Análisis de señales electrocardiográficas para la detección temprana del síndrome metabólico mediante machine learning

Abstract

This paper develops an innovative method for the early identification of metabolic syndrome (MetS) through the analysis of electrocardiographic (ECG) signals and the use of advanced machine learning techniques. MS, characterised by a series of metabolic irregularities that significantly increase the risk of cardiovascular disease, type 2 diabetes and strokes, is currently diagnosed through invasive and tedious methods for patients, such as the oral glucose tolerance test. This project aims to replace these practices with a non-invasive, agile and efficient approach based on automated analysis of ECG signals. Since MS affects the autonomic modulation of the heart, it generates detectable alterations in the ECG that can be used for early detection. The study encompasses the development of a system capable of processing ECG signals, removing noise and extracting relevant parameters in the time, frequency and non-linear domains. These parameters are used in the training and testing of two machine learning models: Random Forest and Support Vector Machine (SVM).