Space-time cluster identification in point processes

Filtering Spatial Point Patterns Using Kernel Densities

Understanding spatial inhomogeneity and clustering in point patterns arises in many contexts, ranging from disease outbreak monitoring to analyzing radiologically-based emphysema in biomedical images. This can often involve classifying individual points as being part of a feature/cluster or as being part of a background noise process. Existing methods for this task can struggle when there are differences in the size and/or density of individual clusters. In this work, we propose employing kernel density estimates of the underlying point process intensity function, using an existing data-driven approach to bandwidth selection, to separate feature points from noise. This is achieved by constructing a null distribution, either through asymptotic properties or Monte Carlo simulation, and comparing kernel density estimates to a given quantile of this distribution. We demonstrate that our method, termed Kernel Density and Simulation based Filtering (KDS-Filt), showed superior performance to existing alternative approaches, especially when there is inhomogeneity in cluster sizes and density. We also show the utility of KDS-Filt for identifying clinically relevant information about the spatial distribution of emphysema in lung computed tomography scans. The KDS-Filt methodology is available as part of the sncp R package, which can be downloaded at https://github.com/stop-pre16/sncp.

Spatio-temporal clustering of the incidence of schizophrenia in Quebec, Canada from 2004 to 2007

Exploring spatio-temporal patterns of disease incidence can help to identify areas of significantly elevated or decreased risk, providing potential etiologic clues. In this study, we present a spatio-temporal analysis of the incidence of schizophrenia in Quebec from 2004 to 2007 using administrative databases from the Régie de l'Assurance Maladie du Quebec and the hospital discharge database. We conducted purely spatial analyses for each age group adjusted by sex for the whole period using SatScan (version 9.1.1). Findings from the study indicated variations in the spatial clustering of schizophrenia according to sex and age. In term of incidence rate, there are high differences between urban and rural-remote areas, as well as between the two main metropolitan areas of the province of Quebec (Island of Montreal and Quebec-City).

Detection of spatial disease clusters with LISA functions

Detection of disease clusters is an important tool in epidemiology that can help to identify risk factors associated with the disease and in understanding its etiology. In this article we propose a method for the detection of spatial clusters where the locations of a set of cases and a set of controls are available. The method is based on local indicators of spatial association functions (LISA functions), particularly on the development of a local version of the product density, which is a second-order characteristic of spatial point processes. The behavior of the method is evaluated and compared with Kulldorff's spatial scan statistic by means of a simulation study. It is shown that the LISA method yields high sensitivity and specificity when it is used to detect simulated clusters of different sizes and shapes. It also performs better than the spatial scan statistic when they are used to detect clusters of irregular shape; however, it presents relatively high type I error in situations where the number of cases is high. Both methods are applied for detecting spatial clusters of kidney disease in the city of Valencia, Spain, in the year 2008.