Desarrollo de un sistema de IoT con machine learning para la detección temprana de crisis epilépticas

Abstract

This research addresses the development of a system based on the Internet of Things (IoT) and optimized through machine learning techniques for the early detection of epileptic seizures. An algorithm was developed to capture biological signals through wearable sensors and process them within an IoT architecture. Data obtained from the Hospital de la Misericordia in Bogotá were used to validate various models, with a decision tree and a multilayer perceptron neural network standing out, achieving accuracies of 92.47% and 93.56%, respectively. In real clinical settings, these models achieved an accuracy of 65%. The proposed architecture allows for real-time preventive alerts, offering a potentially effective tool to improve the quality of life for people with epilepsy.