Generalization of Cardiomyopathy Classification Models Based on Feature Descriptors from Magnetic Resonance Imaging

• Stephani S. H. Costa USP
• Vagner Mendonça Gonçalves USP / IFSP
• Matheus A. O. Ribeiro USP
• Fátima L. S. Nunes USP

DOI: https://doi.org/10.5753/sbcas.2025.7899

Resumo

Supervised Machine Learning (SML) models can help physicians to compose more accurate diagnoses. Due to the diversity of machines, systems, and protocols, exams conducted at different centers may have high variability, decreasing the generalization ability of these models. This paper aims to evaluate generalization ability of SML models for cardiomyopathy classification in Cardiac Magnetic Resonance Imaging (CMRI), using a set of left ventricle morphological and motion features. We performed cross-validation tests on two public CMRI databases, comparing intraand inter-dataset performances. The results are promising and demonstrate that the implemented features can contribute to building generalizable classification models.

Referências

Atehortúa, A., Romero, E., and Garreau, M. (2022). Characterization of motion patterns by a spatio-temporal saliency descriptor in cardiac cine MRI. Computer Methods and Programs in Biomedicine, 218(106714):1–12.

Bergstra, J. and Bengio, Y. (2012). Random search for hyper-parameter optimization. Journal of Machine Learning Research, 13(10):281–305.

Bernard, O. et al. (2018). Deep Learning techniques for automatic MRI cardiac multi-structures segmentation and diagnosis: is the problem solved? IEEE Transactions on Medical Imaging, 37(11):2514–2525.

Breiman, L. (2001). Random forests. Machine Learning, 45(1):5–32.

Chawla, N. V., Bowyer, K. W., Hall, L. O., and Kegelmeyer, W. P. (2002). SMOTE: Synthetic Minority Over-Sampling Technique. Journal of Artificial Intelligence Research, 16(1):321–357.

Chen, T. and Guestrin, C. (2016). XGBoost: a scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 785–794. ACM.

Cortes, C. and Vapnik, V. (1995). Support-vector networks. Machine Learning, 20(3):273–297.

Diao, K. et al. (2023). Multi-channel deep learning model-based myocardial spatial–temporal morphology feature on cardiac MRI cine images diagnoses the cause of LVH. Insights into Imaging, 14(70):1–11.

Gheyas, I. A. and Smith, L. S. (2010). Feature subset selection in large dimensionality domains. Pattern Recognition, 43(1):5–13.

Goto, S. et al. (2023). Multinational federated learning approach to train ECG and echocardiogram models for hypertrophic cardiomyopathy detection. Circulation, 146(10):755–769.

Hastie, T., Tibshirani, R., and Friedman, J. (2009). The Elements of Statistical Learning. Springer, 2nd edition.

Izquierdo, C. et al. (2021). Radiomics-based classification of left ventricular non-compaction, hypertrophic cardiomyopathy, and dilated cardiomyopathy in cardiovascular magnetic resonance. Frontiers in Cardiovascular Medicine, 8(764312):1–10.

Jolliffe, I. T. (2002). Principal Component Analysis. Springer, 2nd edition.

Kagiyama, N., Tokodi, M., and Sengupta, P. P. (2022). Machine learning in cardiovascular imaging. Heart Failure Clinics, 18(2):245–258.

Knoll, F. et al. (2019). Assessment of the generalization of learned image reconstruction and the potential for transfer learning. Magnetic Resonance in Medicine, 81(1):116–128.

Kraskov, A., Stögbauer, H., and Grassberger, P. (2004). Estimating mutual information. Physical Review E, 69(6):066138–1–066138–16.

Lemaı̂tre, G., Nogueira, F., and Aridas, C. K. (2017). Imbalanced-learn: a Python toolbox to tackle the curse of imbalanced datasets in machine learning. Journal of Machine Learning Research, 18(17):1–5.

Linardos, A. et al. (2022). Federated learning for multi-center imaging diagnostics: a simulation study in cardiovascular disease. Scientific Reports, 12(3551):1–12.

Liu, Q. et al. (2023). Papillary-muscle-derived radiomic features for hypertrophic cardiomyopathy versus hypertensive heart disease classification. Diagnostics, 13(9):1544:1–1544:15.

Manning, C. D., Raghavan, P., and Schütze, H. (2008). Introduction to Information Retrieval. Cambridge University Press, Cambridge, England.

Martín-Isla, C. et al. (2023). Deep learning segmentation of the right ventricle in cardiac MRI: the M&Ms challenge. IEEE Journal of Biomedical and Health Informatics, 27(7):3302–3313.

Ministério da Saúde (2023). “Use o coração para vencer as doenças cardiovasculares”: 29/9 – Dia Mundial do Coração. Available at: [link]. Accessed: 4 Mar 2025.

Moreno, A., Rodriguez, J., and Martínez, F. (2019). Regional multiscale motion representation for cardiac disease prediction. In 2019 XXII Symposium on Image, Signal Processing and Artificial Vision (STSIVA), pages 1–5. IEEE.

Ng, W. W. Y. et al. (2007). Image classification with the use of radial basis function neural networks and the minimization of the localized generalization error. Pattern Recognition, 40(1):19–32.

Pedregosa, F. et al. (2011). Scikit-learn: machine learning in Python. Journal of Machine Learning Research, 12(85):2825–2830.

Ribeiro, M. A. O., Gutierrez, M. A., and Nunes, F. L. S. (2023). Improving deep learning shape consistency with a new loss function for left ventricle segmentation in cardiac MRI. In 2023 IEEE 36th International Symposium on Computer-Based Medical Systems (CBMS). IEEE.

Ribeiro, M. A. O. and Nunes, F. L. S. (2021). Evaluating the pre-processing impact on the generalization of deep learning networks for left ventricle segmentation. In 2021 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pages 3505–3512. IEEE.

Shen, D., Wu, G., and Suk, H.-I. (2017). Deep learning in medical image analysis. Annual Review of Biomedical Engineering, 19:221–248.

Sivaprasad, R. et al. (2022). Heart disease prediction and classification using machine learning and transfer learning model. In 2022 International Conference on Automation, Computing and Renewable Systems (ICACRS), pages 595–601. IEEE.

Sociedade Brasileira de Cardiologia (2025). Cardiômetro: monitoramento de mortes por doenças cardiovasculares no Brasil. Available at: [link]. Accessed: 4 Mar 2025.

Sundaram, D. S. B. et al. (2021). Natural language processing based machine learning model using cardiac MRI reports to identify hypertrophic cardiomyopathy patients. In 2021 Design of Medical Devices Conference, pages V001T03A005–1–V001T03A005–5. The American Society of Mechanical Engineers.

Wibowo, A. et al. (2022). Cardiac disease classification using two-dimensional thickness and few-shot learning based on magnetic resonance imaging image segmentation. Journal of Imaging, 8(7):194:1–194:16.

Xia, C. et al. (2019). A multi-modality network for cardiomyopathy death risk prediction with CMR images and clinical information. In Shen, D. et al., editors, Medical Image Computing and Computer Assisted Intervention – MICCAI 2019, volume 11765 of Lecture Notes in Computer Science, pages 577–585. Springer Nature.

Zhang, X. et al. (2023). Cardiac magnetic resonance radiomics for disease classification. European Radiology, 33(4):2312–2323.

Zhou, M. et al. (2023). Echocardiography-based machine learning algorithm for distinguishing ischemic cardiomyopathy from dilated cardiomyopathy. BMC Cardiovascular Disorders, 23(476):1–10.