Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Feb 19;15(1):6116.
doi: 10.1038/s41598-025-90418-0.

A skin disease classification model based on multi scale combined efficient channel attention module

Hui Liu #  1 , Yibo Dou #  2 , Kai Wang  3 , Yunmin Zou  4   5 , Gan Sen  6 , Xiangtao Liu  1 , Huling Li  1
Affiliations

A skin disease classification model based on multi scale combined efficient channel attention module

Hui Liu et al. Sci Rep. .

Abstract

Skin diseases, a significant category in the medical field, have always been challenging to diagnose and have a high misdiagnosis rate. Deep learning for skin disease classification has considerable value in clinical diagnosis and treatment. This study proposes a skin disease classification model based on multi-scale channel attention. The network architecture of the model consists of three main parts: an input module, four processing blocks, and an output module. Firstly, the model has improved the pyramid segmentation attention module to extract multi-scale features of the image entirely. Secondly, the reverse residual structure is used to replace the residual structure in the backbone network, and the attention module is integrated into the reverse residual structure to achieve better multi-scale feature extraction. Finally, the output module consists of an adaptive average pool and a fully connected layer, which convert the aggregated global features into several categories to generate the final output for the classification task. To verify the performance of the proposed model, this study used two commonly used skin disease datasets, ISIC2019 and HAM10000, for validation. The experimental results showed that the accuracy of this study was 77.6% on the ISIC2019 skin disease series dataset and 88.2% on the HAM10000 skin disease dataset. External validation data was added for evaluation to validate the model further, and the comprehensive evaluation results proved the effectiveness of the proposed model in this paper.

PubMed Disclaimer

Conflict of interest statement

Declarations. Competing interests: The authors declare no competing interests.

Figures

Fig. 1
Fig. 1
Network architecture diagram: (a) Representing the detailed structural diagram of Inverted Residuals Bottleneck, (b) represents the detailed structural diagram of Bottleneck.
Fig. 2
Fig. 2
Inverted residual bottleneck structure diagram: (a) Represents the original structure, (b) represents the improved detailed module.
Fig. 3
Fig. 3
Schematic diagram of multi-scale feature extraction,The left side of the figure represents segmented feature extraction, while the right side performs feature concatenation.
Fig. 4
Fig. 4
ECA overall structure diagram (1×1 convolutional layer is directly applied after the global average pooling layer to capture cross channel interaction).
Fig. 5
Fig. 5
Improved PSA overall structure diagram(replaces the SE module in the original PSA block with the ECA module).
Fig. 6
Fig. 6
Schematic diagram of improved multi-scale feature extraction(Replace the ordinary grouped convolution in the original PSA with depthwise separable convolution (DWConv)).
Fig. 7
Fig. 7
ISIC2019 dataset sample.
Fig. 8
Fig. 8
Image visualization with Grad-CAM technique.
Fig. 9
Fig. 9
ISIC2019 dataset sample.
Fig. 10
Fig. 10
ISIC2019 dataset sample.
Fig. 11
Fig. 11
HAM10000 dataset sample.
Fig. 12
Fig. 12
Comparison of models on the HAM10000 dataset.
Fig. 13
Fig. 13
Comparison of models on the real validate dataset.
See this image and copyright information in PMC

References

    1. Chen, M., Cui, D., Haick, H. & Tang, N. Artificial intelligence-based medical sensors for healthcare system. Adv. Sensor Res.3, 2300009 (2024).
    1. Groh, M. et al. Deep learning-aided decision support for diagnosis of skin disease across skin tones. Nat. Med.30, 573–583 (2024). - PMC - PubMed
    1. Rauf, F. et al. Artificial intelligence assisted common maternal fetal planes prediction from ultrasound images based on information fusion of customized convolutional neural networks. Front. Med.11, 1486995 (2024). - PMC - PubMed
    1. Yadav, R. & Bhat, A. A systematic literature survey on skin disease detection and classification using machine learning and deep learning. Multimed. Tools Appl.83, 1–32 (2024).
    1. Saleh, N., Hassan, M. A. & Salaheldin, A. M. Skin cancer classification based on an optimized convolutional neural network and multicriteria decision-making. Sci. Rep.14, 17323 (2024). - PMC - PubMed

LinkOut - more resources