Department of Biostatistics Seminar/Workshop Series

A Repeated Measures Approach to the Statistical Analysis of Left Ventricular MRI Data

Samantha R. Seals, PhD

Biostatistician III, University of Mississippi Medical Center

In 2002, the American Heart Association recommended a 17-segment model for use in the imaging of the left ventricle. This 17-segment model produces multiple observations per subject, and in turn, these observations can be considered at the segment level, regional level (e.g., slices of the LV or by arterial geometry), or the global level. Appropriate statistical methods depend on number of observations per subject, which depends on the level at which the researcher wishes to analyze data. Many researchers opt to employ data reduction in the 17-segment model using a summary statistic, such as the average of all segments; however, we propose retaining all 17 observations to preserve the correlation between segments. We will first discuss frequentist methods appropriate for global-level, regional-level, and segmental-level data, including two-sample t-tests and generalized linear models. Next, we will discuss a family of covariance structures for clustered data, including a specification developed for use in left ventricular data. We will conclude with an application to left ventricular imaging data from a clinical trial.

Topic revision: r1 - 30 Apr 2015, AshleeBartley
 

This site is powered by FoswikiCopyright © 2013-2022 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Vanderbilt Biostatistics Wiki? Send feedback