基于空间分布模型的布鲁氏菌病高风险区域预测OA

In order to accurately predict brucellosis high-risk areas,the spatial distribution characteristics of brucellosis in Rucheng County,Hunan Province were systematically analyzed to predict high-risk areas using Kernel Density Estimation(KDE),spatial autocorrelation analysis and the Maximum Entropy Model(MaxEnt),based on surveillance data of animal brucellosis in the county in 2024,considering the factors such as geographical coordinates,farming density and transportation networks.The prediction results were validated through epidemiological traceability investigation and model verification.The results revealed that brucellosis presented significant spatial clustering(Globa1 Moran's I=0.42,P＜0.01)in major hotspot areas in Daping Town,Jiyi Township and other regions;farming density(contributing 31.2％),distance to provincial borders(28.6％)and frequency of genetic sheep transactions(18.4％)were identified as the risk factors by the MaxEnt model;the high-risk area predicted by the model accounted for 23.7％of the area studied,covering 6 townships,the area under the receiver operating characteristic(ROC)curve(AUC)was 0.892,and the spatial overlap rate with actual positive farms/households reached 78.5％;epidemiological traceability investigation confirmed that animal brucellosis spread mainly via cross-provincial introduction and intra-regional cross-transmission.In conclusion,the brucellosis high-risk areas in the county presented a dual-core clustering pattern characterized by"inter-provincial border zones+intensive farming zones",cross-provincial introduction and high farming density were considered as the primary driving factors;and the established spatial distribution model was also applicable for brucellosis risk prediction in non-traditional endemic areas in southern China.A basis and methodological demonstration were provided for zoning of regional brucellosis eradication efforts.