Conditional Adaptive Bayesian Spectral Analysis of Nonstationary Biomedical Time Series

• Activity Number: 577 - Semiparametric Modeling in Biometric Data
• Type: Contributed
• Date/Time: Wednesday, August 2, 2017 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Scott Bruce* and Martica H Hall and Daniel J Buysse and Robert T Krafty
• Companies: Department of Statistical Science, Temple University and University of Pittsburgh and Department of Psychiatry, University of Pittsburgh and Department of Biostatistics, University of Pittsburgh
• Keywords: Heart rate variability ; Locally stationary ; Replicated time series ; Reversible jump Markov chain Monte Carlo ; Sleep quality ; Spectrum analysis

Abstract

Many studies of biomedical time series signals aim to measure the association between frequency-domain properties of time series and clinical and behavioral covariates. However, the time-varying dynamics of these associations are largely ignored due to a lack of methods that can assess the changing relationship through time. This article introduces a method for the simultaneous and automatic analysis of the association between the time-varying power spectrum and covariates. The procedure adaptively partitions the grid of time and covariate values into an unknown number of approximately stationary blocks and estimates local spectra through penalized splines. The approach is formulated in a fully Bayesian framework, in which the number and locations of partition points are random, and fit using reversible jump MCMC techniques. Estimation and inference averaged over the distribution of partitions allows for the accurate analysis of spectra with both smooth and abrupt changes. This approach is used to analyze the association between the time-varying spectrum of heart rate variability and sleep quality in a study of older adults serving as the primary caregiver for their ill spouse.