Creation of an artificial neural network model to predict diseases based on patient symptomatology

Abstract

The main purpose of this research is the creation of an artificial neural network aimed at predicting common diseases in patients who reside in marginalized or difficult-to-access areas such as: influenza, sinusitis, stomach and urinary tract infections. The basis of this work focuses on information collected from the patient’s clinical history, as well as a detailed symptom questionnaire, both integrated into a telemedicine system developed by the authors and installed in the marginalized areas of northern Veracruz, Mexico. The effectiveness of the artificial neural network is evaluated to determine its accuracy in detecting diseases, through tests carried out on the algorithm. The high precision obtained in the diagnosis of the artificial neural network supports it as an efficient complementary technological tool. The results obtained are contrasted with the disease diagnoses issued by specialist doctors, thus establishing the validity and reliability of the neural network in this specific context of medical care in marginalized areas.