ijsmr
|
|
Abstract: The basket trial is a recent development in the clinical trial practice. It conducts the test of the same treatment on several different related diseases in a single trial, and has the advantage of reduced cost and enhanced efficiency. A natural question is how to assess the performance of the group sequential basket trial against the classical group sequential trial? To our knowledge, a formal assessment hasn’t been seen in the literature, and is the goal of this study. Specifically, we use the receiver operating characteristic curve to assess the performance of the mentioned two trials. We considered two cases, parametric and nonparametric settings. The former is efficient when the parametric model is correctly specified, but can bemis-leading if the model is incorrect; the latter is less efficient but is robust in that it cannot be wrong no matter what the true data generating model is. Simulation studies are conducted to evaluate the experiments, and it suggests that the group sequential basket trial generally outperforms the group sequential trial in either the parametric and nonparametric cases, and that the nonparametric method gives more accurate evaluation than the parametric one for moderate to large sample sizes. Keywords: Basket trial, group sequential clinical trial, nonparametric ROC curve, parametric ROC curve, phase II clinical trial. |
|
|
Abstract: A major objective in microarray experiments is to identify a panel of genes that are associated with a disease outcome or trait. Many statistical methods have been proposed for gene selection within the last fifteen years. While the comparison of some of these methods has been done, most of them concentrated on finding gene signatures based on two groups. This study evaluates four gene selection methods when the outcome of interested is continuous in nature. We provide a comparative review of four methods: the Statistical Analysis of Microarrays (SAM), the Linear Models for Microarray Analysis (LIMMA), the Lassoed Principal Components (LPC), and the Quantitative Trait Analysis (QTA). Comparison is based on the power to identify differentially expressed genes, the predictive ability of the genelists for a continuous outcome (G2 checkpoint function), and the prognostic properties of the genelists for distant metastasis-free survival. A simulated dataset and a publicly available melanoma cell lines dataset are used for simulations and validation, respectively. A primary melanoma dataset is used for assessment of prognosis. No common genes were found among the genelists from the four methods. While the SAM was generally the best in terms of power, the QTA genelist performed the best in the prediction of the G2 checkpoint function. Identification of genelists depends on the choice of the gene selection method. The QTA method would be preferred over the other approaches in predicting a quantitative outcome in melanoma research. We recommend the development of more robust statistical methods for differential gene expression analysis. Keywords: Differential gene expression, Melanoma cell lines, Prediction, Power, Quantitative trait. |
|
|
Abstract: Objective: To demonstrate the use of robust Cox regression in estimating adjusted relative risks (and confidence intervals) when all participants with an identical follow-up time and when a common outcome is investigated. Methods: In this paper, we propose an alternative statistical method, robust Cox regression, to estimate adjusted relative risks in prospective studies. We use simulated cohort data to examine the suitability of robust Cox regression. Results: Robust Cox regression provides estimates that are equivalent to those of modified Poisson regression: regression coefficients, relative risks, 95% confidence intervals, P values. It also yields reasonable probabilities (bounded by 0 and 1). Unlike modified Poisson regression, robust Cox regression allows for four automatic variable selection methods, it directly computes adjusted relative risks for continuous variables, and is able to incorporate time-dependent covariates. Conclusion: Given the popularity of Cox regression in the medical and epidemiological literature, we believe that robust Cox regression may gain wider acceptance and application in the future. We recommend robust Cox regression as an alternative analytical tool to modified Poisson regression. In this study we demonstrated its utility to estimate adjusted relative risks for common outcomes in prospective studies with two or three waves of data collection (spaced similarly). Keywords: Robust Cox regression, Modified Poisson regression, Logistic regression, Relative risk, Odds ratio.Download Full Article |
|
|
Abstract: This article focuses on model based sparse feature extraction of biomedical signals for classification problems, which stems from sparse representation in modern signal processing. In the presented work, a novel approach based on sparse principal component analysis (SPCA) is proposed to extract signal features. This method involves partitioning signals and utilizing SPCA to select only a limited number of signal segments in order to construct signal principal components during the training stage. For signal classification purposes, a set of regression models based on sparse principal components of the selected training signal segments is constructed. Within this approach, model residuals are estimated and used as signal features for classification. The applications of the proposed approach are demonstrated by using both the synthetic data and real EEG signals. The high classification accuracy results suggest that the proposed methods may be useful for automatic event detection using long-term observational signals. keywords: Sparse Principal Component Analysis, Sparse Feature Extraction, Signal Classification, Long-term Signals. Keywords: Sparse Principal Component Analysis, Sparse Representation, Signal Classification, Long-term Signals.Download Full Article |
|
|
Abstract: In the present article our aim is to model the HIV infected adult patients’ dataset. A retrospective cohort study was conducted in Zewditu Referral Hospital located in Addis Ababa, Ethiopia. Records of patients enrolled between September 2010 and August 2014 were reviewed continuously using patients’Antiretroviral Therapy (ART) unique identification numbers as reference. Kaplan-Meier survival curves and Log-Rank test were used to compare the survival experience of different category of patients. Then we attempted to model the above data with the help of four parametric models namely; Exponential, Weibull, Gompertz, and Log-logistic. All fitted models were compared separately by using AIC and log likelihood. The log-logistic model gave a better description of the time-to-death of HIV infected adult patients than the other models. Based on log-logistic model, age, weight, and functional status, TB screen, World Health Organization (WHO) clinical stage and educational level were found to be the most prognostic factors of time-to-death. Furthermore a high risk of death of patients was found to be associated with lower initial weight, WHO clinical stage IV, lower CD4 count, being ambulatory, bedridden, and TB screened and illiterate. Keywords: Human immunodeficiency Virus, Acquired immune deficiency syndrome, Parametric Models, HAART, ARTCD.Download Full Article |