Zenodo (CERN European Organization for Nuclear Research) · 2012 · 10 citations · 0 references
ComputingBiomedical InformaticsData ScienceClinical Decision SupportData MiningMachine LearningCase-based LearningData InterpretationData PreprocessingDiagnostic Decision SupportExpert SystemsVisual Data MiningAutomated Retinal Disease ClassificationCase-based Decision SupportCase-based ReasoningCase-based Reasoning SupportClinical Prediction RulesDiagnostic Prediction ModelsExplanatory Case Study
Diabetes is one of the high prevalence diseases worldwide with increased number of complications, with retinopathy as one of the most common one. This paper describes how data mining and case-based reasoning were integrated to predict retinopathy prevalence among diabetes patients in Malaysia. The knowledge base required was built after literature reviews and interviews with medical experts. A total of 140 diabetes patients- data were used to train the prediction system. A voting mechanism selects the best prediction results from the two techniques used. It has been successfully proven that both data mining and case-based reasoning can be used for retinopathy prediction with an improved accuracy of 85%.