Prediction of Cardiovascular Diseases with Retinal Image Using Deep Learning

Abstract

Cardiovascular diseases (CVDs) are a leading cause of mortality worldwide. Early detection and accurate diagnosis of CVDs are crucial for effective intervention and improved patient outcomes. Retinal imaging has emergedasanon-invasiveandcost-effectivetechniquefor CVD prediction. This study aims to develop a deep learning model using convolutional neural networks (CNNs)andMobileNetarchitecturetopredictCVDsfrom retinal images. The proposed model leverages the capabilities of CNNs to automatically learn relevant features from retinal images and MobileNet\'s lightweight design for efficient deployment

Introduction

Cardiovascular diseases (CVDs) are a leading cause of global mortality, where early detection is key to improving outcomes. Traditional diagnostic methods are often invasive, expensive, and inaccessible, especially in low-resource settings. Retinal imaging offers a promising, non-invasive alternative since the retina’s blood vessels reflect systemic vascular health, including cardiovascular conditions.

This project aims to develop a deep learning model using Convolutional Neural Networks (CNNs) and MobileNet architecture to predict CVDs from retinal images. MobileNet’s lightweight design and CNN’s feature extraction capabilities are leveraged to create an accurate, efficient, and cost-effective diagnostic tool.

The system workflow includes data collection, preprocessing (resizing, normalization, augmentation), model training, saving, and prediction. Users can register, log in, upload retinal images, and view prediction results securely.

Two models were tested: a standard CNN and MobileNet. While the CNN achieved 79% accuracy in classifying retinal images into six categories related to eye and cardiovascular conditions, MobileNet outperformed it with 90% accuracy, showing better classification consistency and suitability for CVD prediction.

Conclusion

This study demonstrates the potential of deep learningmodels in predicting cardiovascular diseases through the classification of retinal images. The comparative analysis between Convolutional Neural Networks (CNN) and MobileNet revealed that while both models are capable of distinguishing between the six classes (ARMD, DN, DR, MH, NORMAL, and ODC), MobileNet significantly outperforms CNN in terms of accuracy, achieving 90% compared to CNN\'s 79%. The higher accuracy and more precise classifications observed with MobileNet suggest that it is better equipped to handle the intricacies of retinal image data, making it a more reliable model for the early detection and prediction of cardiovascular diseases. These findings underscore the importance of choosing the appropriate deep learning architecture for medical image classification tasks, as it can substantially impact the accuracy and reliability of the predictions. Future work could focus on further enhancingthe model\'s performance, exploring additional data augmentation techniques, or integrating more advanced models to improve prediction accuracy even further.

References

[1] Gulshan, V., et al. (2016). \"Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs.\" JAMA, 316(22), 2402–2410. doi:10.1001/jama.2016.17216. [2] Esteva, A., et al. (2017). \"Dermatologist-Level Classification of Skin Cancer with Deep Neural Networks.\" Nature, 542(7639), 115–118. doi:10.1038/nature21056. [3] Liu, S., et al. (2019). \"Retinal Image Classification Using Deep Learning Algorithms for Early Detection of Diabetic Retinopathy.\" IEEE Access, 7, 71534-71547. doi:10.1109/ACCESS.2019.2918671. [4] Tan, M., & Le, Q. (2019). \"EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks.\" Proceedings of the International Conference on Machine Learning (ICML), 6105–6114. Retrieved from https://arxiv.org/abs/1905.11946. [5] He, K., et al. (2016). \"Deep Residual Learning for Image Recognition.\" Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 770–778. doi:10.1109/CVPR.2016.90. [6] Szegedy, C., et al. (2016). \"Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning.\" Proceedings of the AAAI Conference on Artificial Intelligence, 4278–4284. Retrieved from https://arxiv.org/abs/1602.07261. [7] Ronneberger, O., Fischer, P., & Brox, T. (2015). \"U-Net: Convolutional Networks for Biomedical Image Segmentation.\" Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), 234–241. doi:10.1007/978-3-319-24574-4_28. [8] Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). \"ImageNet Classification with Deep Convolutional Neural Networks.\" Advances in Neural InformationProcessing Systems (NIPS), 1097–1105. doi:10.1145/3065386. [9] Simonyan, K., & Zisserman, A. (2015). \"Very Deep Convolutional Networks for Large-Scale Image Recognition.\" International Conference on Learning Representations (ICLR). Retrieved from https://arxiv.org/abs/1409.1556. [10] Szegedy, C., et al. (2015). \"Going Deeper with Convolutions.\" Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 1–9. doi:10.1109/CVPR.2015.7298594. [11] LeCun, Y., Bengio, Y., & Hinton, G. (2015). \"Deep Learning.\" Nature, 521(7553), 436–444. doi:10.1038/nature14539. [12] Litjens, G., et al. (2017). \"A Survey on Deep Learning in Medical Image Analysis.\" Medical Image Analysis, 42, 60–88. doi:10.1016/j.media.2017.07.005. [13] J. Li, X. Liu, and Y. Wang, \"Deep Learning for Medical Image Segmentation: A Review,\" IEEE Transactions on Neural Networks and Learning Systems, vol. 31, no. 8, pp. 2762-2779, 2020. [14] K. He, X. Zhang, S. Ren, and J. Sun, \"Deep Residual Learning for Image Recognition,\" IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 770-778, 2016. [15] M. Abadi, P. Barham, J. Chen, et al., \"TensorFlow: A System for Large-Scale Machine Learning,\" 12th USENIX Symposium on Operating Systems Design and Implementation (OSDI), pp. 265-283, 2016.

Copyright

Copyright © 2025 Padakanti Pranathi, Pathuri Manaswi Reddy, Vanteru Nikesh Reddy, Naini Charan Reddy. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.