EUVIMEDEUVIMED entdecken
Crossref

Attention-Enhanced Deep Learning Framework for Automated Skin Disease Classification

Nasrin Sultana, Samia Tabassum Hrity, Shah Tanzeem Afsar, Shahriar Arefin Zummon

International Journal of Computer Applications · 20.08.2026 · 187 (137) · 1-6

DOI: 10.5120/ijcaedf9f1ac9e27 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Computer Applications, online erschienen am 20.08.2026. Verfasst von Nasrin Sultana und 3 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Skin disease is the most common health problem worldwide.It affects millions of people and often impacts both physical health and quality of life.Early detection of skin conditions is essential to prevent complications and improve treatment outcomes.The availability of Dermatologists is limited in many regions particularly in rural and underserved areas.As a result, patients often experience delays in diagnosis and treatment.In recent times, advances in artificial intelligence, especially deep learning have shown great potential in medical image analysis and automated illness detection.This research presents deep learning based framework for automated skin disease detection using dermoscopic images and incorporates the trained model into a mobile app application for practical use.The suggested system is designed to assists users in identifying possible skin conditions in an early stage that encourages timely medical consultation.This model is based in an EfficientNetV2-S convolutional neural network.Additionally Convolutional Block Attention Module (CBAM) and Generalized Mean (GeM) pooling are used.Six common skin disease categories were trained and evaluated on a dataset.A two-stage training strategy was adopted.Experimental results reveal that the suggested model could efficiently learns visual patterns that is associated with several dermatological disorders, achieving a final classification accuracy of 91.77% and a macro F1-score of 0.8548 across the evaluated disease categories.A mobile healthcare application was constructed for real-time uses.This enables users to shoot or upload skin photos and get real-time analysis provided by the trained model.

Zitieren

APA

Sultana, N., Hrity, S. T., Afsar, S. T., & Zummon, S. A. (2026). Attention-Enhanced Deep Learning Framework for Automated Skin Disease Classification. International Journal of Computer Applications, 187(137), 1-6. https://doi.org/10.5120/ijcaedf9f1ac9e27

Vancouver

Sultana N, Hrity ST, Afsar ST, Zummon SA. Attention-Enhanced Deep Learning Framework for Automated Skin Disease Classification. International Journal of Computer Applications. 2026;187(137):1-6. doi:10.5120/ijcaedf9f1ac9e27

BibTeX
@article{Sultana2026,
  author = {Sultana, Nasrin and Hrity, Samia Tabassum and Afsar, Shah Tanzeem and Zummon, Shahriar Arefin},
  title = {Attention-Enhanced Deep Learning Framework for Automated Skin Disease Classification},
  journal = {International Journal of Computer Applications},
  year = {2026},
  volume = {187},
  number = {137},
  pages = {1--6},
  doi = {10.5120/ijcaedf9f1ac9e27}
}
Kostenlos starten →