Anthropometric Predictors and Artificial Neural Networks in the diagnosis of Hypertension
Krzysztof Pytel, Tadeusz Nawarycz, Wojciech Drygas, Lidia Ostrowska-Nawarycz
DOI: http://dx.doi.org/10.15439/2015F246
Citation: Proceedings of the 2015 Federated Conference on Computer Science and Information Systems, M. Ganzha, L. Maciaszek, M. Paprzycki (eds). ACSIS, Vol. 5, pages 287–290 (2015)
Abstract. Artificial Neural Networks (ANNs) play a vital role in the medical field in solving various health problems like estimating the risk of cardiovascular diseases. The article concerns the process of developing ANNs for estimating the risk of arterial hypertension. ANNs proposed in this article use anthropometrical predictors, easy to control for everybody at home without special equipment. In the article we analyze four different models of ANNs and try to find out which model and set of anthropometrical predictors estimates the risk the most accurately. We use dataset of 2485 real cases of patients from the city of Lodz. The experiment was done in the Matlab environment. The performance of the proposed method in terms of accuracy and facility of use shows that ANNs can be effective tools for preliminary tests of arterial hypertension.