DPA-HairNet: A Dual Encoder Attention Based Network for Hair Artifact Removal in Dermoscopic Images
Hair artifacts in dermoscopic images significantly hinder the accurate diagnosis of melanoma and other skin conditions by obscuring critical lesion details. To address this challenge, we introduce DPA-HairNet, a novel Dual Encoder Attention-Based Network designed specifically for effective hair arti...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11062859/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849714387981434880 |
|---|---|
| author | F M Javed Mehedi Shamrat Mohd Yamani Idna Idris Chowdhury Forhadul Karim Xujuan Zhou Raj Gururajan |
| author_facet | F M Javed Mehedi Shamrat Mohd Yamani Idna Idris Chowdhury Forhadul Karim Xujuan Zhou Raj Gururajan |
| author_sort | F M Javed Mehedi Shamrat |
| collection | DOAJ |
| description | Hair artifacts in dermoscopic images significantly hinder the accurate diagnosis of melanoma and other skin conditions by obscuring critical lesion details. To address this challenge, we introduce DPA-HairNet, a novel Dual Encoder Attention-Based Network designed specifically for effective hair artifact removal while preserving lesion integrity. The model features a dual encoder architecture to separately process hair-specific and lesion-specific features, an attention mechanism to prioritize diagnostically relevant regions, and a dual-output design that generates precise hair segmentation masks and high-quality reconstructed images. This study leverages the ISIC 2018 and HAM10000 datasets, incorporating advanced data augmentation techniques to enhance dataset diversity and ensure robust model training. DPA-HairNet was evaluated against six state-of-the-art segmentation models using comprehensive metrics, including Accuracy, Dice Coefficient, Jaccard Index, PSNR, MAE, Specificity, Precision, Recall, and F1-Score, where the proposed model outperformed all models. Furthermore, classification performance was assessed before and after hair artifact removal using twelve pre-trained classifiers, demonstrating significant improvements in diagnostic accuracy. Explainable AI techniques, Grad-CAM, UMAP and attention heatmaps, were employed to interpret and validate the model’s focus. These results underscore DPA-HairNet’s effectiveness and potential integration into clinical workflows to enhance dermoscopic image analysis and diagnosis. Future work will explore generalization to additional artifact types and optimization for real-time clinical deployment. |
| format | Article |
| id | doaj-art-ceae0295977d4df3814f2efcbd377eb6 |
| institution | DOAJ |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-ceae0295977d4df3814f2efcbd377eb62025-08-20T03:13:43ZengIEEEIEEE Access2169-35362025-01-011311818511821110.1109/ACCESS.2025.358535311062859DPA-HairNet: A Dual Encoder Attention Based Network for Hair Artifact Removal in Dermoscopic ImagesF M Javed Mehedi Shamrat0https://orcid.org/0000-0001-9176-3537Mohd Yamani Idna Idris1https://orcid.org/0000-0003-4894-0838Chowdhury Forhadul Karim2https://orcid.org/0000-0002-4663-0353Xujuan Zhou3Raj Gururajan4Department of Computer System and Technology, Universiti Malaya, Kuala Lumpur, MalaysiaDepartment of Computer System and Technology, Universiti Malaya, Kuala Lumpur, MalaysiaDepartment of Physiology, Faculty of Medicine, Universiti Malaya, Kuala Lumpur, MalaysiaSchool of Business, University of Southern Queensland, Springfield Campus, Toowoomba, QLD, AustraliaSchool of Business, University of Southern Queensland, Springfield Campus, Toowoomba, QLD, AustraliaHair artifacts in dermoscopic images significantly hinder the accurate diagnosis of melanoma and other skin conditions by obscuring critical lesion details. To address this challenge, we introduce DPA-HairNet, a novel Dual Encoder Attention-Based Network designed specifically for effective hair artifact removal while preserving lesion integrity. The model features a dual encoder architecture to separately process hair-specific and lesion-specific features, an attention mechanism to prioritize diagnostically relevant regions, and a dual-output design that generates precise hair segmentation masks and high-quality reconstructed images. This study leverages the ISIC 2018 and HAM10000 datasets, incorporating advanced data augmentation techniques to enhance dataset diversity and ensure robust model training. DPA-HairNet was evaluated against six state-of-the-art segmentation models using comprehensive metrics, including Accuracy, Dice Coefficient, Jaccard Index, PSNR, MAE, Specificity, Precision, Recall, and F1-Score, where the proposed model outperformed all models. Furthermore, classification performance was assessed before and after hair artifact removal using twelve pre-trained classifiers, demonstrating significant improvements in diagnostic accuracy. Explainable AI techniques, Grad-CAM, UMAP and attention heatmaps, were employed to interpret and validate the model’s focus. These results underscore DPA-HairNet’s effectiveness and potential integration into clinical workflows to enhance dermoscopic image analysis and diagnosis. Future work will explore generalization to additional artifact types and optimization for real-time clinical deployment.https://ieeexplore.ieee.org/document/11062859/Hair artifact removaldermoscopic image analysisneural networkexplainable AIhealthcare |
| spellingShingle | F M Javed Mehedi Shamrat Mohd Yamani Idna Idris Chowdhury Forhadul Karim Xujuan Zhou Raj Gururajan DPA-HairNet: A Dual Encoder Attention Based Network for Hair Artifact Removal in Dermoscopic Images IEEE Access Hair artifact removal dermoscopic image analysis neural network explainable AI healthcare |
| title | DPA-HairNet: A Dual Encoder Attention Based Network for Hair Artifact Removal in Dermoscopic Images |
| title_full | DPA-HairNet: A Dual Encoder Attention Based Network for Hair Artifact Removal in Dermoscopic Images |
| title_fullStr | DPA-HairNet: A Dual Encoder Attention Based Network for Hair Artifact Removal in Dermoscopic Images |
| title_full_unstemmed | DPA-HairNet: A Dual Encoder Attention Based Network for Hair Artifact Removal in Dermoscopic Images |
| title_short | DPA-HairNet: A Dual Encoder Attention Based Network for Hair Artifact Removal in Dermoscopic Images |
| title_sort | dpa hairnet a dual encoder attention based network for hair artifact removal in dermoscopic images |
| topic | Hair artifact removal dermoscopic image analysis neural network explainable AI healthcare |
| url | https://ieeexplore.ieee.org/document/11062859/ |
| work_keys_str_mv | AT fmjavedmehedishamrat dpahairnetadualencoderattentionbasednetworkforhairartifactremovalindermoscopicimages AT mohdyamaniidnaidris dpahairnetadualencoderattentionbasednetworkforhairartifactremovalindermoscopicimages AT chowdhuryforhadulkarim dpahairnetadualencoderattentionbasednetworkforhairartifactremovalindermoscopicimages AT xujuanzhou dpahairnetadualencoderattentionbasednetworkforhairartifactremovalindermoscopicimages AT rajgururajan dpahairnetadualencoderattentionbasednetworkforhairartifactremovalindermoscopicimages |