Neural Networks for Automated Smart Health Platforms oriented on Heart Predictive Diagnostic Big Data Systems

摘要

The proposed study is related to an experimental application of a multilayer perceptron (MLP) artificial neural network (ANN) oriented to predictive diagnostic and to automated patient monitoring in homecare assistance. By means of a control room it is possible to check periodically the heart rate of patient sending data from home. The ANN-MLP model is designed and developed by a KNIME workflow providing heart rate prediction with a good data processing performance. The paper describes all the steps of the data processing by highlighting the procedures to read the workflow outputs according with optimal heart rate values. The ANN-MLP predictive diagnostic model is suitable to de-hospitalization process enabling telemedicine and smart sensors able to measure patient physiological data. Finally, the paper provides solutions for big data system integration.