Biomedical subjects

Eric D Kolaczyk

Publications and source records attributed to Eric D Kolaczyk.

3 recordsLinked to original sources

PubMed2006

A multiscale method for disease mapping in spatial epidemiology.

The effects of spatial scale in disease mapping are well-recognized, in that the information conveyed by such maps varies with scale. Here we provide an inferential framework, in the context of tract count data, for describing the distribution of relative risk simultaneously across a hierarchy of multiple scales. In particular, we offer a multiscale extension of the canonical standardized mortality ratio (SMR), consisting of Bayesian posterior-based strategies for both estimation and characterization of uncertainty. As a result, a hierarchy of informative disease and confidence maps can be produced, without the need to first try to identify a single appropriate scale of analysis. We explore the behaviour of the proposed methodology in a small simulation study, and we illustrate its usage through an application to data on gastric cancer in Tuscany.

Bayes Theorem↗

PubMed2006

Multiscale detection of localized anomalous structure in aggregate disease incidence data.

We present a modelling framework for detection of potentially anomalous structure in aggregate spatial disease incidence data in a manner sensitive to localization at multiple scales and/or positions. The key technical contribution is the re-casting of the components of a multiscale disease mapping methodology, recently introduced by the authors in an earlier paper, into a form appropriate for hypothesis testing. In particular, we describe how hypotheses of spatially clustered variations in disease incidence may be linked in one-to-one correspondence with collections of hypotheses on the values of certain multiscale parameters associated with a user-defined hierarchy of nested partitions of an overall spatial region. A Bayesian hypothesis testing methodology is developed in the context of a standard Poisson measurement model, over the collection of possible multiscale hypotheses. We discuss the specification of hyper parameters and prior distributions on the space of models. The methodology is illustrated on both simulated and real data.

Cluster Analysis↗

PubMed2002

Wavelet-based multifractal analysis of human balance.

Center of pressure (COP) traces have been used to investigate the dynamics of human balance. In this paper we employ a wavelet-based multifractal methodology to identify structural differences in mediolateral and anterioposterior sway between COP traces of healthy and Parkinson's patients. Two statistical techniques are used to summarize the differences in multifractal spectrum (MFS) for both groups. The first technique is a multivariate repeated measures analysis on estimated MF spectra for subjects. The second technique obtains two characteristic measures from each subject's estimated MFS: (i) location and (ii) half-width of the spectrum. These measures present an intuitive summary of the MFS for each subject, allowing for statistical comparisons between the two groups. Both analyzes lead to significant discrimination between Parkinson versus healthy subject's MFS. We find that COP time series of Parkinson patients exhibit a greater degree of roughness as compared to healthy subjects' COP traces. Furthermore, MFS for Parkinson patients are narrower, suggesting a reduction in complexity as compared to the healthy group. The methodology presented here may be helpful in development of clinically relevant measures, including the assessment of severity of conditions as the measures developed here correlate with standard severity measures.

Analysis of Variance↗