BMI Department Seminars

Upcoming Seminars:

Upcoming BMI Seminar Events
Title Date Location Presenter Abstract
Biostat Seminar - Moo K. Chung Dec 4 2015 - 12:00pm to 1:00pm Biotechnology Center Auditorium

Moo K. Chung
Biostat Seminar - Nicholas Tatonetti Dec 11 2015 - 12:00pm to 1:00pm Biotechnology Center Auditorium

Nicholas Tatonetti
Columbia University

BMI Trainee Seminar Dec 18 2015 - 12:00pm to 1:00pm Biotechnology Center Auditorium

BMI Trainee Seminar

BMI Seminar List:

Title Presenter(s) Date Location Abstract
BMI Trainee Seminar

BMI Trainee Seminar

Friday, April 29, 2016 - 12:00pm to 1:00pm Biotechnology Center Auditorium
Biostat Seminar - Hernando Ombao

Hernando Ombao

Friday, April 22, 2016 - 12:00pm to 1:00pm Biotechnology Center Auditorium
Biostat Seminar - Josh Stuart

Josh Stuart
UC Santa Cruz

Friday, April 15, 2016 - 12:00pm to 1:00pm Biotechnology Center Auditorium
Biostat Seminar - Mike Daniels

Mike Daniels
UT Austin

Friday, April 8, 2016 - 12:00pm to 1:00pm Biotechnology Center Auditorium
BMI Trainee Seminar

BMI Trainee Seminar

Friday, December 18, 2015 - 12:00pm to 1:00pm Biotechnology Center Auditorium
Biostat Seminar - Nicholas Tatonetti

Nicholas Tatonetti
Columbia University

Friday, December 11, 2015 - 12:00pm to 1:00pm Biotechnology Center Auditorium
Biostat Seminar - Moo K. Chung

Moo K. Chung
UW Madison

Friday, December 4, 2015 - 12:00pm to 1:00pm Biotechnology Center Auditorium
Steve Qin - Improving Hierarchal models using historical data

Steve Qin

Associate Professor
Emory University, Atlanta, GA

Modern high throughput biotechnologies such as microarray and next generation sequencing produce massive amount of information for each sample assayed. However, in a typical high throughput experiment, only very limited amount of data are observed for each individual feature, thus the classical large p, small n problem. Bayesian hierarchical model, capable of borrowing strength across features within the same dataset, has been recognized as an effective tool in analyzing such data. However, the shrinkage effect, the most prominent feature of hierarchical features, can lead to undesirable over-correction for some features. In this work, we discuss possible causes of the over-correction problem and propose several alternative solutions. Our strategy is rooted in the facts that in Big Data era, large amount of historical data are available which should be taken advantage of. Our strategies present a new framework to enhance the Bayesian hierarchical models. Through simulation and real data analysis, we demonstrated superior performance of the proposed strategies.

Friday, November 20, 2015 - 12:00pm to 1:00pm Biotechnology Center Auditorium PDF icon Qin_Seminar.pdf
Affiliated Seminar - Wilcox

Lauren Wilcox
School of Interactive Computing, Georgia Institute of Technology
"Supporting New Avenues of Technology between Patients and their Caregivers"

Seminar series: Industrial and Systems Engineering Colloquium

ABSTRACT: The recent trend toward patient participation in their own healthcare has opened up numerous challenges and opportunities for computing research. Dr. Wilcox’s research focuses on how technology can be designed to foster this participation---in particular, how user interfaces can be designed and developed to facilitate health-related information awareness and understanding. In this talk, she will provide an overview of her work on effective design and use of technology to inform lay people about aspects of their health status and health care.

Dr. Wilcox will describe insights into how computing systems can better support communication with patients in inpatient settings, and discuss emerging work focused on a critical yet under-supported group: teens with complex chronic illnesses. Finally, she will discuss opportunities in the emerging area of consumer-centered health informatics, and describe a future in which transformative interventions will better support communication of multi-faceted health-related information to a variety of end users.

BIO: Lauren Wilcox is an assistant professor in the School of Interactive Computing at Georgia Institute of Technology. Her research focuses on designing, building, and evaluating technology to support the needs of people working both individually and together to achieve health-related goals. In her talk, she will discuss results of a study focused on the impact of a patient-centered information portal that she developed with collaborators in the Department of Biomedical Informatics at Columbia University and NewYork-Presbyterian Hospital, to provide hospital patients with access to timely health-related data and electronic note capture tools. She will also report on formative studies focused on the design of technology to support teens with complex chronic illness in communicating health-related information with their parents and their clinical caregivers at Children’s Healthcare of Atlanta (CHOA).

Lauren received her PhD in Computer Science from Columbia University in 2013. Her studies related to communicating patient-centered health information have been recognized by the Agency for Healthcare Research and Quality (AHRQ) through a Dissertation Award in 2012 and by the NSF through a CISE Research Initiation Initiative award in 2015. Prior to her academic career, Lauren worked as a Software Engineer at IBM, where she was recognized with an Early Tenure Inventor award. She is a member of the technical program committees for PervasiveHealth 2016 and ACM CHI 2016.

Friday, November 6, 2015 - 12:00pm to 1:00pm Mechanical Engineering, 1513 Univ Ave., Room 1163 PDF icon Wilcox Colloquium Flyer.pdf
Extending Propensity Score for General Treatment Regimes

Qi Long
Emory University

The propensity score plays an essential role in causal inference when observational data are used. However, a number of challenges arise when using the propensity score to deal with general treatment regimes that are not categorical. In this talk I will present two extensions of the propensity score for general treatment regimes. First, I will discuss our work on improving the covariate balancing propensity score (CBPS; Imai and Ratkovic, 2014 and Fong et al. 2015) for continuous treatment regimes. Our proposed approach is numerically more stable and can handle both continuous and discrete covariates. Our simulation results show that the proposed approach outperforms the original CBPS. Second, I will introduce the concept of the propensity process when time to treatment is of interest in the presence of time-dependent covariates and show that the propensity process balances the entire time-dependent covariate history, which cannot be achieved by the existing propensity score methods. I will illustrate both methods using analysis of the clinical data from the Emory Amyotrophic Lateral Sclerosis (ALS) Center. These are joint work with Samantha M. Noreen, Pallavi Mishra-Kalyani, and Brent A. Johnson.

Friday, October 23, 2015 - 12:00pm to 1:00pm Biotechnology Center Auditorium