Dynamic Prototype Memory Network for Robust Multiclass Skin Lesion Diagnosis from Dermoscopic Images
International Journal of Medical Toxicology and Forensic Medicine,
Vol. 16 (2026),
1 January 2026
,
Page 1-9
https://doi.org/10.22037/ijmtfm.v16.53420
Abstract
Background: Automated skin lesion diagnosis from dermoscopic images is challenging because lesions of the same disease can vary substantially visually, while visually similar patterns can occur across different disease categories. This article introduces a Dynamic Prototype Memory Network (DPM-Net) for multiclass skin lesion classification.
Methods: We first employ a deep visual encoder to extract discriminative feature representations from dermoscopic images. We then compare the extracted lesion embeddings with disease-specific prototypes using a similarity-based prototype assignment mechanism. To enhance the method's representational capability, we dynamically update the prototypes based on the feature distribution of the lesion samples. We incorporate a prototype diversity mechanism to prevent prototype collapse, while a prototype compactness constraint encourages lesion embeddings to remain close to the representative prototypes of their corresponding disease categories.
Results: Performance validation shows improved results, with a maximum accuracy of 95.39%.
Conclusion: Therefore, the proposed model integrates different components to perform automated multiclass skin lesion diagnosis from dermoscopic images.
- Dynamic Prototype Memory
- Skin Lesion Classification
- Dermoscopic Images
- HAM10000
- Prototype Learning
How to Cite
References
[1] Singh ID, Bansal S, Datta P. Automated Multi-Class Skin Disease Diagnosis Using Deep Convolutional Neural Networks. In: 2026 Fourth International Conference on Secure Cyber Computing and Communications (ICSCCC). IEEE; 2026. p. 1308-13. [DOI: 10.1109/ICSCCC69031.2026.11600163]
[2] Ravisankar S, Karthikeyan M. Deep learning-based multi-class classification of cutaneous lesions for dermatological assessment. Cutan Ocul Toxicol. 2026:1-19. [DOI: 10.1080/15569527.2026.2702002]
[3] Christy SJ, Kumari GRN. An Optimized Deep Learning Framework for Robust Multi-Class Skin Cancer Detection from Dermoscopic Images. In: 2026 13th International Conference on Computing for Sustainable Global Development (INDIACom). IEEE; 2026. p. 1-5. [DOI: 10.23919/INDIACom70271.2026.11525716]
[4] Nikhitha K, Mayee MK, Bhavya G, Chandana MM, Subhashini T, Sailaja M. Multi-Class Skin Diseases Detection Using Dermoscopic Images. In: 2026 Fourth International Conference on Augmented Intelligence and Sustainable Systems (ICAISS). IEEE; 2026. p. 882-7. [DOI: 10.1109/ICAISS68683.2026.11526239]
[5] Saju JK, Thakur D, Sharma P, Pandey A, Tajammul M. A Deep Learning Framework for Automated Multi-Class Skin Disease Detection Using Dermoscopic Images. In: 2026 5th International Conference on Innovative Practices in Technology and Management (ICIPTM). IEEE; 2026. p. 1-8. [DOI: 10.1109/ICIPTM69057.2026.11465623]
[6] Gabani V, Navamani TM, Shyamala K, Vaswani Rajpal VK. Multimodal skin lesion classification for early cancer diagnosis using deep learning. Front Physiol. 2026;17:1717517. [DOI: 10.3389/fphys.2026.1717517]
[7] Chand HV, Sabharwal S, Jiang W, Islam SM, Cengiz K. A stacking-based deep learning ensemble for multi-class skin lesion classification. Discov Comput. 2026;29(1):417. [DOI: 10.1007/s10791-026-10288-6]
[8] Singh MK, Singh AK, Choudhary P, Singh AK, Tyagi LK. A practical deep learning framework for multi-class skin cancer detection using U-Net segmentation and ResNet50 transfer learning. J Niger Soc Phys Sci. 2026:3530. [DOI: 10.46481/jnsps.2026.3530]
[9] Fiaz M, Shoaib Khan MB, Khan AH, Bilal A, Abdullah M, Darem AA, et al. An explainable hybrid deep learning framework for precise skin lesion segmentation and multi-class classification. Front Med. 2025;12:1681542. [DOI: 10.3389/fmed.2025.1681542]
[10] Arshad M, Khan MA, Almujally NA, Alasiry A, Marzougui M, Nam Y. Multiclass skin lesion classification and localziation from dermoscopic images using a novel network-level fused deep architecture and explainable artificial intelligence. BMC Med Inform Decis Mak. 2025;25(1):215. [DOI: 10.1186/s12911-025-03051-2]
[11] Owida HA, El-Fattah IA, Abuowaida S, Alshdaifat N, Mashagba HA, Abd Aziz AB, et al. A deep learning-based dual-branch framework for automated skin lesion segmentation and classification via dermoscopic Images. Sci Rep. 2025;15(1):37823. [DOI: 10.1038/s41598-025-21783-z]
[12] Al-Waisy AS, Al-Fahdawi S, Khalaf MI, Mohammed MA, Al-Attar B, Al-Andoli MN. A deep learning framework for automated early diagnosis and classification of skin cancer lesions in dermoscopy images. Sci Rep. 2025;15(1):31234. [DOI: 10.1038/s41598-025-15655-9]
[13] Skin Cancer MNIST: HAM10000 Dataset. Kaggle. Accessed August 1, 2026. [Link]
[14] ISIC 2019 Skin Lesion images for classification dataset. Kaggle. Accessed August 1, 2026. [Link]
[15] Liu Z, Mao H, Wu CY, Feichtenhofer C, Darrell T, Xie S. A convnet for the 2020s. In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE; 2022. p. 11966-76. [DOI: 10.1109/CVPR52688.2022.01167]
[16] Snell J, Swersky K, Zemel R. Prototypical networks for few-shot learning. Adv Neural Inf Process Syst. 2017;30.
[17] Li X, Ouyang J, Chen Y, Diao Z, Li K, He D, et al. Accurate skin lesion classification on imbalanced dermoscopic images with high variance via the SCTFD framework. Sci Rep. 2026;16(1):12302. [DOI: 10.1038/s41598-026-37846-8]
[18] Sarhan AM, Ali HA, Yasser S, Gobara M, Kandil AA, Sherif G, et al. Achieving high-accuracy skin cancer classification with deep learning optimized by ant colony algorithm. J Electr Syst Inf Technol. 2025;12(1):49. [DOI: 10.1186/s43067-025-00243-8]
- Abstract Viewed: 17 times
- 53420 Downloaded: 8 times