Detecção simultânea de múltiplas regiões de alto e baixo risco em mapas de dados pontuais de caso-controle

Abstract

The spatial scan statistic is the most commonly used technique for detecting clusters. Several extensions of this technique have been developed, seeking flexibility in the search space of the clusters, as well as improvement in the accuracy of the detection. One of the recents extensions of the scan statistic, called Voronoi Based Scan (VBScan), aims to detect clusters in a map consisting of point event data (case-control). Other efforts are being employed in order to use the scan statistic in the problem of detection of multiple clusters. In this work, we developed a method based on the VBScan to detect the partition of a map consisting of point-event data (case-control). The method aims to identify all significant multiple anomalies, which can be of high or low risk. In this new method, we use the VBScan recursively on the map with case-control point-event data, atrav´es de um procedimento bi-objetivo.Our method was tested on different simulated maps and partitioned intodifferent numbers of components. We evaluate the power of detection and matching (a measure of overlap between the true and detected partitions). We also apply the method on two case studies.The method is fast, with good accuracy determination.