DIAGNOSIS AND CLASSIFICATION OF OCULAR DISEASES USING SPATIAL CORRELATION NETWORKS
Main Article Content
Keywords
Ocular disease, spatial correlation network, CNN, eye disease
Abstract
Ocular diseases, ranging from diabetic retinopathy to glaucoma, pose a substantial global healthcare challenge, demanding accurate and timely diagnosis. The conventional manual interpretation of ocular imaging, while practiced with expertise, is subjective and labor-intensive. This research addresses the pressing need for automation in ocular disease diagnosis and classification through the innovative application of Spatial Correlation Networks (SCNs).
Our objectives encompass the development of a dedicated SCN framework, comprehensive dataset collection, standardized annotation, and the automated detection and classification of ocular diseases. Performance evaluation considers accuracy, sensitivity, specificity, and computational efficiency, alongside the potential for clinical integration. Ethical and regulatory dimensions are thoughtfully explored. The author has developed a patient-level multi-label OD (PLML_ODs) classification model based on a spatial correlation network (SCNet). This model takes into account the patient-level diagnosis by combining data from both eyes and performing multi-label ODs classification. The PLML_ODs model comprises three key components: a backbone convolutional neural network (CNN) used for feature extraction, specifically DenseNet-169; the SCNet for capturing feature correlations; and a classifier for generating classification scores.
DenseNet-169 is responsible for extracting two distinct sets of attributes, one from each of the left and right CFI. Subsequently, the SCNet records correlations between these two sets of features at a pixel-by-pixel level. After the attributes have been analyzed, they are integrated to create a patient-level representation. This patient-level representation is utilized throughout the entire process of ODs categorization.
To evaluate the effectiveness of the PLML_ODs model, a soft margin loss function is applied to a publicly accessible dataset. The results demonstrate a significant improvement in classification performance when compared to several baseline approaches.
References
2. W. Xiao, H. Xi, H. W. Jing, R. L. Duo, Z. Yi et al., “Screening and identifying hepatobiliary diseases through deep learning using ocular images: A prospective, multicentre study,” The Lancet Digital Health, vol. 3, no. 2, pp. e88–e97, 2021.
3. Z. Wang, Z. Yuanfu, Y. Mudi, M. Yan, Z. Wenping et al., “Automated segmentation of macular edema for the diagnosis of ocular disease using deep learning method,” Scientific Reports, vol. 11, no. 1, pp. 1–12, 2021.
4. H. Malik, S. F. Muhammad, K. Adel, A. Adnan, N. Q. Junaid et al., “A comparison of transfer learning performance versus health experts in disease diagnosis from medical imaging,” IEEE Access, vol. 8, no. 1, pp. 139367–139386, 2020.
5. D. Wang and W. Liejun, “On OCT image classification via deep learning,” IEEE Photonics Journal, vol. 11, no. 5, pp. 1–14, 2019.
6. T. Nazir, I. Aun, J. Ali, M. Hafiz, H. Dildar et al., “Retinal image analysis for diabetes-based eye disease detection using deep learning,” Applied Sciences, vol. 10, no. 18, pp. 6185, 2020.
7. X. Meng, X. Xiaoming, Y. Lu, Z. Guang, Y. Yilong et al., “Fast and effective optic disk localization based on convolutional neural network,” Neurocomputing, vol. 312, no. 2, pp. 285–295, 2018.
8. L. Faes, K. W. Siegfried, J. F. Dun, L. Xiaoxuan, K. Edward et al., “Automated deep learning design for medical image classification by health-care professionals with no coding experience: A feasibility study,” The Lancet Digital Health, vol. 1, no. 5, pp. 232–242, 20
9. J. Kanno, S. Takuhei, I. Hirokazu, I. Hisashi, Y. Yuji et al., “Deep learning with a dataset created using kanno saitama macro, a self-made automatic foveal avascular zone extraction program,” Journal of Clinical Medicine, vol. 12, no. 1, pp. 183, 2023.
10. P. Li, L. Lingling, G. Zhanheng and W. Xin, “AMD-Net: Automatic subretinal fluid and hemorrhage segmentation for wet age-related macular degeneration in ocular fundus images,” Biomedical Signal Processing and Control, vol. 80, no. 2, pp. 104262, 2023.
11. J. He, L. Cheng, Y. Jin, O. Yu and G. Lixu, “Multi-label ocular disease classification with a dense correlation deep neural network,” Biomedical Signal Processing and Control, vol. 63, no. 24, pp. 102167, 2021.
12. A.H. Khan, H. Malik, W. Khalil, S. K. Hussain, T. Anees et al., “Spatial correlation module for classification of multi-label ocular diseases using color fundus images,” Computers, Materials & Continua, vol. 76, no. 1, pp. 133–150, 2023.
13. S. Ahlam, E. M. Senan and H. S. A. Shatnawi, “Automatic classification of colour fundus images for prediction eye disease types based on hybrid features,” Diagnostics, vol. 13, no. 10, pp. 1706, 2023.
14. C. C. Yang, D. Y. Hsu and C. H. Chou, “Predicting the onset of diabetes with machine learning methods,” Journal of Personalized Medicine, vol. 13, no. 3, pp. 406, 2023.
15. J. Y. Choi, K. Y. Tae, G. S. Jeong, K. Jiyong, T. U. Terry et al., “Multi-categorical deep learning neural network to classify retinal images: A pilot study employing small database,” PLoS One, vol. 12, no. 11, pp. e0187336, 2017.
16. R. Fan, A. Kamran, B. Christopher, C. Mark, B. Nicole et al., “Detecting glaucoma from fundus photographs using deep learning without convolutions: Transformer for improved generalization,” Ophthalmology Science, vol. 3, no. 1, pp. 100233, 2023.
17. V. Mayya, K. Uma, K. S. Divyalakshmi and U. A. Rajendra, “An empirical study of preprocessing techniques with convolutional neural networks for accurate detection of chronic ocular diseases using fundus images,” Applied Intelligence, vol. 53, no. 2, pp. 1548–1566, 2023.
Article Sidebar
Article Details
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
All articles published in JPTCP are licensed under Copyright Creative Commons Attribution-NonCommercial 4.0 International License.
Sayyid Kamran Hussain
Department of Computer Science, TIMES Institute, Multan, Pakistan.
Sadaqat Ali Ramay
Department of Computer Science, TIMES Institute, Multan, Pakistan
Madiha Sarwar
Department of Information Technology, University of Sargodha, Sargodha, Pakistan.
Tahir Abbas
Department of Computer Science, TIMES Institute, Multan, Pakistan.
Muhammad Kaleem
Department of Information Technology, University of Sargodha, Sargodha, Pakistan.