Home // International Journal On Advances in Life Sciences, volume 16, numbers 1 and 2, 2024 // View article
Authors:
Carlos Barroso-Moreno
Hector Espinos Morato
Enrique Puertas
Juan José Beunza Nuin
José Vicente Monmeneu
María P. López-Lereu
David Moratal
Keywords: Cardiomyopathy; Machine Learning Algorithms; right ventricular; Pulmonary Vascular Resistance.
Abstract:
Cardiomyopathy is a condition affecting the heart muscle, poses challenges to effective blood pumping by the heart. While prior research predominantly concentrated on the left ventricle, recent investigations underscore the significance of the right ventricle. This study aims to ascertain the clinical and cardiac parameters influencing right ventricular engagement in both ischaemic and non-ischaemic cardiomyopathy. A database comprising 56,447 subjects, collected between 2008 and 2020 by the ASCIRES Biomedical Group, forms the basis of this investigation. The methodology encompasses two main blocks: the clinical aspect utilizes decision trees for enhanced interpretability, while the technical aspect employs Machine Learning to achieve a higher degree of prediction accuracy. Power Business Intelligence and RapidMiner are the main software tools enabling data transformation and deep data analysis within the EU legal framework for health data. The outcomes reveal the pivotal influence of disparities in aortic artery beat volume, pulmonary vascular volume and aortic arch as key factors. Remarkably, the RapidMiner tool, employing the decision trees algorithm, attains an impressive Area Under the Curve (AUC) of 0.873 with XGBoost and decreases to AUC= 0.791 with SVM. In conclusion, the study underscores the ability to identify crucial clinical variables associated with right ventricular involvement, offering the potential to streamline diagnostic procedures and reduce associated timeframes in cardiomyopathy scenarios.
Pages: 44 to 55
Copyright: Copyright (c) to authors, 2024. Used with permission.
Publication date: June 30, 2024
Published in: journal
ISSN: 1942-2660