EUVIMEDEUVIMED entdecken
Crossref

Attention-Guided Learning With Feature Reconstruction for Skin Lesion Diagnosis Using Clinical and Ultrasound Images

Chunlun Xiao, Anqi Zhu, Chunmei Xia, Zifeng Qiu, Yuanlin Liu, Cheng Zhao, Weiwei Ren, Lifan Wang, Lei Dong, Tianfu Wang, Lehang Guo, Baiying Lei

IEEE Transactions on Medical Imaging · 2025-01 · 44 (1) · 543-555

DOI: 10.1109/tmi.2024.3450682 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Medical Imaging, 2025-01. Verfasst von Chunlun Xiao und 11 weiteren Autor:innen. Hauptthemen laut MEDLINE: Deep Learning, Image Interpretation, Computer-Assisted, Skin Diseases. Die Arbeit wurde bisher 17-mal zitiert (Crossref) und nennt 73 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Shenzhen Science and Technology Program, Science and Technology Commission of Shanghai Municipality.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Skin lesion is one of the most common diseases, and most categories are highly similar in morphology and appearance. Deep learning models effectively reduce the variability between classes and within classes, and improve diagnostic accuracy. However, the existing multi-modal methods are only limited to the surface information of lesions in skin clinical and dermatoscopic modalities, which hinders the further improvement of skin lesion diagnostic accuracy. This requires us to further study the depth information of lesions in skin ultrasound. In this paper, we propose a novel skin lesion diagnosis network, which combines clinical and ultrasound modalities to fuse the surface and depth information of the lesion to improve diagnostic accuracy. Specifically, we propose an attention-guided learning (AL) module that fuses clinical and ultrasound modalities from both local and global perspectives to enhance feature representation. The AL module consists of two parts, attention-guided local learning (ALL) computes the intra-modality and inter-modality correlations to fuse multi-scale information, which makes the network focus on the local information of each modality, and attention-guided global learning (AGL) fuses global information to further enhance the feature representation. In addition, we propose a feature reconstruction learning (FRL) strategy which encourages the network to extract more discriminative features and corrects the focus of the network to enhance the model's robustness and certainty. We conduct extensive experiments and the results confirm the superiority of our proposed method. Our code is available at: https://github.com/XCL-hub/AGFnet.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Deep LearningImage Interpretation, Computer-AssistedSkin DiseasesHumansUltrasonographySkinSkin NeoplasmsAlgorithmsDatabases, Factual

Förderung

  • National Natural Science Foundation of China
  • Shenzhen Science and Technology Program
  • Science and Technology Commission of Shanghai Municipality
  • Shenzhen Medical Research Funds
  • National Natural Science Foundation of Guangdong Province
  • China Postdoctoral Science Foundation

Zitieren

APA

Xiao, C., Zhu, A., Xia, C., Qiu, Z., Liu, Y., Zhao, C., Ren, W., Wang, L., Dong, L., Wang, T., Guo, L., & Lei, B. (2025). Attention-Guided Learning With Feature Reconstruction for Skin Lesion Diagnosis Using Clinical and Ultrasound Images. IEEE Transactions on Medical Imaging, 44(1), 543-555. https://doi.org/10.1109/tmi.2024.3450682

Vancouver

Xiao C, Zhu A, Xia C, Qiu Z, Liu Y, Zhao C, et al. Attention-Guided Learning With Feature Reconstruction for Skin Lesion Diagnosis Using Clinical and Ultrasound Images. IEEE Transactions on Medical Imaging. 2025;44(1):543-555. doi:10.1109/tmi.2024.3450682

BibTeX
@article{Xiao2025,
  author = {Xiao, Chunlun and Zhu, Anqi and Xia, Chunmei and Qiu, Zifeng and Liu, Yuanlin and Zhao, Cheng and Ren, Weiwei and Wang, Lifan and Dong, Lei and Wang, Tianfu and Guo, Lehang and Lei, Baiying},
  title = {Attention-Guided Learning With Feature Reconstruction for Skin Lesion Diagnosis Using Clinical and Ultrasound Images},
  journal = {IEEE Transactions on Medical Imaging},
  year = {2025},
  volume = {44},
  number = {1},
  pages = {543--555},
  doi = {10.1109/tmi.2024.3450682}
}
Kostenlos starten →