Using a Bayesian belief network to aid differential diagnosis of tropical bovine diseases. McKendrick, I., Gettinby, G., Gu, Y., Reid, S., & Revie, C. Preventive Veterinary Medicine, 47(3):141-156, 11, 2000.
Using a Bayesian belief network to aid differential diagnosis of tropical bovine diseases [link]Website  doi  abstract   bibtex   
The examination of presenting signs has always played an important role in the diagnosis of diseases in animal populations. In the case of diseases of tropical cattle, such expertise is often scarce and confined to those experts with many years of experience. To capture, conserve and disseminate such valuable expert knowledge remains a key challenge to the application of knowledge-based systems in veterinary medicine. In this communication, we explore the use of a Bayesian belief network to quantify expert opinion with a view to estimating the likelihood of various diseases in the presence and absence of certain signs. Information was elicited from a panel of 44 experienced veterinarians to provide the response matrix of 27 signs associated with 20 commonly occurring diseases in sub-Saharan cattle. Using this prior information, estimates of the probability of certain signs occurring with each disease were calculated from which the Bayesian belief network was able to propagate the posterior probability of each of the diseases based on the observed signs. The method as an aid in making diagnosis is discussed. It is recognised that such an approach is but one strand in the process of arriving at a diagnosis. For ease of use and accessibility, the approach has been converted into the software program CaDDiS (Cattle Disease Diagnosis System) which is available for consultation on the World Wide Web.
@article{
 title = {Using a Bayesian belief network to aid differential diagnosis of tropical bovine diseases},
 type = {article},
 year = {2000},
 keywords = {Africa,Bayesian belief network,Cattle disease,Differential diagnosis,Expert system},
 pages = {141-156},
 volume = {47},
 websites = {http://www.sciencedirect.com/science/article/pii/S0167587700001720},
 month = {11},
 id = {6c2d4651-008c-38ba-9985-0715a04f671f},
 created = {2015-04-11T18:46:32.000Z},
 accessed = {2015-04-11},
 file_attached = {false},
 profile_id = {95e10851-cdf3-31de-9f82-1ab629e601b0},
 group_id = {db853ba9-040a-35e0-8f4f-fe9d1f87c7b5},
 last_modified = {2017-03-14T14:28:38.949Z},
 read = {false},
 starred = {false},
 authored = {false},
 confirmed = {true},
 hidden = {false},
 private_publication = {false},
 abstract = {The examination of presenting signs has always played an important role in the diagnosis of diseases in animal populations. In the case of diseases of tropical cattle, such expertise is often scarce and confined to those experts with many years of experience. To capture, conserve and disseminate such valuable expert knowledge remains a key challenge to the application of knowledge-based systems in veterinary medicine. In this communication, we explore the use of a Bayesian belief network to quantify expert opinion with a view to estimating the likelihood of various diseases in the presence and absence of certain signs. Information was elicited from a panel of 44 experienced veterinarians to provide the response matrix of 27 signs associated with 20 commonly occurring diseases in sub-Saharan cattle. Using this prior information, estimates of the probability of certain signs occurring with each disease were calculated from which the Bayesian belief network was able to propagate the posterior probability of each of the diseases based on the observed signs. The method as an aid in making diagnosis is discussed. It is recognised that such an approach is but one strand in the process of arriving at a diagnosis. For ease of use and accessibility, the approach has been converted into the software program CaDDiS (Cattle Disease Diagnosis System) which is available for consultation on the World Wide Web.},
 bibtype = {article},
 author = {McKendrick, I.J and Gettinby, G and Gu, Y and Reid, S.W.J and Revie, C.W},
 doi = {10.1016/S0167-5877(00)00172-0},
 journal = {Preventive Veterinary Medicine},
 number = {3}
}

Downloads: 0