Dynamic Prototype Memory Network for Robust Multiclass Skin Lesion Diagnosis from Dermoscopic Images
International Journal of Medical Toxicology and Forensic Medicine · 06.10.2026
DOI: 10.22037/ijmtfm.v16.53420 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Medical Toxicology and Forensic Medicine, erschienen am 06.10.2026. Verfasst von Sharmila Varadan und 3 weiteren Autor:innen.
Abstract (Originalsprache)
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.
Schlagwörter
Zitieren
APA
Varadan, S., Shalini, S., Thirumaran, C. W., & Tupili, S. (2026). Dynamic Prototype Memory Network for Robust Multiclass Skin Lesion Diagnosis from Dermoscopic Images. International Journal of Medical Toxicology and Forensic Medicine. https://doi.org/10.22037/ijmtfm.v16.53420
Vancouver
Varadan S, Shalini S, Thirumaran CW, Tupili S. Dynamic Prototype Memory Network for Robust Multiclass Skin Lesion Diagnosis from Dermoscopic Images. International Journal of Medical Toxicology and Forensic Medicine. 2026. doi:10.22037/ijmtfm.v16.53420
BibTeX
@article{Varadan2026,
author = {Varadan, Sharmila and Shalini, Shanker and Thirumaran, Chembian Woothukadu and Tupili, Sangeetha},
title = {Dynamic Prototype Memory Network for Robust Multiclass Skin Lesion Diagnosis from Dermoscopic Images},
journal = {International Journal of Medical Toxicology and Forensic Medicine},
year = {2026},
doi = {10.22037/ijmtfm.v16.53420}
}
EUVIMED