 |
North Texas Chapter
|
|
North Texas Chapter
|
Next Meeting: Thursday, February 2
nd , 2012Refreshments: 6:30 pm Meeting: 7:00pm
Southern Methodist University Heroy Bldg. Rm 153
Speaker: Chul Ahn, PhD
Presentation: Sample size calculations for clustered and longitudinal outcomes
Abstract: Correlated outcomes commonly occur in biomedical research. Examples include repeated measurement studies, clustered outcome studies such as dental studies, ophthalmologic studies, and community intervention studies. The correlations among observations must be taken into consideration in the design due to the potential association for observations within a cluster. I present parametric and nonparametric methods to calculate sample sizes and powers for studies with correlated outcomes. Design and analysis of correlated outcomes will be illustrated using examples from studies of diagnostic sensitivity and specificity, community intervention studies and repeated measurement studies.
Biography:
Prof. Chul Ahn received the Ph.D. degree in Statistics in 1986 from Carnegie Mellon University. He has over twenty-five years of experience as a lead biostatistician on large federal funded grants and has been a biostatistics consultant for numerous research projects and clinical protocols. Dr. Ahn has over 320 peer-reviewed publications. He is now working as the Director of Biostatistics and Research Design Key Function of the NIH-funded Clinical and Translational Sciences Award (CTSA) at the university of Texas Southwestern Medical Center, a Biostatistics Core Director of the CPRIT funded multi-center clinical trial, a Director of Shared Research Resources Core of PROSPR cancer screening grant, and a Co-Director of an NIH-funded SPORE grant. Dr. Ahn has been working on CTSA BERD evaluation committee. He will also work on the important and timely CTSA BERD evaluation metrics survey as a project leader.
Announcement:
ASA chapters, sections, and outreach groups are invited to join a wide array of other statistical groups in celebrating the International Year of Statistics (aka Statistics2013). Please visit
http://statistics2013.org. Professor Linda Hynan has registered our chapter on the website.Chapter Officers
President – Yahya Daoud
Vice-President/Program officer – Jing Cao
Secretary and Treasurer - - John Blankenbaker
Chapter Representative - - Linda Hynan
Administrative Assistant - - Sheila Crain
The North Texas Chapter of the American Statistical Association maintains a web page at URL http://www.smu.edu/asant
Next Meeting: Thursday, November 10th, 2011
Refreshments at 6:30 pm and Meeting begins at 7:00pm in Rm 153 Heroy Bldg. on SMU Campus.
Speaker: Kai-Sheng Song, PhD
Presentation: Statistical Modeling and Analysis of the Echo Envelope Distributions in Medical Ultrasound Imaging
Abstract:
Ultrasonic imaging is arguably the second most popular medical imaging modality today next only to conventional X-ray. B-mode ultrasound imaging is a widely used clinical diagnostic tool for the detection and identification of abnormalities in the liver, breast, and other organs/soft tissue structures in the human body. However, the presence of ultrasound speckle makes it difficult to visually or auto-matically interpret clinical ultrasound images. In this talk, we discuss some statistical issues in speckle analysis and present various existing approaches in modelling the envelope distribution of backscattered echo received from tissues. We propose a flexible family of envelope distributions, illustrate its finite sample performance in parameter estimation and testing, and demonstrate its effectiveness in analysis of B-scan ultrasound images. The proposed method provides a potentially powerful and practical procedure to be used in ultrasound image segmentation, tissue typing, and target identification.
Biography: Kai-Sheng Song received the Ph.D. degree in Statistics in 1993 from the University of California, Davis. He was a visiting Assistant Professor in the Department of Statistics at Purdue University, West Lafayette, during 1993-1994 and at Texas A&M University, College Station, during 1994-1995, respectively. From 1995 to 2001, he was an Assistant Professor of Statistics at the Florida State University, Tallahassee, and in 2001, he was promoted to the rank of Associate Professor of Statistics with Tenure. He joined the faculty of the University of North Texas in 2007. Dr. Song has published more than 30 articles and has been awarded 2 patents. He is an Associate Editor for the Biometrical Journal. His current research interests include Signal and Image Processing, ROC Curve and Surface Analysis, Statistical Theory for Algorithms, and Nonparametric (Minimax) Hypothesis Testing. Dr. Song is the recipient of several awards including the IBM Faculty Award in 2005 and the Florida State University Teaching Award in 2000.
Chapter Officers
President – Yahya Daoud
Vice-President/Program officer – Jing Cao
Secretary and Treasurer - - John Blankenbaker
Chapter Representative - - Linda Hynan
Administrative Assistant - - Sheila Crain
Last Meeting: Thursday, September 29th, 2011
Refreshments: 6:30 pm Meeting: 7:00pm
Southern Methodist University Heroy Bldg. Rm 153
Speaker: Tony Ng, PhD
Presentation: Optimal Sample Size Allocation for Multi-Stress Tests using Extreme Value Regression
Abstract: In this talk, I will discuss the optimal sample size allocation in a multi-group life-testing experiment for complete sample and Type-II censored sample. The extreme value regression model is commonly used for statistical analysis of data arising from such a multi-stress experiment, for example, the books by Nelson (1982) and Meeker and Escobar (1998). By considering this situation, we will derive the maximum likelihood estimators (MLEs), expected Fisher information and the asymptotic variance-covariance matrix of the MLEs. Three optimality criteria will be introduced and the optimal allocation of units for two- and k-stress level situations will then be determined. Then I will demonstrate the efficiency of this optimal allocation rule by using the real experimental situation considered earlier by Nelson and Meeker (1978). Finally, I will present some Monte Carlo simulations to show that the optimality results hold for small sample sizes as well.
References:
C. Y. Ka, P. S. Chan, H. K. T. Ng and N. Balakrishnan (2011). Optimal Sample Size Allocation for Multi-level Stress Testing with Extreme Value Regression under Type-II Censoring, Statistics, 45, 257 - 279.
H. K. T. Ng, N. Balakrishnan and P. S. Chan (2007). Optimal Sample Size Allocation in Extreme Value Regression, Naval Research Logistics, 54, 237-249.
Last meeting was Thursday May 19, 2011 at 7:00 PM
Meeting Place: Meeting will be held in Room 153 Heroy on the SMU Campus
Social: Starting at 6:30 PM
Agenda:
1. Election of North Texas (NT) ASA Chapter Officers for 2011-2012
The following people have agreed to run for office – are there others who would like to run for any of these offices?
a. President Yahya Daoud
b. Vice President/ Program Chair
Candidates:
Jing Cao, Ph.D. Assistant Professor, Department of Statistical Science, Southern Methodist University
c. Secretary-Treasurer: John Blankenbaker
d. Chapter Representative: Linda Hynan
2. Questionnaire to members to find out interests and topics. (Linda)
3. Thank the NT ASA Chapter President Dr. Xinlei (Sherry) Wang for her great leadership and contributions to the Chapter. (Linda)
4. Presentation: Going from Data Curation to Analysis from the Perspective of the Clinical Researcher and the Statistician
Title: Going from Data Curation to Analysis from the Perspective of the Clinical Researcher and the Statistician
Abstract:
Two case studies will be presented highlighting the interaction between the clinical researcher and statistician. The first case study outlines the steps of a medical comparative effectiveness research project. The discussion will include defining the population cohort, accounting for treatment assignment biases, and ultimately using an inverse probability weighted Cox model for analysis.
For the second case study, temperature telemetry data was collected on macaque monkeys before and after a flu challenge. A time series analysis was conducted to determine if a change in temperature occurred. Results of the analysis will be presented as well as a continuation of the discussion on the challenges of obtaining a mutual understanding between the researcher and statistician in regards to the data and appropriate analyses.
Speakers:
Andrew Masica, MD, MSCI
Dr. Masica is the Director of Clinical Effectiveness for the Baylor Health Care System, where he conducts operational work and patient-oriented research aimed at improving health outcomes. Following undergraduate studies at Harvard University, Dr. Masica attended the Indiana University School of Medicine and completed a residency in Internal Medicine at the University of Texas-Southwestern Medical Center in Dallas. He subsequently received formal fellowship training in pharmacology and a master’s degree in clinical investigation at Vanderbilt.
Dr. Masica has developed an active and productive research program at BHCS, obtaining over $1.2 million in peer-reviewed funding from the Agency for Healthcare Research and Quality (AHRQ), as well as additional pharma-sponsored grants. He has extensive experience coordinating clinical research studies in a principal investigator role, and has translated this work into multiple publications. His current areas of investigative focus include rational use of pharmacotherapy and comparative effectiveness.
Dr. Masica’s appointment at BHCS as the Director of Clinical Effectiveness centers on generation and implementation of evidence that will promote patient safety and enhance quality of care, with an emphasis on “real-world” application. He continues to practice clinically as a hospital-based internist.
Derek Blankenship, PhD
Dr. Derek Blankenship is the Director of Biostatistics, Institute for Health Care Research and Improvement, Baylor Health Care System (BHCS) in Dallas, Texas. Prior to joining BHCS as a Biostatistician in 2007, Derek was a Senior Graduate Research Assistant and Teaching Assistant at the University of Oklahoma Health Science Center (OUHSC) in Oklahoma City, OK. In his work at BHCS, Derek provides statistical research support for a number of grants and projects, including research on the use of microarrays to understand a wide variety of diseases and conditions.
Next Chapter Meeting: March 31, 2011 at 7:00 pm in Rm 153 Heroy Hall, SMU Campus
Title: Detecting epistatic SNPs associated with complex diseases via a Bayesian classification tree search method
Speaker: Min Chen, Ph.D.
Abstract:
Complex phenotypes are known to be associated with interactions among genetic factors. A growing body of evidence suggests that gene-gene interactions contribute to many common human diseases. Identifying potential interactions of multiple polymorphisms thus may be important to understand the biology and biochemical processes of the disease etiology. However, despite the great success of genome-wide association studies that mostly focus on single locus analysis, it is challenging to detect these interactions, especially when the marginal effects of the susceptible loci are weak and/or they involve several genetic factors. Here we describe a Bayesian classification tree model to detect such interactions in case-control association studies. We show that this method has the potential to uncover interactions involving polymorphisms showing weak to moderate marginal effects as well as multi factorial interactions involving more than two loci.
Biography:
|
|
Name: Min Chen, Ph.D. Academic Title: Assistant Professor Primary Appointment: Clinical Sciences, UT Southwestern Medical Center at Dallas Email: Min.Chen@UTSouthwestern.edu |
EDUCATION
2006 Ph.D. Decision Science and Statistics, University of Texas at Austin
FELLOWSHIPS
2008 - 2010 Postdoctoral Associate, Yale University
HONORS AND AWARDS
2006 David Bruton, Jr. Fellowship, University of Texas at Austin
2006 R. L. Anderson Student Paper Award
Southern Regional Council on Statistics (SRCOS)/ASA Summer Research Conference on Statistics
1999 Best Junior Graduate Student Award, Dept. of Statistics, University of Pittsburgh
RESEARCH INTERESTS
Statistical Genomics: Genome wide association and risk prediction; Cancer genetics/epigenetics; Next generation sequencing data analysis; eQTL mapping; Methods for preprocessing genomic data.
Bayesian Methodologies and Applications: Spatial Modeling; Model Selection; Statistical Decision Theory.
PUBLICATIONS
Work has been published in PLoS Genetics, PLoS ONE, Annals of Human Genetics, Computational Statistics and Data Analysis, Communications in Statistics, Journal of American Statistical Association, Journal of Statistical Planning and Inference, Journal of Statistical Theory and Applications, etc.
PROFESSIONAL ASSOCIATIONS / AFFILIATIONS
American
Statistical Association, American Society of Human Genetics, International
Chinese Statistical Association.
TITLE: Bayesian Sample Size and Power Calculations for ROC Studies: Comparing Parametric and Nonparametric Robustness
SPEAKER: Dunlei Cheng, Biostatistician in the Institute of Health Care Research and Improvement at Baylor Health Care System in Dallas, TX
Abstract
The receiver operating characteristics (ROC) curve is a popular tool which displays diagnostic accuracy expressed in terms of a medical test’s true positive fraction (sensitivity) versus its false positive fraction (1 – specificity). An important problem in designing an ROC study is determining an appropriate sample size that will ensure adequate statistical power to measure performance of a continuous test. Methods for ROC sample size studies often assume independent normal distributions for test scores among the diseased and non-diseased populations. A Bayesian average power criterion is used here to compare sample sizes under the default two-group normal model when the data distribution for the diseased population is either skewed or multimodal. For these two common scenarios we investigate the potential for robustness of power calculations under the mis-specified normal model and compare to power when calculated under a more flexible nonparametric Dirichlet process mixture model. We also highlight the utility of flexible models for ROC data analysis and their importance to study design. When non-standard distributional shapes are anticipated, our Bayesian nonparametric approach allows investigators to power their studies based on the use of more appropriate distributional assumptions than are generally in use. Our simulation-based procedure is easily implemented using the WinBUGS and R software packages.
Biography
Dunlei Cheng works as a Biostatistician in the Institute of Health Care Research and Improvement at Baylor Health Care System in Dallas, TX where he is mainly involved in collaborative research. He got Ph.D. in Statistics from Baylor University in 2007. His research interests include sample size calculation, diagnostic tests, and Bayesian parametric and nonparametric inferences. His work has been published in Annals of Epidemiology, Biometrical Journal, Computational Statistics and Data Analysis, Statistics in Medicine, Baylor University Medical Center Proceedings, Journal for Specialists in Pediatric Nursing, Journal of Hospital Medicine, and The American Journal of Cardiology. Dunlei is the reviewer of Annals of Epidemiology and Journal of Applied Statistics. He is the member of American Statistical Association, ENAR, and International Chinese Statistical Society.Previous Meeting -- September 30, 2010
Program: Statistical Process Control and Modeling Using Homogenous Finite Mixtures Issues with the current statistical methods for Statistical Process Control (SPC) will be discussed. Homogenous finite mixtures (HFM) will be used for SPC new framework to model process data for Statistical Process Control. The framework assumes that what we observe is a finite mixture of homogenous distributions. The distributions are among a set of user defined distributions. The number and the type of distributions are selected using Bayesian Information Criteria and parameters are estimated using an EM method. Three examples will be discussed to show that the proposed method: (1) provides statistical models that better represent what we observe, compared to traditional models (based on the normality assumption or transformation to normal distributions); (2) provides better estimates for Control Limits and for probability to have non conforming units; and (3) provides more insights on the sources of variation.
Speaker: Dr. Dario Nappa
Send any changes in mailing and e-mail addresses to Sheila Crain (scrain@smu.edu) and she will have them changed in our chapter data base.
Disclaimer: The contents of this Secondary Information Source are the sole responsibility of the contributors, and do not necessarily represent the opinions or policies of Southern Methodist University.
This page is maintained by Sheila Crain.
The space to store materials in this homepage is provided by Southern Methodist University. Last updated : 11/07/11..