onset Still's disease (AOSD). METHODS: AOSD patients who were treated as in-patients from 2004 through 2015 were grouped according to 24-hour body temperature (BT) by hierarchical clustering using a Euclidean distance metric with complete linkage. The clinical and laboratory characteristics of the groups were then examined. RESULTS: Hierarchical clustering partitioned 70 AOSD patients into three distinct groups. Group 1 (n=14) had the highest mean BT (38.1± 0.4°C) and the widest variation in BT (2.7±0.9°C). Group 2 (n=35) had a lower mean BT (37.4±0.3°C) and a smaller variation (2.1±0.7°C). Group 3 (n=21) had the lowest mean BT (36.7±0.3°C) and the smallest variation (1.5±0.6°C). Clinical features and extent of organ involvement did not differ significantly between groups. However, Group 1 had lower platelet counts and higher lactate dehydrogenase, ferritin levels, and prothrombin time than the other groups. In addition, Group 1 exhibited higher risk of having a macrophage activation syndrome (MAS) and tended to require more intense treatment with corticosteroids and immunosuppressant to achieve clinical remission as compared to other groups. CONCLUSIONS: Hierarchical clustering identified three distinct fever patterns in patients with AOSD. Higher BT was associated with wider variations in diurnal temperature, higher risk of developing MAS, more intense treatment, and longer time to clinical remission, suggesting that fever pattern is a prognostic factor for AOSD.

A Methodological Perspective on the Longitudinal Cognitive Change after Stroke.

BACKGROUND/AIMS: Most studies of poststroke cognitive impairment (PSCI) have analyzed cognitive levels at specific time points rather than their changes over time. Furthermore, they seldom consider correlations between cognitive domains. We aimed to investigate the effects of these methodological considerations on determining significant PSCI predictors in a longitudinal stroke cohort. METHODS: In patients who underwent neuropsychological tests at least twice after stroke, we adopted a multilevel hierarchical mixed-effects model with domain-specific cognitive changes and a multivariate model for multiple outcomes to reflect their correlations. RESULTS: We enrolled 375 patients (median follow-up of 34.1 months). Known predictors of PSCI were generally associated with cognitive levels; however, most of the statistical significances disappeared when cognitive changes were set as outcomes, except age for memory, prior stroke and baseline cognition for executive/attention domain, and baseline cognition for visuospatial function. The multivariate analysis which considered multiple outcomes simultaneously further altered these associations. CONCLUSIONS: This study shows that defining outcomes as changes over time and reflecting correlations between outcomes may affect the identification of predictors of PSCI.

H-likelihood approach for joint modeling of longitudinal outcomes and time-to-event data.

In longitudinal studies, a subject may have different types of outcomes that could be correlated. For example, a response variable of interest would be measured repeatedly over time on the same subject and at the same time, an event time representing a single event or competing-risks event is also observed. In this paper, we propose a joint modeling framework that accounts for the inherent association between such multiple outcomes via frailties (unobserved random effects). Among outcomes, at least one outcome is an event time that has a type of a single event or competing-risks event. For inference we use the hierarchical likelihood (h-likelihood) that provides an unified efficient fitting procedure for the joint models. Numerical studies are provided to show the performance of the proposed method and two data examples are shown.

A Hierarchical Generalized Linear Model in Combination with Dispersion Modeling to Improve Sib-Pair Linkage Analysis.

OBJECTIVE: We explored a hierarchical generalized linear model (HGLM) in combination with dispersion modeling to improve the sib-pair linkage analysis based on the revised Haseman-Elston regression model for a quantitative trait. METHODS: A dispersion modeling technique was investigated for sib-pair linkage analysis using simulation studies and real data applications. We considered 4 heterogeneous dispersion settings according to a signal-to-noise ratio (SNR) in the various statistical models based on the Haseman-Elston regression model. RESULTS: Our numerical studies demonstrated that susceptibility loci could be detected well by modeling the dispersion parameter appropriately. In particular, the HGLM had better performance than the linear regression model and the ordinary linear mixed model when the SNR is low, i.e., when substantial noise was present in the data. CONCLUSION: The study shows that the HGLM in combination with dispersion modeling can be utilized to identify multiple markers showing linkage to familial complex traits accurately. Appropriate dispersion modeling might be more powerful to identify markers closest to the major genes which determine a quantitative trait.

Optimal likelihood-ratio multiple testing with application to Alzheimer's disease and questionable dementia.

BACKGROUND: Controlling the false discovery rate is important when testing multiple hypotheses. To enhance the detection capability of a false discovery rate control test, we applied the likelihood ratio-based multiple testing method in neuroimage data and compared the performance with the existing methods. METHODS: We analysed the performance of the likelihood ratio-based false discovery rate method using simulation data generated under independent assumption, and positron emission tomography data of Alzheimer's disease and questionable dementia. We investigated how well the method detects extensive hypometabolic regions and compared the results to those of the conventional Benjamini Hochberg-false discovery rate method. RESULTS: Our findings show that the likelihood ratio-based false discovery rate method can control the false discovery rate, giving the smallest false non-discovery rate (for a one-sided test) or the smallest expected number of false assignments (for a two-sided test). Even though we assumed independence among voxels, the likelihood ratio-based false discovery rate method detected more extensive hypometabolic regions in 22 patients with Alzheimer's disease, as compared to the 44 normal controls, than did the Benjamini Hochberg-false discovery rate method. The contingency and distribution patterns were consistent with those of previous studies. In 24 questionable dementia patients, the proposed likelihood ratio-based false discovery rate method was able to detect hypometabolism in the medial temporal region. CONCLUSIONS: This study showed that the proposed likelihood ratio-based false discovery rate method efficiently identifies extensive hypometabolic regions owing to its increased detection capability and ability to control the false discovery rate.

Korean, Japanese, and Chinese populations featured similar genes encoding drug-metabolizing enzymes and transporters: a DMET Plus microarray assessment.

BACKGROUND: Interethnic differences in genetic polymorphism in genes encoding drug-metabolizing enzymes and transporters are one of the major factors that cause ethnic differences in drug response. This study aimed to investigate genetic polymorphisms in genes involved in drug metabolism, transport, and excretion among Korean, Japanese, and Chinese populations, the three major East Asian ethnic groups. METHODS: The frequencies of 1936 variants representing 225 genes encoding drug-metabolizing enzymes and transporters were determined from 786 healthy participants (448 Korean, 208 Japanese, and 130 Chinese) using the Affymetrix Drug-Metabolizing Enzymes and Transporters Plus microarray. To compare allele or genotype frequencies in the high-dimensional data among the three East Asian ethnic groups, multiple testing, principal component analysis (PCA), and regularized multinomial logit model through least absolute shrinkage and selection operator were used. RESULTS: On microarray analysis, 1071 of 1936 variants (>50% of markers) were found to be monomorphic. In a large number of genetic variants, the fixation index and Pearson's correlation coefficient of minor allele frequencies were less than 0.034 and greater than 0.95, respectively, among the three ethnic groups. PCA identified 47 genetic variants with multiple testing, but was unable to discriminate ethnic groups by the first three components. Multinomial least absolute shrinkage and selection operator analysis identified 269 genetic variants that showed different frequencies among the three ethnic groups. However, none of those variants distinguished between the three ethnic groups during subsequent PCA. CONCLUSION: Korean, Japanese, and Chinese populations are not pharmacogenetically distant from one another, at least with regard to drug disposition, metabolism, and elimination.

Variable selection in subdistribution hazard frailty models with competing risks data.

The proportional subdistribution hazards model (i.e. Fine-Gray model) has been widely used for analyzing univariate competing risks data. Recently, this model has been extended to clustered competing risks data via frailty. To the best of our knowledge, however, there has been no literature on variable selection method for such competing risks frailty models. In this paper, we propose a simple but unified procedure via a penalized h-likelihood (HL) for variable selection of fixed effects in a general class of subdistribution hazard frailty models, in which random effects may be shared or correlated. We consider three penalty functions, least absolute shrinkage and selection operator (LASSO), smoothly clipped absolute deviation (SCAD) and HL, in our variable selection procedure. We show that the proposed method can be easily implemented using a slight modification to existing h-likelihood estimation approaches. Numerical studies demonstrate that the proposed procedure using the HL penalty performs well, providing a higher probability of choosing the true model than LASSO and SCAD methods without losing prediction accuracy. The usefulness of the new method is illustrated using two actual datasets from multi-center clinical trials.

Sparse partial least-squares regression for high-throughput survival data analysis.

The partial least-square (PLS) method has been adapted to the Cox's proportional hazards model for analyzing high-dimensional survival data. But because the latent components constructed in PLS employ all predictors regardless of their relevance, it is often difficult to interpret the results. In this paper, we propose a new formulation of sparse PLS (SPLS) procedure for survival data to allow simultaneous sparse variable selection and dimension reduction. We develop a computing algorithm for SPLS by modifying an iteratively reweighted PLS algorithm and illustrate the method with the Swedish and the Netherlands Cancer Institute breast cancer datasets. Through the numerical studies, we find that our SPLS method generally performs better than the standard PLS and sparse Cox regression methods in variable selection and prediction.

A Model for Determining Predictors of the MUAC in Acute Malnutrition in Ghana.

The issue of malnutrition is perhaps the most important public health determinant of global wellbeing. It is one of the main causes of improper mental and physical development as well as death of many children. The Mid Upper Arm Circumference (MUAC) rapid text setup is able to diagnose malnutrition due to the fact that the human arm contains subcutaneous fat and muscle mass. When proportional food intake increases or reduces, the corresponding increase or reduction in the subcutaneous fat and muscle mass leads to an increase or decrease in the MUAC. In this study, the researchers attempt to develop a model for determining the performance of MUAC in predicting Child malnutrition in Ghana. It focuses on the Joint Generalized Linear Model (Joint-GLM) instead of the traditional Generalized Linear Model (GLM). The analysis is based on primary data measured on children under six years, who were undergoing nutritional treatment at the Princess Marie Louise (PML) Children's Hospital in the Ashiedu Keteke sub-metro area of Accra Metropolis. The study found that a precisely measured weight of a child, height, and albumen levels were positive determinants of the predicted MUAC value. The study also reveals that, of all the variables used in determining the MUAC outcome, the hemoglobin and total protein levels of a child would be the main causes of any variation between the exact nutritional status of a child and that suggested by the MUAC value. The final Joint-GLM suggests that, if there are occasions where the MUAC gave false results, it could be a result of an imbalance in the child's hemoglobin and protein levels. If these two are within acceptable levels in a child, the MUAC is most likely to be consistent in predicting the child's nutritional status accurately. This study therefore recommends the continued use of MUAC in diagnosis of child malnutrition but urges Ghana and countries in Sub-Saharan Africa to roll out an effective nutrition intervention plan targeting the poor and vulnerable suburbs so that the nutritional status of children under five years of age, who were the focus of the current study, may be improved.

Semiparametric estimation for nonparametric frailty models using nonparametric maximum likelihood approach.

A consequence of using a parametric frailty model with nonparametric baseline hazard for analyzing clustered time-to-event data is that its regression coefficient estimates could be sensitive to the underlying frailty distribution. Recently, there has been a proposal for specifying both the baseline hazard and the frailty distribution nonparametrically, and estimating the unknown parameters by the maximum penalized likelihood method. Instead, in this paper, we propose the nonparametric maximum likelihood method for a general class of nonparametric frailty models, i.e. models where the frailty distribution is completely unspecified but the baseline hazard can be either parametric or nonparametric. The implementation of the estimation procedure can be based on a combination of either the Broyden-Fletcher-Goldfarb-Shanno or expectation-maximization algorithm and the constrained Newton algorithm with multiple support point inclusion. Simulation studies to investigate the performance of estimation of a regression coefficient by several different model-fitting methods were conducted. The simulation results show that our proposed regression coefficient estimator generally gives a reasonable bias reduction when the number of clusters is increased under various frailty distributions. Our proposed method is also illustrated with two data examples.

Super-sparse principal component analyses for high-throughput genomic data.

BACKGROUND: Principal component analysis (PCA) has gained popularity as a method for the analysis of high-dimensional genomic data. However, it is often difficult to interpret the results because the principal components are linear combinations of all variables, and the coefficients (loadings) are typically nonzero. These nonzero values also reflect poor estimation of the true vector loadings; for example, for gene expression data, biologically we expect only a portion of the genes to be expressed in any tissue, and an even smaller fraction to be involved in a particular process. Sparse PCA methods have recently been introduced for reducing the number of nonzero coefficients, but these existing methods are not satisfactory for high-dimensional data applications because they still give too many nonzero coefficients. RESULTS: Here we propose a new PCA method that uses two innovations to produce an extremely sparse loading vector: (i) a random-effect model on the loadings that leads to an unbounded penalty at the origin and (ii) shrinkage of the singular values obtained from the singular value decomposition of the data matrix. We develop a stable computing algorithm by modifying nonlinear iterative partial least square (NIPALS) algorithm, and illustrate the method with an analysis of the NCI cancer dataset that contains 21,225 genes. CONCLUSIONS: The new method has better performance than several existing methods, particularly in the estimation of the loading vectors.

Clustering with varying risks of false assignments in discrete latent variable model.

In clustering problems, to model the intrinsic structure of unlabeled data, the latent variable models are frequently used. These model-based clustering methods often provide a clustering rule minimizing the total false assignment error. However, in many clustering applications, it is desirable to treat false assignment errors for a certain cluster differently. In this paper, we introduce the false assignment rate for clustering and estimate it by using the extended likelihood approach. We propose VRclust, a novel clustering rule that controls various errors differently across clusters. Real data examples illustrate the usage of estimation of false assignment rate and a simulation study shows that error controls are consistent as the sample size increases.

Cluster-specific nonignorably missing, endogenous, and continuous regressors in multilevel model for binary outcome.

In multilevel regression models for observational clustered data, regressors can be correlated with cluster-level error components, namely endogenous, due to omitted cluster-level covariates, measurement error, and simultaneity. When endogeneity is ignored, regression coefficient estimators can be severely biased. To deal with endogeneity, instrument variable methods have been widely used. However, the instrument variable method often requires external instrument variables with certain conditions that cannot be verified empirically. Methods that use the within-cluster variations of the endogenous variable work under the restriction that either the outcome or the endogenous variable has a linear relationship with the cluster-level random effect. We propose a new method for binary outcome when it follows a logistic mixed-effects model and the endogenous variable is normally distributed but not linear in the random effect. The proposed estimator capitalizes on the nested data structure without requiring external instrument variables. We show that the proposed estimator is consistent and asymptotically normal. Furthermore, our method can be applied when the endogenous variable is missing in a cluster-specific nonignorable mechanism, without requiring that the missing mechanism be correctly specified. We evaluate the finite sample performance of the proposed approach via simulation and apply the method to a health care study using a San Diego inpatient dataset. Our study demonstrates that the clustered structure can be exploited to draw valid analysis of multilevel data with correlated effects.

Investigation of Correlated Internet and Smartphone Addiction in Adolescents: Copula Regression Analysis.

Internet and smartphone addiction have become important social issues. Various studies have demonstrated their association with clinical and psychological factors, including depression, anxiety, aggression, anger expression, and behavioral inhibition, and behavioral activation systems. However, these two addictions are also highly correlated with each other, so the consideration of the relationship between internet and smartphone addiction can enhance the analysis. In this study, we considered the copula regression model to regress the bivariate addictions on clinical and psychological factors. Real data analysis with 555 students (age range: 14-15 years; males, N = 295; females, N = 265) from South Korean public middle schools is illustrated. By fitting the copula regression model, we investigated the dependency between internet and smartphone addiction and determined the risk factors associated with the two addictions. Furthermore, by comparing the model fits of the copula model with linear regression and generalized linear models, the best copula model was proposed in terms of goodness of fit. Our findings revealed that internet and smartphone addiction are not separate problems, and that associations between them should be considered. Psychological factors, such as anxiety, the behavioral inhibition system, and aggression were also significantly associated with both addictions, while ADHD symptoms were related to internet addiction only. We emphasize the need to establish policies on the prevention, management, and education of addiction.