EUVIMEDDescubrir EUVIMED
Open access · CC BY 4.0

A Dual-Branch Multi-Scale Network for Skin Lesion Classification

Ying Liu, Xinyu Feng, Yuchai Wan, Huifu Li, Xun Zhang, Abdureyim Raxidin

Electronics · 2026-03-08 · 15 (5) · 1118

DOI: 10.3390/electronics15051118 ↗

At a glance

Journal article in Electronics, published online on 2026-03-08. Written by Ying Liu and 5 co-authors. 38 references. The full text is freely available (open access, CC BY 4.0). Funded by: Beijing Natural Science Foundation, The Project of Cultivation for Young Top-notch Talents of Beijing Municipal Institution, Project of Social Science Foundation of Xinjiang Uygur Autonomous Region.

ResumenDatos bibliográficosCiteRelated studies

Abstract (original language)

Dermoscopic images are widely used for diagnosing skin diseases, and automatic classification of lesion types using deep learning can significantly enhance diagnostic efficiency. However, challenges such as variations in imaging conditions, subtle differences between classes, high variability within classes, and severe class imbalance complicate skin lesion analysis. This paper introduces a dual-branch deep learning model where two branches independently process high-frequency and low-frequency image features to generate multi-scale fused representations. To address class imbalance, the model employs cosine similarity to strengthen inter-class discrimination and incorporates a bias term to improve recognition of minority lesion classes. Experiments conducted on the ISIC 2017 and ISIC 2018 datasets demonstrate that the proposed method surpasses state-of-the-art approaches, achieving accuracies of 97.0% and 91.9%, respectively, with sensitivity and specificity both exceeding 90% on the two datasets.

Funding

  • Beijing Natural Science Foundation
  • The Project of Cultivation for Young Top-notch Talents of Beijing Municipal Institution
  • Project of Social Science Foundation of Xinjiang Uygur Autonomous Region
  • Project of Natural Science Foundation of Xinjiang Uygur Autonomous Region

Cite

APA

Liu, Y., Feng, X., Wan, Y., Li, H., Zhang, X., & Raxidin, A. (2026). A Dual-Branch Multi-Scale Network for Skin Lesion Classification. Electronics, 15(5), 1118. https://doi.org/10.3390/electronics15051118

Vancouver

Liu Y, Feng X, Wan Y, Li H, Zhang X, Raxidin A. A Dual-Branch Multi-Scale Network for Skin Lesion Classification. Electronics. 2026;15(5):1118. doi:10.3390/electronics15051118

BibTeX
@article{Liu2026,
  author = {Liu, Ying and Feng, Xinyu and Wan, Yuchai and Li, Huifu and Zhang, Xun and Raxidin, Abdureyim},
  title = {A Dual-Branch Multi-Scale Network for Skin Lesion Classification},
  journal = {Electronics},
  year = {2026},
  volume = {15},
  number = {5},
  pages = {1118},
  doi = {10.3390/electronics15051118}
}
Start for free →