Artificial Neural Networks: Prediction of mortality/survival in gastroenterology

Abstract Artificial Neural Networks have increasingly been used for building prognostic models in several medical domains. We discuss these networks along with their drawbacks and review their use in the prediction of survival and mortality compared to conventional models or gold standard disease scoring techniques in the field of gastroenterology. Our paper also emphasizes the development of web enabled multi-institutional neural network models based on two sets of variables for harnessing the potential of machine tools in the prognosis of patients.