Comparación de modelos de inteligencia artificial para la detección de cardiopatías usando imágenes de resonancia magnética
DOI:
https://doi.org/10.24054/rcta.v1i47.4285Palabras clave:
cardiopatías, resonancia magnética cardíaca, YOLOv8, Random Forest, CNN, inteligencia artificialResumen
En este trabajo se compararon tres técnicas de inteligencia artificial para la detección binaria (sano/enfermo) de cardiopatías mediante cortes axiales de resonancia magnética cardíaca. Usando un conjunto de 150 pacientes en formato NIfTI, las imágenes se preprocesaron (normalización, reescalado a 128 × 128, conversión a RGB y aumento de datos) y se dividieron en una proporción 80/20 por paciente. Se evaluaron Random Forest con descriptores radiómicos GLCM y estadísticos de primer orden, una red neuronal convolucional (CNN) y un modelo basado en YOLOv8 adaptado a clasificación binaria. Los modelos se compararon mediante accuracy, precision, recall, F1-score y AUC, y se aplicaron técnicas de explicabilidad (SHAP, Grad-CAM, Integrated Gradients y sensibilidad a la oclusión) para validar la coherencia anatómica de las predicciones. En conjunto, los resultados indican que los enfoques de aprendizaje profundo preentrenado, como YOLOv8, ofrecen ventajas sustanciales en términos de precisión e interpretabilidad, posicionándose como una alternativa prometedora para el desarrollo de sistemas inteligentes de apoyo al diagnóstico de cardiopatías estructurales.
Descargas
Referencias
World Health Organization, “Cardiovascular diseases (CVDs),” 2023. [Online]. Available: https://www.who.int/news-room/fact-sheets/detail/cardiovascular-diseases-(cvds)
Ministerio de Salud y Protección Social, “Enfermedades cardiovasculares.” [Online]. Available: https://www.minsalud.gov.co/salud/Paginas/Enfermedades-cardiovasculares.aspx
Ministerio de Salud y Protección Social, “Minsalud conmemora el Día Mundial del Corazón,” Sep. 29, 2022. [Online]. Available: https://www.minsalud.gov.co/Paginas/Minsalud-conmemora-el-dia-mundial-del-Corazon.aspx
Ministerio de Salud y Protección Social, “Plan Nacional de Salud Pública 2022–2031.” [Online]. Available: https://www.minsalud.gov.co/plandecenal/Paginas/PDSP-2022-2031.aspx
D. De Santis, A. Trivisonno, and E. G. Caiani, “Accelerated deep learning-based function assessment in cardiovascular magnetic resonance: a clinical comparison with conventional cine sequences,” La Radiologia Medica, 2025, doi: https://doi.org/10.1007/s11547-025-02019-6
A. Argentiero, G. Muscogiuri, M. G. Rabbat, C. Martini, N. Soldato, P. Basile, et al., “The applications of artificial intelligence in cardiovascular magnetic resonance—A comprehensive review,” Journal of Clinical Medicine, vol. 11, no. 10, p. 2866, 2022, doi: https://doi.org/10.3390/jcm11102866
O. Bernard et al., “Deep learning techniques for automatic MRI cardiac multi-structures segmentation and diagnosis: Is the problem solved?” IEEE Trans. Med. Imaging, vol. 37, no. 11, pp. 2514–2525, 2018. https://doi.org/10.1109/TMI.2018.2837502.
M. Alsharqi and E. R. Edelman, “Artificial Intelligence in Cardiovascular Imaging and Interventional Cardiology: Emerging Trends and Clinical Implications,” J. Soc. Cardiovasc. Angiogr. Interv., vol. 4, no. 3, p. 102558, 2025. https://doi.org/10.1016/j.jscai.2024.102558
Y.-R. Wang, K. Yang, Y. Wen, P. Wang, Y. Hu, Y. Lai, Y. Wang, K. Zhao, S. Tang, A. Zhang, H. Zhan, M. Lu, X. Chen, S. Yang, Z. Dong, Y. Wang, H. Liu, L. Zhao, L. Huang, Y. Li, L. Wu, Z. Chen, Y. Luo, D. Liu, et al., “Screening and diagnosis of cardiovascular disease using artificial intelligence-enabled cardiac magnetic resonance imaging,” Nature Medicine, vol. 30, pp. 1471–1480, 2024, doi: https://doi.org/10.1038/s41591-024-02971-2
R. Jafari, R. Azad, and S. Minaee, “Automatic diagnosis of myocarditis disease in cardiac MRI modality using deep transformers and explainable artificial intelligence,” arXiv preprint arXiv:2210.14611, 2022. https://doi.org/10.48550/arXiv.2210.14611
D. Li, H. Zhang, and C. Wu, “An algorithm for cardiac disease detection based on magnetic resonance imaging using an improved SA-YOLO model,” Scientific Reports, vol. 15, 88567, 2025, doi: https://doi.org/10.1038/s41598-025-88567-3
M. U. Rehman, S. Akbar, and A. Imran, “Predicting coronary heart disease with advanced machine learning classifiers for improved cardiovascular risk assessment,” Scientific Reports, vol. 15, 96437, 2025, doi: https://doi.org/10.1038/s41598-025-96437-1
T. Tsampras, T. Karamanidou, G. Papanastasiou, and T. G. Stavropoulos, “Deep learning for cardiac imaging: Focus on myocardial diseases: A narrative review,” Hellenic Journal of Cardiology, 2024. https://doi.org/10.1016/j.hjc.2024.12.002
F. Cheng et al., “Learning directional feature maps for cardiac MRI segmentation,” in Proc. Int. Conf. Med. Image Comput. Comput.-Assist. Intervent. (MICCAI), Cham, Switzerland: Springer, Sep. 2020, pp. 108–117. https://doi.org/10.1007/978-3-030-59719-1_11
D. M. Alsekait et al., “Heart-Net: A multi-modal deep learning approach for diagnosing cardiovascular diseases,” Comput. Mater. Contin., vol. 80, no. 3, 2024. https://doi.org/10.32604/cmc.2024.054591
R. K. Sahu, D. R. Sahu, and B. K. Pattanayak, “A hybrid segmentation and classification CAD framework for myocardial infarction detection using cardiac MRI,” Scientific Reports, vol. 15, 98893, 2025, doi: https://doi.org/10.1038/s41598-025-98893-1
X. Liu, F. Xing, H. K. Gaggin, C. C. J. Kuo, G. El Fakhri, and J. Woo, “Successive subspace learning for cardiac disease classification with two-phase deformation fields from cine MRI,” in Proc. 2023 IEEE 20th Int. Symp. Biomed. Imaging (ISBI), Apr. 2023, pp. 1–5. https://doi.org/10.1109/ISBI53787.2023.10230746
J. Xing, S. Wang, K. C. Bilchick, A. R. Patel, and M. Zhang, “Joint deep learning for improved myocardial scar detection from cardiac MRI,” in Proc. 2023 IEEE 20th Int. Symp. Biomed. Imaging (ISBI), Apr. 2023, pp. 1–5. https://doi.org/10.1109/ISBI53787.2023.10230541
G. P. Veldhuizen, T. Lenz, D. Cifci, M. van Treeck, J. Clusmann, Y. Chen, C. V. Schneider, T. Luedde, P. W. de Leeuw, A. El-Armouche, D. Truhn, and J. N. Kather, “Deep learning can predict cardiovascular events from liver imaging,” JHEP Reports, 2025, Art. no. 101427. https://doi.org/10.1016/j.jhepr.2025.101427
R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, and D. Batra, “Grad-CAM: Visual explanations from deep networks via gradient-based localization,” International Journal of Computer Vision, vol. 128, pp. 336–359, 2019, https://doi.org/10.1007/s11263-019-01228-7
J. Liu, A. Wei, L. Cao, X. He, and C. Tang, “Contrastive trustworthy prototype learning for multi-modality myocardial pathology segmentation,” Applied Soft Computing, vol. 173, Art. no. 112909, 2025, doi: https://doi.org/10.1016/j.asoc.2025.112909
Z. Dong, Y. Tang, P. Sun, G. Yin, K. Zhao, X. Ma, and S. Zhao, “Early identification of myocardial microstructural alterations in hypertrophic cardiomyopathy with in vivo cardiac diffusion-tensor imaging,” Radiology: Cardiothoracic Imaging, vol. 7, no. 1, Art. no. e240009, 2025, doi: https://doi.org/10.1148/ryct.240009
Descargas
Publicado
Número
Sección
Licencia
Derechos de autor 2026 Jorge Eliecer Gomez Gomez, Juan José Gúzman Pineda, Velssy Liliana Hernández Riaño
Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0.
