TY  - GEN
AB  - Accurate wound segmentation is crucial for the precise diagnosis and treatment of various skin conditions through image analysis. In this paper, we introduce a novel dual attention U-Net model designed for precise wound segmentation. Our proposed architecture integrates two widely used deep learning models, VGG16 and U-Net, incorporating dual attention mechanisms to focus on relevant regions within the wound area. Initially trained on diabetic foot ulcer images, we fine-tuned the model to acute and chronic wound images and conducted a comprehensive comparison with other state-of-the-art models. The results highlight the superior performance of our proposed dual attention model, achieving a Dice coefficient and IoU of 94.1% and 89.3%, respectively, on the test set. This underscores the robustness of our method and its capacity to generalize effectively to new data.
AD  - Computer Science Department, University of Geneva, Geneva, Switzerland
AD  - imito AG, Zurich, Switzerland
AD  - Geneva School of Health Sciences, HES-SO University of Applied Sciences and Arts Western Switzerland
AD  - Computer Science Department, University of Geneva, Geneva, Switzerland
AD  - Geneva School of Health Sciences, HES-SO University of Applied Sciences and Arts Western Switzerland ; Care Directorate, Geneva University Hospitals, Geneva, Switzerland
AD  - Geneva School of Health Sciences, HES-SO University of Applied Sciences and Arts Western Switzerland ; Institute of Industrial and IT Engineering, HEPIA, HES-SO Geneva University of Applied Sciences and Arts, Western Switzerland, Geneva, Switzerland
AD  - Institute of Industrial and IT Engineering, HES-SO Geneva University of Applied Sciences and Arts, Western Switzerland, Geneva (HEPIA), Switzerland ; Computer Science Department, University of Geneva, Geneva, Switzerland
AU  - Niri, Rania
AU  - Zahia, Sofia
AU  - Stefanelli, Alessio
AU  - Sharma, Kaushal
AU  - Probst, Sebastian
AU  - Pichon, Swann
AU  - Chanel, Guillaume
CY  - États-Unis
DA  - 2025-01
DO  - 10.1007/s10278-025-01386-w
DO  - DOI
ID  - 15453
JF  - Journal of imaging informatics in medicine
KW  - U-Net
KW  - attention networks
KW  - wound segmentation
KW  - medical imaging
KW  - deep learning
L1  - https://arodes.hes-so.ch/record/15453/files/Probst_2025_wound_segmentation_transfert_learning.pdf
L2  - https://arodes.hes-so.ch/record/15453/files/Probst_2025_wound_segmentation_transfert_learning.pdf
L4  - https://arodes.hes-so.ch/record/15453/files/Probst_2025_wound_segmentation_transfert_learning.pdf
LA  - eng
LK  - https://arodes.hes-so.ch/record/15453/files/Probst_2025_wound_segmentation_transfert_learning.pdf
N2  - Accurate wound segmentation is crucial for the precise diagnosis and treatment of various skin conditions through image analysis. In this paper, we introduce a novel dual attention U-Net model designed for precise wound segmentation. Our proposed architecture integrates two widely used deep learning models, VGG16 and U-Net, incorporating dual attention mechanisms to focus on relevant regions within the wound area. Initially trained on diabetic foot ulcer images, we fine-tuned the model to acute and chronic wound images and conducted a comprehensive comparison with other state-of-the-art models. The results highlight the superior performance of our proposed dual attention model, achieving a Dice coefficient and IoU of 94.1% and 89.3%, respectively, on the test set. This underscores the robustness of our method and its capacity to generalize effectively to new data.
PB  - Springer
PP  - États-Unis
PY  - 2025-01
SN  - 2948-2925
T1  - Wound segmentation with U-Net using a dual attention mechanism and transfer learning
TI  - Wound segmentation with U-Net using a dual attention mechanism and transfer learning
UR  - https://arodes.hes-so.ch/record/15453/files/Probst_2025_wound_segmentation_transfert_learning.pdf
VL  - January 2025, to be published
Y1  - 2025-01
ER  -