Deep Learning Framework for Automatic Cardiac Diagnosis
Keywords:
Cardiovascular Diseases, Convolutional Neural Networks, Deep Learning, Diagnosis, Medical ImagingAbstract
Cardiovascular diseases (CVDs) continue a leading basis of death globally, underscoring the critical necessity for efficient and accurate investigative tools. Deep learning, a subset of artificial intelligence, has arose as an encouraging approach for automatic cardiac diagnosis due to its capability to extract complex features from medical imaging data. In this study, we propose a novel deep learning framework tailored for the automatic diagnosis of cardiac conditions from medical images, such as echocardiograms, MRI scans, or CT scans. The framework employs convolutional neural networks (CNNs) for feature mining and classification tasks. It comprises several key components, including data preprocessing, feature extraction using pre-trained CNN architectures (such as VGG, ResNet, or DenseNet), fine-tuning, and classification using fully connected layers. To address the challenges of limited annotated medical imaging data, transfer learning techniques are incorporated to adapt the pre-trained models to the specific cardiac diagnosis task. Furthermore, to enhance model generalization and interpretability, attention mechanisms and explainable AI techniques are incorporated into the framework. Attention methodologies enable the model to emphasis on relevant regions within the medical images, aiding in more accurate diagnosis. Explainable AI techniques provide insights into the decision-making process of the deep learning model, increasing trust and transparency in its predictions. The proposed framework is evaluated on a diverse dataset comprising cardiac imaging data from multiple modalities and cardiac conditions. Performance metrics such as sensitivity, accuracy, area under the receiver operating characteristic curve (AUC-ROC), and specificity are used to measure the diagnostic accuracy of the model. Experimental results validate the effectiveness of the proposed framework in accurately diagnosing various cardiac conditions, including myocardial infarction, cardiomyopathy, and valvular heart diseases. In conclusion, the developed deep learning framework shows promising potential as an automated tool for cardiac diagnosis, offering rapid and accurate assessment of cardiac conditions from medical imaging data. By leveraging the supremacy of deep learning and incorporating attention mechanisms and explainable AI techniques, the framework aims to improve clinical decision-making and patient results in the field of cardiology.
Downloads
References
Krittanawong, C., Johnson, K. W., Rosenson, R. S., & Wang, Z. (2017). Deep learning for cardiovascular medicine: A practical primer. European Heart Journal, 39(27), 2058-2069.
Madani, A., Arnaout, R., Mofrad, M., & Arnaout, R. (2018). Fast and accurate view classification of echocardiograms using deep learning. NPJ Digital Medicine, 1(1), 1-8.
Ouyang, D., He, B., Ghorbani, A., Yuan, N., Ebinger, J., Langlotz, C. P., & Liu, J. (2020). Video-based AI for beat-to-beat assessment of cardiac function. Nature, 580(7802), 252-256.
Yang, F., Hu, W., Xu, Z., Cao, J., Li, J., & Liu, Z. (2020). An interpretable CNN for identifying cardiac morphology using attention mechanism. Pattern Recognition Letters, 132, 63-70.
Davy, A., Crowe, J. A., & Boyd, A. (2020). The role of explainable AI in cardiovascular disease diagnosis. Journal of Cardiovascular Computed Tomography, 14(3), 206-208.
Wolterink, J. M., Leiner, T., Viergever, M. A., & Isgum, I. (2020). Generative adversarial networks for noise reduction in low-dose CT. IEEE Transactions on Medical Imaging, 36(12), 2536-2545.
Al'Aref, S. J., Anchouche, K., Singh, G., Slomka, P. J., Kolli, K. K., Kumar, A & Min, J. K. (2020). Clinical applications of machine learning in cardiovascular disease and its relevance to cardiac imaging. European Heart Journal, 41(47), 4489-4499.
Hannun, A. Y., Rajpurkar, P., Haghpanahi, M., Tison, G. H., Bourn, C., Turakhia, M. P., & Ng, A. Y. (2019). Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network. Nature Medicine, 25(1), 65-69.
Attia, Z. I., Kapa, S., Lopez-Jimenez, F., McKie, P. M., Ladewig, D. J., Satam, G., & Noseworthy, P. A. (2019). Screening for cardiac contractile dysfunction using an artificial intelligence–enabled electrocardiogram. Nature Medicine, 25(1), 70-74.
Avendi, M. R., Kheradvar, A., Jafarkhani, H. (2016). A combined deep-learning and deformable-model approach to fully automatic segmentation of the left ventricle in cardiac MRI. Medical Image Analysis, 30, 108-119.
Isensee, F., Jaeger, P. F., Kohl, S. A., Petersen, J., & Maier-Hein, K. H. (2018). nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nature Methods, 18(2), 203-211.
Kerkstra, S., van Ginneken, B., & Staring, M. (2018). Deep learning for fully automatic segmentation of the left ventricle cavity from cardiac cine MRI. Medical Image Analysis, 39, 35-46.
Irvin, J., Rajpurkar, P., Ko, M., Yu, Y., Ciurea-Ilcus, S., Chute, C., ... & Lungren, M. P. (2019). CheXpert: A large chest radiograph dataset with uncertainty labels and expert comparison. arXiv preprint arXiv:1901.07031.
Choi, E., Bahadori, M. T., Schuetz, A., Stewart, W. F., & Sun, J. (2016). Doctor AI: Predicting clinical events via recurrent neural networks. arXiv preprint arXiv:1511.05942.
Jaušovec, N., Jaušovec, K., & Pahor, A. (2019). EEG beta asymmetry in stroke patients and its association with post-stroke depression: A pilot study. Clinical Neurophysiology, 130(9), 1679-1685.
Ioffe, S., & Szegedy, C. (2015). Batch normalization: Accelerating deep network training by reducing internal covariate shift. arXiv preprint arXiv:1502.03167.
Rajpurkar, P., Irvin, J., Zhu, K., Yang, B., Mehta, H., Duan, T., ... & Lungren, M. P. (2017). Chexnet: Radiologist-level pneumonia detection on chest x-rays with deep learning. arXiv preprint arXiv:1711.05225.
Zhuang, X., Shen, J., & Yang, Q. (2019). A local encoding-decoding framework for high-fidelity and high-resolution mesh reconstruction. arXiv preprint arXiv:1905.10054.
Gupta, A., & Ayhan, M. S. (2019). Orthogonality-preserving decomposition networks. arXiv preprint arXiv:1910.11945.
Motallebi, S., Baratloo, A., Negida, A., & Lajevardi, L. (2017). Value of plasma/serum neutrophil gelatinase-associated lipocalin in detection of pediatric acute kidney injury; a systematic review and meta-analysis. International Journal of Pediatrics, 5(6), 5089-5103.
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
All papers should be submitted electronically. All submitted manuscripts must be original work that is not under submission at another journal or under consideration for publication in another form, such as a monograph or chapter of a book. Authors of submitted papers are obligated not to submit their paper for publication elsewhere until an editorial decision is rendered on their submission. Further, authors of accepted papers are prohibited from publishing the results in other publications that appear before the paper is published in the Journal unless they receive approval for doing so from the Editor-In-Chief.
IJISAE open access articles are licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. This license lets the audience to give appropriate credit, provide a link to the license, and indicate if changes were made and if they remix, transform, or build upon the material, they must distribute contributions under the same license as the original.