Combining data and meta-analysis to build Bayesian networks for clinical decision support. Yet, B., Perkins, Z., B., Rasmussen, T., E., Tai, N., R., M., & Marsh, D., W., R. Journal of biomedical informatics, 52:373-85, 12, 2014.
Combining data and meta-analysis to build Bayesian networks for clinical decision support. [link]Website  doi  abstract   bibtex   
Complex clinical decisions require the decision maker to evaluate multiple factors that may interact with each other. Many clinical studies, however, report 'univariate' relations between a single factor and outcome. Such univariate statistics are often insufficient to provide useful support for complex clinical decisions even when they are pooled using meta-analysis. More useful decision support could be provided by evidence-based models that take the interaction between factors into account. In this paper, we propose a method of integrating the univariate results of a meta-analysis with a clinical dataset and expert knowledge to construct multivariate Bayesian network (BN) models. The technique reduces the size of the dataset needed to learn the parameters of a model of a given complexity. Supplementing the data with the meta-analysis results avoids the need to either simplify the model - ignoring some complexities of the problem - or to gather more data. The method is illustrated by a clinical case study into the prediction of the viability of severely injured lower extremities. The case study illustrates the advantages of integrating combined evidence into BN development: the BN developed using our method outperformed four different data-driven structure learning methods, and a well-known scoring model (MESS) in this domain.
@article{
 title = {Combining data and meta-analysis to build Bayesian networks for clinical decision support.},
 type = {article},
 year = {2014},
 keywords = {Bayesian networks,Clinical decision support,Evidence synthesis,Evidence-based medicine,Meta-analysis},
 pages = {373-85},
 volume = {52},
 websites = {http://www.sciencedirect.com/science/article/pii/S1532046414001816},
 month = {12},
 id = {74ba4371-b97e-3f7a-ba19-6b7c62005699},
 created = {2015-04-12T20:17:35.000Z},
 accessed = {2015-04-11},
 file_attached = {false},
 profile_id = {95e10851-cdf3-31de-9f82-1ab629e601b0},
 group_id = {838ecfe2-7c01-38b2-970d-875a87910530},
 last_modified = {2017-03-14T14:27:28.880Z},
 read = {false},
 starred = {false},
 authored = {false},
 confirmed = {true},
 hidden = {false},
 private_publication = {false},
 abstract = {Complex clinical decisions require the decision maker to evaluate multiple factors that may interact with each other. Many clinical studies, however, report 'univariate' relations between a single factor and outcome. Such univariate statistics are often insufficient to provide useful support for complex clinical decisions even when they are pooled using meta-analysis. More useful decision support could be provided by evidence-based models that take the interaction between factors into account. In this paper, we propose a method of integrating the univariate results of a meta-analysis with a clinical dataset and expert knowledge to construct multivariate Bayesian network (BN) models. The technique reduces the size of the dataset needed to learn the parameters of a model of a given complexity. Supplementing the data with the meta-analysis results avoids the need to either simplify the model - ignoring some complexities of the problem - or to gather more data. The method is illustrated by a clinical case study into the prediction of the viability of severely injured lower extremities. The case study illustrates the advantages of integrating combined evidence into BN development: the BN developed using our method outperformed four different data-driven structure learning methods, and a well-known scoring model (MESS) in this domain.},
 bibtype = {article},
 author = {Yet, Barbaros and Perkins, Zane B and Rasmussen, Todd E and Tai, Nigel R M and Marsh, D William R},
 doi = {10.1016/j.jbi.2014.07.018},
 journal = {Journal of biomedical informatics}
}

Downloads: 0