|
Activity Number:
|
19
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Sunday, August 6, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #306913 |
|
Title:
|
Bayesian Predictive Probability as a Diagnostic Assessment of the Likelihood of Coronary Artery Disease in Collateral Arteries
|
|
Author(s):
|
Laura Thompson*+
|
|
Companies:
|
U.S. Food and Drug Administration
|
|
Address:
|
FDA/CDRH/OSB/DBS, HFZ-550 , Rockville, MD, 20850,
|
|
Keywords:
|
Bayesian ; gender differences ; angiography ; nonparametric ; coronary artery disease
|
|
Abstract:
|
We assume a binary tree model for the human coronary artery tree and apply a reinforced random walk model to model movement of lipid particles through the artery tree. Data might consist of a set of angiographic images---with one or more stenoses---morphometric measurements of the vessels, and biological characteristics of the patients from whom the images were obtained. We apply nonparametric Bayesian methods to obtain the predictive probability of plaque, given the presence of plaque in a collateral region of the vessel that has already stenosed. This predictive probability can be used as a diagnostic tool to identify risk of recurrent coronary artery disease (CAD). We also explore the special case of female CAD.
|
