We present a technology for the accurate and fast geo-location of medical data and the creation of central data archives, specifically designed for the development of disease risk maps and of other functions for modern epidemiology and surveillance. A WEBGIS system is configured as an Internet web service integrated with connectivity to a Geographical Information System (GIS). We developed for the ULSS Belluno a WEBGIS for the accurate mapping of tick-borne diseases, with specific attention to Lyme borreliosis, which may cause cardiac manifestations as atrioventricular conduction abnormalities, complete atrioventricular block, myocarditis and dilated cardiomiopathy. A first tree-based predictive model has been developed for risk classification of tick bites from 256 samples (data gathered through the Belluno Lyme WEBGIS), with a descriptive accuracy of 91.9% and a predictive accuracy of 75.1%. An experimental risk GIS map was therefore obtained from the model by considering altitude, week of sampling and vegetation type as predictor variables
New WEBGIS technologies for geolocation of epidemiological data: an application for the surveillance of the risk of Lyme borreliosis disease
Furlanello, Cesare;Merler, Stefano;Menegon, Stefano;
2002-01-01
Abstract
We present a technology for the accurate and fast geo-location of medical data and the creation of central data archives, specifically designed for the development of disease risk maps and of other functions for modern epidemiology and surveillance. A WEBGIS system is configured as an Internet web service integrated with connectivity to a Geographical Information System (GIS). We developed for the ULSS Belluno a WEBGIS for the accurate mapping of tick-borne diseases, with specific attention to Lyme borreliosis, which may cause cardiac manifestations as atrioventricular conduction abnormalities, complete atrioventricular block, myocarditis and dilated cardiomiopathy. A first tree-based predictive model has been developed for risk classification of tick bites from 256 samples (data gathered through the Belluno Lyme WEBGIS), with a descriptive accuracy of 91.9% and a predictive accuracy of 75.1%. An experimental risk GIS map was therefore obtained from the model by considering altitude, week of sampling and vegetation type as predictor variablesI documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.