Aplicações de machine learning no diagnóstico e prognóstico da doença de Chagas via eletrocardiograma: uma revisão sistemática

Abstract

Chagas disease, caused by the protozoan Trypanosoma cruzi, remains a significant public health challenge, particularly in Latin America, where it affects millions of people and leads to severe cardiac complications such as chronic Chagas cardiomyopathy (CCC). Early diagnosis and accurate prognosis are critical for improving patient outcomes, yet traditional diagnostic methods, such as electrocardiogram (ECG) interpretation, are often subjective and dependent on clinical expertise. In recent years, machine learning (ML) techniques have emerged as powerful tools for analyzing biomedical signals, offering the potential to automate ECG analysis and enhance diagnostic accuracy. This systematic review explores the state-of-the-art in ML applications for Chagas disease diagnosis and prognosis using ECG data. Key ML approaches, including convolutional neural networks (CNNs), long short-term memory networks (LSTMs), and support vector machines (SVMs), are discussed, highlighting their effectiveness in detecting cardiac abnormalities and predicting adverse outcomes. Despite promising results, challenges such as the need for large, annotated datasets, model interpretability, and clinical validation remain. This review consolidates existing research, identifies gaps, and provides insights for future developments in ML-based solutions for Chagas disease, aiming to support the implementation of these technologies in clinical practice.