Jeongyoun Ahn

Blood transfusion is an independent predictor of increased mortality in nonoperatively managed blunt hepatic and splenic injuries.

BACKGROUND Management strategies for blunt solid viscus injuries often include blood transfusion. However, transfusion has previously been identified as an independent predictor of mortality in unselected trauma admissions. We hypothesized that transfusion would adversely affect mortality and outcome in patients presenting with blunt hepatic and splenic injuries after controlling for injury severity and degree of shock. METHODS We retrospectively reviewed records from all adults with blunt hepatic and/or splenic injuries admitted to a Level I trauma center over a 4-year period. Demographics, physiologic variables, injury severity, and amount of blood transfused were analyzed. Univariate and multivariate analysis with logistic and linear regression were used to identify predictors of mortality and outcome. RESULTS One hundred forty-three (45%) of 316 patients presenting with blunt hepatic and/or splenic injuries received blood transfusion within the first 24 hours. Two hundred thirty patients (72.8%) were selected for nonoperative management, of whom 75 (33%) required transfusion in the first 24 hours. Transfusion was an independent predictor of mortality in all patients (odds ratio [OR], 4.75; 95% confidence interval [CI], 1.37-16.4; p = 0.014) and in those managed nonoperatively (OR, 8.45; 95% CI, 1.95-36.53; p = 0.0043) after controlling for indices of shock and injury severity. The risk of death increased with each unit of packed red blood cells transfused (OR per unit, 1.16; 95% CI, 1.10-1.24; p < 0.0001). Blood transfusion was also an independent predictor of increased hospital length of stay (coefficient, 5.45; 95% CI, 1.64-9.25; p = 0.005). CONCLUSION Blood transfusion is a strong independent predictor of mortality and hospital length of stay in patients with blunt liver and spleen injuries after controlling for indices of shock and injury severity. Transfusion-associated mortality risk was highest in the subset of patients managed nonoperatively. Prospective examination of transfusion practices in treatment algorithms of blunt hepatic and splenic injuries is warranted.

A resampling approach for interval-valued data regression

We consider interval-valued data that frequently appear with advanced technologies in current data collection processes. Interval-valued data refer to the data that are observed as ranges instead of single values. In the last decade, several approaches to the regression analysis of interval-valued data have been introduced, but little work has been done on relevant statistical inferences concerning the regression model. In this paper, we propose a new approach to fit a linear regression model to interval-valued data using a resampling idea. A key advantage is that it enables one to make inferences on the model such as the overall model significance test and individual coefficient test. We demonstrate the proposed approach using simulated and real data examples, and also compare its performance with those of existing methods.

HDLSS Discrimination With Adaptive Data Piling

We propose new discrimination methods for classification of high dimension, low sample size (HDLSS) data that regularize the degree of data piling. The within-class scatter of the HDLSS data, when projected onto a low-dimensional discriminant subspace, can be selected to be arbitrarily small. Using this fact, we develop two different ways of tuning the amount of within-class scatter, or equivalently, the degree of data piling. In the first approach, we consider a linear path connecting the maximal data piling and the least data piling directions. We also formulate a problem of finding the optimal classifier under a constraint on data piling. The data piling regularization methods are extended to multicategory problems. Simulated and real data examples show competitive performances of the proposed classification methods. Supplementary materials for this article are available online on the journal web site.

General sparse multi-class linear discriminant analysis

Discrimination with high dimensional data is often more effectively done with sparse methods that use a fraction of predictors rather than using all the available ones. In recent years, some effective sparse discrimination methods based on Fishers linear discriminant analysis (LDA) have been proposed for binary class problems. Extensions to multi-class problems are suggested in those works; however, they have some drawbacks such as the heavy computational cost for a large number of classes. We propose an approach to generalize a binary LDA solution into a multi-class solution while avoiding the limitations of the existing methods. Simulation studies with various settings, as well as real data examples including next generation sequencing data, confirm the effectiveness of the proposed approach.

Covariance Adjustment for Batch Effect in Gene Expression Data

Batch bias has been found in many microarray gene expression studies that involve multiple batches of samples. A serious batch effect can alter not only the distribution of individual genes but also the inter-gene relationships. Even though some efforts have been made to remove such bias, there has been relatively less development on a multivariate approach, mainly because of the analytical difficulty due to the high-dimensional nature of gene expression data. We propose a multivariate batch adjustment method that effectively eliminates inter-gene batch effects. The proposed method utilizes high-dimensional sparse covariance estimation based on a factor model and a hard thresholding. Another important aspect of the proposed method is that if it is known that one of the batches is produced in a superior condition, the other batches can be adjusted so that they resemble the target batch. We study high-dimensional asymptotic properties of the proposed estimator and compare the performance of the proposed method with some popular existing methods with simulated data and gene expression data sets.

A Nonparametric Kernel Approach to Interval-Valued Data Analysis

This article concerns datasets in which variables are in the form of intervals, which are obtained by aggregating information about variables from a larger dataset. We propose to view the observed set of hyper-rectangles as an empirical histogram, and to use a Gaussian kernel type estimator to approximate its underlying distribution in a nonparametric way. We apply this idea to both univariate density estimation and regression problems. Unlike many existing methods used in regression analysis, the proposed method can estimate the conditional distribution of the response variable for any given set of predictors even when some of them are not interval-valued. Empirical studies show that the proposed approach has a great flexibility in various scenarios with complex relationships between the location and width of intervals of the response and predictor variables.

Analysis of Long Period Variable Stars With Nonparametric Tests for Trend Detection

In astronomy the study of variable stars—that is, stars characterized by showing significant variation in their brightness over time—has made crucial contributions to our understanding of many phenomena, from stellar birth and evolution to the calibration of the extragalactic distance scale. In this article, we develop a method for analyzing multiple, (pseudo)-periodic time series with the goal of detecting temporal trends in their periods. We allow for nonstationary noise and for clustering among the various time series. We apply this method to the long-standing astronomical problem of identifying variable stars whose regular brightness fluctuations have periods that change over time. The results of our analysis show that such changes can be substantial, raising the possibility that astronomers’ estimates of galactic distances can be refined. Two significant contributions of our approach, relative to existing methods for this problem, are as follows: 1. The method is nonparametric, making minimal assumptions about both the temporal trends themselves but also the covariance structure of the nonstationary noise. 2. Our proposed test has higher power than existing methods. The test is based on inference for a high-dimensional normal mean, with control of the false discovery rate to account for multiplicity. We present theory and simulations to demonstrate the performance of our method. We also analyze data from the American Association of Variable Star Observers and find a monotone relationship between mean period and strength of trend similar to that identified by Hart, Koen, and Lombard (2007).

Sparse HDLSS discrimination with constrained data piling

Regularization is a key component in high dimensional data analyses. In high dimensional discrimination with binary classes, the phenomenon of data piling occurs when the projection of data onto a discriminant vector is dichotomous, one for each class. Regularizing the degree of data piling yields a new class of discrimination rules for high dimension-low sample size data. A discrimination method that regularizes the degree of data piling while achieving sparsity is proposed and solved via a linear programming. Computational efficiency is further improved by a sign-preserving regularization that forces the signs of the estimator to be the same as the mean difference. The proposed classifier shows competitive performances for simulated and real data examples including speech recognition and gene expressions.

Sparse generalized eigenvalue problem with application to canonical correlation analysis for integrative analysis of methylation and gene expression data: SELP for CCA

We present a method for individual and integrative analysis of high dimension, low sample size data that capitalizes on the recurring theme in multivariate analysis of projecting higher dimensional data onto a few meaningful directions that are solutions to a generalized eigenvalue problem. We propose a general framework, called SELP (Sparse Estimation with Linear Programming), with which one can obtain a sparse estimate for a solution vector of a generalized eigenvalue problem. We demonstrate the utility of SELP on canonical correlation analysis for an integrative analysis of methylation and gene expression profiles from a breast cancer study, and we identify some genes known to be associated with breast carcinogenesis, which indicates that the proposed method is capable of generating biologically meaningful insights. Simulation studies suggest that the proposed method performs competitive in comparison with some existing methods in identifying true signals in various underlying covariance structures.

Distance-based outlier detection for high dimension, low sample size data

ABSTRACT Despite the popularity of high dimension, low sample size data analysis, there has not been enough attention to the sample integrity issue, in particular, a possibility of outliers in the data. A new outlier detection procedure for data with much larger dimensionality than the sample size is presented. The proposed method is motivated by asymptotic properties of high-dimensional distance measures. Empirical studies suggest that high-dimensional outlier detection is more likely to suffer from a swamping effect rather than a masking effect, thus yields more false positives than false negatives. We compare the proposed approaches with existing methods using simulated data from various population settings. A real data example is presented with a consideration on the implication of found outliers.