Deep Transfer Learning for Automated Diagnosis of Rare Diseases in Medical Imaging
DOI:
https://doi.org/10.64137/XXXXXXXX/IJMST-V1I1P102Keywords:
Deep learning, Transfer learning, Rare diseases, Medical imaging, Convolutional neural networks, Grad-CAM, Classification, Fine-tuningAbstract
Finding and identifying rare diseases through images is a lasting challenge in routine medical work because of the diseases' uncommon nature and limited data. When it comes to seeing how many diagnostic tools work, Convolutional Neural Networks (CNNs) stand out with their great potential. One difficulty with these models is that their outcomes are less impressive when the data they receive is not adequate, which is often true for rare diseases. To overcome this drawback, this research looks at applying deep transfer learning to support stronger identification of rare diseases using knowledge from common or large-scale images. Here, we review and analyze various models (e.g., VGG, ResNet, DenseNet) fully, and we study how fine-tuning these models works for datasets such as ocular melanoma, Gaucher disease and neurofibromatosis. We rely on a broad method that uses data filtering, adapts the model to the clinical environment, measures the results using metrics and includes the Grad-CAM technique for interpreting what the model learns. It is clear from the findings that using transfer learning leads to better classification accuracy, sensitivity and specificity for every dataset. The highest accuracy, of 92.4%, was found in Gaucher disease classification using the modified DenseNet201 model with a hybrid loss function. Being able to show the problem areas in patient images with Grad-CAM made it easier for medical experts to trust the model’s performance. The research confirms that transfer learning is helpful for diagnosing rare diseases with the help of images and suggests it could improve early diagnosis and treatment
References
[1] Rajpurkar, P., Irvin, J., Zhu, K., Yang, B., Mehta, H., Duan, T., ... & Ng, A. Y. (2017). Chexnet: Radiologist-level pneumonia detection on chest x-rays with deep learning. arXiv preprint arXiv:1711.05225.
[2] Esteva, A., Kuprel, B., Novoa, R. A., Ko, J., Swetter, S. M., Blau, H. M., & Thrun, S. (2017). Dermatologist-level classification of skin cancer with deep neural networks. nature, 542(7639), 115-118.
[3] Gulshan, V., Peng, L., Coram, M., Stumpe, M. C., Wu, D., Narayanaswamy, A., ... & Webster, D. R. (2016). Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. jama, 316(22), 2402-2410.
[4] Litjens, G., Kooi, T., Bejnordi, B. E., Setio, A. A. A., Ciompi, F., Ghafoorian, M., ... & Sánchez, C. I. (2017). A survey on deep learning in medical image analysis. Medical image analysis, 42, 60-88.
[5] Shin, H. C., Roth, H. R., Gao, M., Lu, L., Xu, Z., Nogues, I., ... & Summers, R. M. (2016). Deep convolutional neural networks for computer-aided detection: CNN architectures, dataset characteristics and transfer learning. IEEE transactions on medical imaging, 35(5), 1285-1298.
[6] Tajbakhsh, N., Shin, J. Y., Gurudu, S. R., Hurst, R. T., Kendall, C. B., Gotway, M. B., & Liang, J. (2016). Convolutional neural networks for medical image analysis: Full training or fine tuning?. IEEE transactions on medical imaging, 35(5), 1299-1312.
[7] Yosinski, J., Clune, J., Bengio, Y., & Lipson, H. (2014). How transferable are features in deep neural networks?. Advances in neural information processing systems, 27.
[8] Raghu, M., Zhang, C., Kleinberg, J., & Bengio, S. (2019). Transfusion: Understanding transfer learning for medical imaging. Advances in neural information processing systems, 32.
[9] Chartsias, A., Joyce, T., Giuffrida, M. V., & Tsaftaris, S. A. (2017). Multimodal MR synthesis via modality-invariant latent representation. IEEE transactions on medical imaging, 37(3), 803-814.
[10] Albarqouni, S., Baur, C., Achilles, F., Belagiannis, V., Demirci, S., & Navab, N. (2016). Aggnet: deep learning from crowds for mitosis detection in breast cancer histology images. IEEE transactions on medical imaging, 35(5), 1313-1321.
[11] Antoniou, A., Storkey, A., & Edwards, H. (2017). Data augmentation generative adversarial networks. arXiv preprint arXiv:1711.04340.
[12] Wang, Y., Yao, Q., Kwok, J. T., & Ni, L. M. (2020). Generalizing from a few examples: A survey on few-shot learning. ACM computing surveys (csur), 53(3), 1-34.
[13] Selvaraju, R. R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., & Batra, D. (2017). Grad-cam: Visual explanations from deep networks via gradient-based localization. In Proceedings of the IEEE international conference on computer vision (pp. 618-626).
[14] Ribeiro, M. T., Singh, S., & Guestrin, C. (2016, August). " Why should i trust you?" Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining (pp. 1135-1144).
[15] Lundberg, S. M., & Lee, S. I. (2017). A unified approach to interpreting model predictions. Advances in neural information processing systems, 30.
[16] Khan, S., Noor, S., Awan, H.H. et al. “Deep-ProBind: binding protein prediction with transformer-based deep learning model”. BMC Bioinformatics 26, 88 (2025). https://doi.org/10.1186/s12859-025-06101-8.
[17] Aragani, Venu Madhav and Maroju, Praveen Kumar and Mudunuri, Lakshmi Narasimha Raju, Efficient Distributed Training through Gradient Compression with Sparsification and Quantization Techniques (September 29, 2021). Available at SSRN: https://ssrn.com/abstract=5022841 or http://dx.doi.org/10.2139/ssrn.5022841
[18] Swathi Chundru, Siva Subrahmanyam Balantrapu, Praveen Kumar Maroju, Naved Alam, Pushan Kumar Dutta, Pawan Whig, (2024/12/1), AGSQTL: adaptive green space quality transfer learning for urban environmental monitoring, 8th IET Smart Cities Symposium (SCS 2024), 2024, 551-556, IET.
[19] Islam Uddin, Salman A. AlQahtani, Sumaiya Noor, Salman Khan. “Deep-m6Am: a deep learning model for identifying N6, 2′-O-Dimethyladenosine (m6Am) sites using hybrid features[J]”. AIMS Bioengineering, 2025, 12(1): 145-161. doi: 10.3934/bioeng.2025006.
[20] S. Bama, P. K. Maroju, S. Banala, S. Kumar Sehrawat, M. Kommineni and D. Kodi, "Development of Web Platform for Home Screening of Neurological Disorders Using Artificial Intelligence," 2025 First International Conference on Advances in Computer Science, Electrical, Electronics, and Communication Technologies (CE2CT), Bhimtal, Nainital, India, 2025, pp. 995-999, doi: 10.1109/CE2CT64011.2025.10939414.
[21] Nirali Shah, "Validation and Verification of Artificial Intelligence Containing Products Across the Regulated Healthcare or Medical Device Industries", International Journal of Science and Research (IJSR), Volume 13 Issue 7, July 2024, pp. 66-71, https://www.ijsr.net/getabstract.php?paperid=ES24701081833, DOI: https://www.doi.org/10.21275/ES24701081833
[22] Kirti Vasdev. (2024). “Deep Learning for High-Resolution Geospatial Image Analysis”. International Journal of Innovative Research in Engineering & Multidisciplinary Physical Sciences, 12(6), 1–8. https://doi.org/10.5281/zenodo.14535586
[23] Khan, S., AlQahtani, S.A., Noor, S. et al. “PSSM-Sumo: deep learning based intelligent model for prediction of sumoylation sites using discriminative features”. BMC Bioinformatics 25, 284 (2024). https://doi.org/10.1186/s12859-024-05917-0.
[24] Deep Learning-Based Animal Intrusion Detection And Warning System For Railroad Tracks, Sree Lakshmi Vineetha Bitragunta, International Journal of Core Engineering & Management, Volume-6, Issue-11, 2021, PP-292-301.
