APPLICATIONS OF MACHINE LEARNING IN CARDIOLOGY FOR OBESE PATIENTS UNDERGOING WEIGHT LOSS THERAPIES: A SYSTEMATIC REVIEW

Abstract

This systematic review aimed to analyze the applications and results of machine learning (ML) techniques in cardiology for obese patients undergoing weight loss therapies. The research was conducted according to the PRISMA 2020 guidelines and used the PICO framework to formulate the guiding question. Searches were performed in the PubMed, Scopus, Web of Science, ScienceDirect, SciELO, and CAPES databases, covering the period from January 2019 to November 2025. 1,642 records were identified, of which 13 studies met the eligibility criteria. Methodological quality was assessed using the QUADAS-2 instrument, allowing the identification of bias risks and applicability concerns. The results showed that ML techniques have transformed cardiovascular diagnosis and prognosis, especially through supervised algorithms such as support vector machines, random forests, and convolutional neural networks, which achieved accuracies greater than 90% in predicting heart failure and detecting cardiac abnormalities. Significant progress was observed in the integration of clinical, genetic, and imaging data, promoting greater accuracy in risk stratification. However, a scarcity of specific studies focusing on obese patients undergoing weight-loss therapies was found, which limits the generalizability of the results and highlights a relevant scientific gap.