Online Program Home
My Program

Abstract Details

Activity Number: 84
Type: Contributed
Date/Time: Sunday, July 31, 2016 : 4:00 PM to 5:50 PM
Sponsor: Section on Bayesian Statistical Science
Abstract #321413
Title: Bayesian Inference in a PDE-Based Model of Exhaled Nitric Oxide
Author(s): Patrick Muchmore* and Sandrah P. Eckel
Companies: and University of Southern California
Keywords: Markov chain Monte Carlo ; Estimation ; Differential equations

The fractional concentration of nitric oxide in exhaled breath (FeNO) has been recognized as a clinically useful biomarker (e.g. in the diagnosis and treatment of asthma). Commercially available analyzers can sample both FeNO and flow rate at 10-100+ Hz, so the output generated during sustained exhalation is a time series with 100s-1000s of measurements. Most existing statistical models reduce these time series to a single, "steady state", estimate of FeNO. The steady state assumption demands that subjects maintain sustained exhalations, during which both FeNO, and the overall rate of exhalation, must remain constant. In practice, the steady state assumption is often violated. In this work we drop the steady state assumption, and consider the resulting dynamic (time-varying) models in a nonlinear regression framework. The dynamics of NO in the airway are assumed to satisfy an advection-diffusion-reaction partial differential equation (PDE), and the solution of this PDE determines the expected NO concentration. Deviations from the mean are assumed to have a parametric form, such as independent log-normal, and the resulting likelihood used as the basis for MCMC sampling.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

Copyright © American Statistical Association