Research on Dermoscopic Image Classification Method Integrating Multi-Scale Features and Attention Mechanism

Yuchen Jiang; Xuewen Ding

doi:10.54097/vqk89w83

Authors

Yuchen Jiang School of Electronic Engineering Tianjin University of Technology and Education Tianjin, China
Xuewen Ding School of Electronic Engineering Tianjin University of Technology and Education Tianjin, China

DOI:

https://doi.org/10.54097/vqk89w83

Keywords:

Dermoscopic Images; Skin Lesion Classification; EfficientNetV2; Multi-scale Feature Fusion; Attention Mechanism; Deep Learning.

Abstract

To address the issues of lesion scale differences, easy loss of fine-grained pigment information, and background interference in dermoscopic images, this paper constructs a multi-scale attention fusion classification network MSAF-EfficientNetV2 based on EfficientNetV2-S. The network extracts features in four stages. AMSF achieves dynamic fusion through channel mapping, spatial alignment, and sample-related scale weights. LAA strengthens the main body of the lesion and irregular boundaries with channel-spatial attention, and uses weighted cross-entropy to alleviate the long-tailed distribution of the seven classes. The experimental results used publicly available data to form two levels of evidence: In the MedMNIST+ 224×224 end-to-end benchmark, the accuracy of DenseNet121 was 84.74±0.51%, and the AUC of DINO ViT-B/16 was 96.50±0.51%; in the histopathologically confirmed true melanoma ISIC_0000013, the Otsu dark region accounted for 22.29%, and the darkest cluster in Lab-K-means accounted for 16.60%, with a Dice concordance of 85.38%. The publicly available ISIC 2017 segmentation experiment further showed that the Dice/Jaccard ratio of VGG16-U-Net was 91.5%/84.6%, and the Dice/Jaccard ratio in this case reached 96.2%/92.6%. The total number of network parameters is 20.069 M and the number of FLOPs is 5.775 G, indicating that the newly added fusion and attention structures maintain controllable computational overhead.

Downloads

Download data is not yet available.

References

[1] Shetty, B., Fernandes, R., Rodrigues, A. P., Chengoden, R., Bhattacharya, S., & Lakshmanna, K. (2022). Skin lesion classification of dermoscopic images using machine learning and convolutional neural network. Scientific Reports, 12, 18134. https://doi.org/10.1038/s41598-022-22644-9

[2] Yao, P., Shen, S., Xu, M., Liu, P., Zhang, F., Xing, J., Shao, P., Kaffenberger, B., & Xu, R. X. (2022). Single model deep learning on imbalanced small datasets for skin lesion classification. IEEE Transactions on Medical Imaging, 41(5), 1242–1254. https://doi.org/10.1109/TMI.2021.3136682

[3] Pérez, E., & Ventura, S. (2023). A framework to build accurate convolutional neural network models for melanoma diagnosis. Knowledge Based Systems, 260, 110157. https://doi.org/10.1016/j.knosys.2022.110157

[4] Qian, S., Ren, K., Zhang, W., & Ning, H. (2022). Skin lesion classification using CNNs with grouping of multi scale attention and class specific loss weighting. Computer Methods and Programs in Biomedicine, 226, 107166. https://doi.org/10.1016/j.cmpb.2022.107166

[5] Wang, L., Zhang, L., Shu, X., & Yi, Z. (2023). Intra class consistency and inter class discrimination feature learning for automatic skin lesion classification. Medical Image Analysis, 85, 102746. https://doi.org/10.1016/j.media.2023.102746

[6] Yue, G., Wei, P., Zhou, T., Jiang, Q., Yan, W., & Wang, T. (2023). Toward multicenter skin lesion classification using deep neural network with adaptively weighted balance loss. IEEE Transactions on Medical Imaging, 42(1), 119–131. https://doi.org/10.1109/TMI.2022.3204646

[7] Ajmal, M., Khan, M. A., Akram, T., Alqahtani, A., Alhaisoni, M., Armghan, A., Althubiti, S. A., & Alenezi, F. (2023). BF²SkNet: Best deep learning features fusion assisted framework for multiclass skin lesion classification. Neural Computing and Applications, 35(30), 22115–22131. https://doi.org/10.1007/s00521-022-08084-6

[8] Hu, Z., Mei, W., Chen, H., & Hou, W. (2024). Multi scale feature fusion and class weight loss for skin lesion classification. Computers in Biology and Medicine, 176, 108594. https://doi.org/10.1016/j.compbiomed.2024.108594

[9] Hosny, K. M., Said, W., Elmezain, M., & Kassem, M. A. (2024). Explainable deep inherent learning for multi classes skin lesion classification. Applied Soft Computing, 159, 111624. https://doi.org/10.1016/j.asoc.2024.111624

[10] Yadav, D. P., Sharma, B., Chauhan, S., Webber, J. L., & Mehbodniya, A. (2024). Dual scale light weight cross attention transformer for skin lesion classification. PLOS ONE, 19(12), e0312598. https://doi.org/10.1371/journal.pone.0312598

[11] Yang, J., Shi, R., Wei, D., Liu, Z., Zhao, L., Ke, B., Pfister, H., & Ni, B. (2023). MedMNIST v2 A large scale lightweight benchmark for 2D and 3D biomedical image classification. Scientific Data, 10, 41. https://doi.org/10.1038/s41597-022-01721-8

[12] Afza, F., Sharif, M., Khan, M. A., Tariq, U., Yong, H. S., & Cha, J. (2022). Multiclass skin lesion classification using hybrid deep features selection and extreme learning machine. Sensors, 22(3), 799. https://doi.org/10.3390/s22030799

[13] Abhishek, K., Jain, A., & Hamarneh, G. (2025). Investigating the quality of DermaMNIST and Fitzpatrick17k dermatological image datasets. Scientific Data, 12, 196. https://doi.org/10.1038/s41597-025-04382-5

[14] Doerrich, S., Di Salvo, F., Brockmann, J., & Ledig, C. (2025). Rethinking model prototyping through the MedMNIST+ dataset collection. Scientific Reports, 15, 7669. https://doi.org/10.1038/s41598-025-92156-9

[15] Tschandl, P., Rosendahl, C., & Kittler, H. (2018). The HAM10000 dataset, a large collection of multi source dermatoscopic images of common pigmented skin lesions. Scientific Data, 5, 180161. https://doi.org/10.1038/sdata.2018.161

[16] International Skin Imaging Collaboration. (n.d.). ISIC Archive image ISIC_0000013: histopathology confirmed invasive melanoma, dermoscopic image, CC0. Retrieved August 18, 2026, from https://www.isic archive.org

[17] Thanh, D. N. H., Hai, N. H., Hieu, L. M., Tiwari, P., & Prasath, V. B. S. (2021). Skin lesion segmentation method for dermoscopic images with convolutional neural networks and semantic segmentation. Computer Optics, 45(1), 122–129. https://doi.org/10.18287/2412-6179-CO-748.