A novel approach to skin disease segmentation using a visual selective state spatial model with integrated spatial constraints
Abstract Accurate segmentation of skin lesions is crucial for reliable clinical diagnosis and effective treatment planning. Automated techniques for skin lesion segmentation assist dermatologists in early detection and ongoing monitoring of various skin diseases, ultimately improving patient outcome...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-02-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-85301-x |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823862698374332416 |
|---|---|
| author | Yu Bai Hai Zhou Hongjie Zhu Shimin Wen Binbin Hu Haotian Li Huazhang Wang Daji Ergu Fangyao Liu |
| author_facet | Yu Bai Hai Zhou Hongjie Zhu Shimin Wen Binbin Hu Haotian Li Huazhang Wang Daji Ergu Fangyao Liu |
| author_sort | Yu Bai |
| collection | DOAJ |
| description | Abstract Accurate segmentation of skin lesions is crucial for reliable clinical diagnosis and effective treatment planning. Automated techniques for skin lesion segmentation assist dermatologists in early detection and ongoing monitoring of various skin diseases, ultimately improving patient outcomes and reducing healthcare costs. To address limitations in existing approaches, we introduce a novel U-shaped segmentation architecture based on our Residual Space State Block. This efficient model, termed ‘SSR-UNet,’ leverages bidirectional scanning to capture both global and local features in image data, achieving strong performance with low computational complexity. Traditional CNNs struggle with long-range dependencies, while Transformers, though excellent at global feature extraction, are computationally intensive and require large amounts of data. Our SSR-UNet model overcomes these challenges by efficiently balancing computational load and feature extraction capabilities. Additionally, we introduce a spatially-constrained loss function that mitigates gradient stability issues by considering the distance between label and prediction boundaries. We rigorously evaluated SSR-UNet on the ISIC2017 and ISIC2018 skin lesion segmentation benchmarks. The results showed that the accuracy of Mean Intersection Over Union, Classification Accuracy and Specificity indexes in ISIC2017 datasets reached 80.98, 96.50 and 98.04, respectively, exceeding the best indexes of other models by 0.83, 0.99 and 0.38, respectively. The accuracy of Mean Intersection Over Union, Dice Coefficient, Classification Accuracy and Sensitivity on ISIC2018 datasets reached 82.17, 90.21, 95.34 and 88.49, respectively, exceeding the best indicators of other models by 1.71, 0.27, 0.65 and 0.04, respectively. It can be seen that SSR-UNet model has excellent performance in most aspects. |
| format | Article |
| id | doaj-art-e75b85c8537a4d989e60cc16b937fe82 |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-02-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-e75b85c8537a4d989e60cc16b937fe822025-02-09T12:28:37ZengNature PortfolioScientific Reports2045-23222025-02-0115111610.1038/s41598-025-85301-xA novel approach to skin disease segmentation using a visual selective state spatial model with integrated spatial constraintsYu Bai0Hai Zhou1Hongjie Zhu2Shimin Wen3Binbin Hu4Haotian Li5Huazhang Wang6Daji Ergu7Fangyao Liu8College of Electronic and Information, Southwest Minzu UniversityCollege of Electronic and Information, Southwest Minzu UniversityCollege of Electronic and Information, Southwest Minzu UniversityCollege of Electronic and Information, Southwest Minzu UniversityCollege of Electronic and Information, Southwest Minzu UniversitySchool of Biological Sciences, University of Nebraska-LincolnKey Laboratory of Electronic Information Engineering, Southwest Minzu UniversityCollege of Electronic and Information, Southwest Minzu UniversityCollege of Electronic and Information, Southwest Minzu UniversityAbstract Accurate segmentation of skin lesions is crucial for reliable clinical diagnosis and effective treatment planning. Automated techniques for skin lesion segmentation assist dermatologists in early detection and ongoing monitoring of various skin diseases, ultimately improving patient outcomes and reducing healthcare costs. To address limitations in existing approaches, we introduce a novel U-shaped segmentation architecture based on our Residual Space State Block. This efficient model, termed ‘SSR-UNet,’ leverages bidirectional scanning to capture both global and local features in image data, achieving strong performance with low computational complexity. Traditional CNNs struggle with long-range dependencies, while Transformers, though excellent at global feature extraction, are computationally intensive and require large amounts of data. Our SSR-UNet model overcomes these challenges by efficiently balancing computational load and feature extraction capabilities. Additionally, we introduce a spatially-constrained loss function that mitigates gradient stability issues by considering the distance between label and prediction boundaries. We rigorously evaluated SSR-UNet on the ISIC2017 and ISIC2018 skin lesion segmentation benchmarks. The results showed that the accuracy of Mean Intersection Over Union, Classification Accuracy and Specificity indexes in ISIC2017 datasets reached 80.98, 96.50 and 98.04, respectively, exceeding the best indexes of other models by 0.83, 0.99 and 0.38, respectively. The accuracy of Mean Intersection Over Union, Dice Coefficient, Classification Accuracy and Sensitivity on ISIC2018 datasets reached 82.17, 90.21, 95.34 and 88.49, respectively, exceeding the best indicators of other models by 1.71, 0.27, 0.65 and 0.04, respectively. It can be seen that SSR-UNet model has excellent performance in most aspects.https://doi.org/10.1038/s41598-025-85301-xDeep learningState spatial residual modelComputer visionSkin lesion segmentationModel parameter updating and optimization |
| spellingShingle | Yu Bai Hai Zhou Hongjie Zhu Shimin Wen Binbin Hu Haotian Li Huazhang Wang Daji Ergu Fangyao Liu A novel approach to skin disease segmentation using a visual selective state spatial model with integrated spatial constraints Scientific Reports Deep learning State spatial residual model Computer vision Skin lesion segmentation Model parameter updating and optimization |
| title | A novel approach to skin disease segmentation using a visual selective state spatial model with integrated spatial constraints |
| title_full | A novel approach to skin disease segmentation using a visual selective state spatial model with integrated spatial constraints |
| title_fullStr | A novel approach to skin disease segmentation using a visual selective state spatial model with integrated spatial constraints |
| title_full_unstemmed | A novel approach to skin disease segmentation using a visual selective state spatial model with integrated spatial constraints |
| title_short | A novel approach to skin disease segmentation using a visual selective state spatial model with integrated spatial constraints |
| title_sort | novel approach to skin disease segmentation using a visual selective state spatial model with integrated spatial constraints |
| topic | Deep learning State spatial residual model Computer vision Skin lesion segmentation Model parameter updating and optimization |
| url | https://doi.org/10.1038/s41598-025-85301-x |
| work_keys_str_mv | AT yubai anovelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT haizhou anovelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT hongjiezhu anovelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT shiminwen anovelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT binbinhu anovelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT haotianli anovelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT huazhangwang anovelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT dajiergu anovelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT fangyaoliu anovelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT yubai novelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT haizhou novelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT hongjiezhu novelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT shiminwen novelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT binbinhu novelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT haotianli novelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT huazhangwang novelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT dajiergu novelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints AT fangyaoliu novelapproachtoskindiseasesegmentationusingavisualselectivestatespatialmodelwithintegratedspatialconstraints |