{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,21]],"date-time":"2026-10-21T00:16:21Z","timestamp":1792541781608,"version":"4.5.0"},"reference-count":45,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,9,30]],"date-time":"2026-09-30T00:00:00Z","timestamp":1790726400000},"content-version":"vor","delay-in-days":91,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Sci Rep"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>AI-driven skin lesion diagnosis systems are revolutionizing dermatology practice but perform worse on darker skin populations, which threatens diagnostic equity in dermatology. Existing debiasing strategies rely on explicit skin tone annotations or adversarial removal of demographic information, which may be unavailable in practice and can harm diagnostic accuracy. We aimed to design a dermatology AI system that reduces skin-tone-related performance disparities across diverse skin populations without using skin tone labels during training. We propose a novel sketch-guided multimodal fusion framework that combines color (RGB) images with algorithmically generated structural sketches. Separate encoders extract representations from each modality, which are integrated by a gated fusion module that adaptively weighs color and structure features. A feature distillation loss aligns color features with their sketch counterparts to encourage structure-aware representations while retaining clinically relevant color cues. We trained and evaluated the model on the Fitzpatrick17k and Diverse Dermatology Images (DDI) datasets. The fairness performance was assessed with Equalized Opportunity, Equalized Odds, and Predictive Quality Disparity across skin tone groups. Out-of-domain robustness was examined using a DermaAmin and Atlas Dermatologico split. On Fitzpatrick17k, the proposed model yielded competitive accuracy and F1-score, while showing lower mean disparity in fairness evaluation than baseline methods. It also demonstrated reduced subgroup disparity across skin tone groups on the evaluated fairness metrics. In the out-of-domain evaluation setup, the model also exhibited improved fairness. On the DDI dataset, the framework showed consistent performance across different skin tone groups with reduced variation. Our proposed model shows promise in reducing skin-tone-related bias in dermatology while preserving utility without relying on explicit skin tone annotations. The observed improvements in skin tone fairness across two datasets suggest that our approach may help reduce measured subgroup disparities in automated skin lesion assessment, although clinical utility and real-world impact remain to be established through prospective validation.<\/jats:p>","DOI":"10.1038\/s41598-026-59316-x","type":"journal-article","created":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T12:18:15Z","timestamp":1782908295000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["Reducing skin tone bias in dermatology AI via sketch-guided multimodal fusion"],"prefix":"10.1038","volume":"16","author":[{"given":"Muhammad","family":"Nasir","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9104-4102","authenticated-orcid":false,"given":"Muhammad Zeeshan","family":"Tahir","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shaohua Kevin","family":"Zhou","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,7,1]]},"reference":[{"key":"59316_CR1","unstructured":"American Cancer Society. Cancer facts & figures. https:\/\/www.cancer.org\/content\/dam\/cancer-org\/research\/cancer-facts-and-statistics\/annual-cancer-facts-and-figures\/2025\/2025-cancer-facts-and-figures-acs.pdf, 2025. Accessed: 2025-06-22 (2025)"},{"issue":"36","key":"59316_CR2","doi-asserted-by":"publisher","first-page":"6199","DOI":"10.1200\/JCO.2009.23.4799","volume":"27","author":"CM Balch","year":"2009","unstructured":"Balch, C. M. et al. Final version of 2009 ajcc melanoma staging and classification. Journal of clinical oncology 27(36), 6199\u20136206 (2009).","journal-title":"Journal of clinical oncology"},{"key":"59316_CR3","doi-asserted-by":"publisher","first-page":"148","DOI":"10.1016\/j.ejca.2019.02.005","volume":"111","author":"J Titus","year":"2019","unstructured":"Titus, J. et al. A convolutional neural network trained with dermoscopic images performed on par with 145 dermatologists in a clinical melanoma image classification task. European Journal of Cancer 111, 148\u2013154 (2019).","journal-title":"European Journal of Cancer"},{"key":"59316_CR4","doi-asserted-by":"crossref","unstructured":"Pakzad, A., Abhishek, K. & Hamarneh, G. Circle: Color invariant representation learning for unbiased classification of skin lesions. In European Conference on Computer Vision, 203\u2013219. (Springer, 2022).","DOI":"10.1007\/978-3-031-25069-9_14"},{"key":"59316_CR5","doi-asserted-by":"crossref","unstructured":"Groh, M., et al. Evaluating deep neural networks trained on clinical images in dermatology with the fitzpatrick 17k dataset. In Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, 1820\u20131828 (2021).","DOI":"10.1109\/CVPRW53098.2021.00201"},{"key":"59316_CR6","doi-asserted-by":"crossref","unstructured":"Abbasi-Sureshjani, S., Raumanns, R., Michels, B.\u00a0E., Schouten, G. & Cheplygina, V. Risk of training diagnostic algorithms on data with demographic bias. In Interpretable and Annotation-Efficient Learning for Medical Image Computing: Third International Workshop, iMIMIC 2020, Second International Workshop, MIL3ID 2020, and 5th International Workshop, LABELS 2020, Held in Conjunction with MICCAI 2020, Lima, Peru, October 4\u20138, 2020, Proceedings 3, 183\u2013192 (Springer, 2020).","DOI":"10.1007\/978-3-030-61166-8_20"},{"key":"59316_CR7","doi-asserted-by":"crossref","unstructured":"Alipour, N., Burke, T. & Courtney, J. Skin type diversity: a case study in skin lesion datasets. Preprint available at https:\/\/www.researchsquare.com\/article\/rs-3160120\/v1 (2023).","DOI":"10.21203\/rs.3.rs-3160120\/v1"},{"key":"59316_CR8","doi-asserted-by":"crossref","unstructured":"Kinyanjui, N.\u00a0M., et al. Fairness of classifiers across skin tones in dermatology. In International conference on medical image computing and computer-assisted intervention, 320\u2013329. (Springer, 2020).","DOI":"10.1007\/978-3-030-59725-2_31"},{"key":"59316_CR9","doi-asserted-by":"crossref","unstructured":"Bevan, P.\u00a0J. & Atapour-Abarghouei, A. Detecting melanoma fairly: Skin tone detection and debiasing for skin lesion classification. In MICCAI Workshop on Domain Adaptation and Representation Transfer, 1\u201311 (Springer, 2022).","DOI":"10.1007\/978-3-031-16852-9_1"},{"issue":"3","key":"59316_CR10","doi-asserted-by":"publisher","first-page":"44","DOI":"10.3390\/computers11030044","volume":"11","author":"EC Adjobo","year":"2022","unstructured":"Adjobo, E. C., Mahama, A. T. S., Gouton, P. & Tossa, J. Towards accurate skin lesion classification across all skin categories using a pcnn fusion-based data augmentation approach. Computers 11(3), 44 (2022).","journal-title":"Computers"},{"key":"59316_CR11","unstructured":"Li, X., Cui, Z., Wu, Y., Gu, L. & Harada, T. Estimating and improving fairness with adversarial learning. arXiv preprint arXiv:2103.04243 (2021)."},{"issue":"15","key":"59316_CR12","doi-asserted-by":"publisher","first-page":"23771","DOI":"10.1007\/s11042-022-14211-1","volume":"82","author":"Adam Corbin","year":"2023","unstructured":"Corbin, Adam & Marques, Oge. Exploring strategies to generate fitzpatrick skin type metadata for dermoscopic images using individual typology angle techniques. Multimedia Tools and Applications 82(15), 23771\u201323795 (2023).","journal-title":"Multimedia Tools and Applications"},{"key":"59316_CR13","doi-asserted-by":"crossref","unstructured":"Salimi, B., Rodriguez, L., Howe, B. & Suciu, D. Interventional fairness: Causal database repair for algorithmic fairness. In Proceedings of the 2019 International Conference on Management of Data, 793\u2013810 (2019).","DOI":"10.1145\/3299869.3319901"},{"key":"59316_CR14","unstructured":"Creager, E., et al. Flexibly fair representation learning by disentanglement. In International conference on machine learning, 1436\u20131445 (PMLR, 2019)."},{"key":"59316_CR15","doi-asserted-by":"crossref","unstructured":"Wang, T., Zhao, J., Yatskar, M., Chang, K.-W. & Ordonez, V. Balanced datasets are not enough: Estimating and mitigating gender bias in deep image representations. In Proceedings of the IEEE\/CVF international conference on computer vision, 5310\u20135319 (2019).","DOI":"10.1109\/ICCV.2019.00541"},{"key":"59316_CR16","doi-asserted-by":"crossref","unstructured":"Wang, Z., et al. Fairness-aware adversarial perturbation towards bias mitigation for deployed deep models. In Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, 10379\u201310388 (2022).","DOI":"10.1109\/CVPR52688.2022.01013"},{"key":"59316_CR17","doi-asserted-by":"crossref","unstructured":"Kim, B., Kim, H., Kim, K., Kim, S. & Kim, J. Learning not to learn: Training deep neural networks with biased data. In Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, 9012\u20139020 (2019).","DOI":"10.1109\/CVPR.2019.00922"},{"key":"59316_CR18","doi-asserted-by":"crossref","unstructured":"Du, S., Hers, B., Bayasi, N., Hamarneh, G. & Garbi R. Fairdisco: Fairer ai in dermatology via disentanglement contrastive learning. In European Conference on Computer Vision, 185\u2013202. (Springer, 2022).","DOI":"10.1007\/978-3-031-25069-9_13"},{"key":"59316_CR19","doi-asserted-by":"crossref","unstructured":"Wu, Y., Zeng, D., Xu, X., Shi, Y. & Hu, J. Fairprune: Achieving fairness through pruning for dermatological disease diagnosis. In International Conference on Medical Image Computing and Computer-Assisted Intervention, 743\u2013753 (Springer, 2022).","DOI":"10.1007\/978-3-031-16431-6_70"},{"key":"59316_CR20","doi-asserted-by":"crossref","unstructured":"Xu, Z., Zhao, S., Quan, Q., Yao, Q. & Zhou, S.\u00a0K. Fairadabn: Mitigating unfairness with adaptive batch normalization and its application to dermatological disease classification. In International Conference on Medical Image Computing and Computer-Assisted Intervention, 307\u2013317. (Springer, 2023).","DOI":"10.1007\/978-3-031-43895-0_29"},{"key":"59316_CR21","unstructured":"Yao, R., Cui, Z., Li, X. & Gu, L. Improving fairness in image classification via sketching. arXiv preprint arXiv:2211.00168 (2022)."},{"issue":"4","key":"59316_CR22","doi-asserted-by":"publisher","first-page":"551","DOI":"10.1016\/S0190-9622(94)70061-3","volume":"30","author":"F Nachbar","year":"1994","unstructured":"Nachbar, F. et al. The abcd rule of dermatoscopy: high prospective value in the diagnosis of doubtful melanocytic skin lesions. Journal of the American Academy of Dermatology 30(4), 551\u2013559 (1994).","journal-title":"Journal of the American Academy of Dermatology"},{"issue":"12","key":"59316_CR23","doi-asserted-by":"publisher","first-page":"1563","DOI":"10.1001\/archderm.134.12.1563","volume":"134","author":"Giuseppe Argenziano","year":"1998","unstructured":"Argenziano, Giuseppe et al. Epiluminescence microscopy for the diagnosis of doubtful melanocytic skin lesions: comparison of the abcd rule of dermatoscopy and a new 7-point checklist based on pattern analysis. Archives of dermatology 134(12), 1563\u20131570 (1998).","journal-title":"Archives of dermatology"},{"key":"59316_CR24","doi-asserted-by":"crossref","unstructured":"Kawahara, J., BenTaieb, A. & Hamarneh, G. Deep features to classify skin lesions. In 2016 IEEE 13th international symposium on biomedical imaging (ISBI), 1397\u20131400 (IEEE, 2016).","DOI":"10.1109\/ISBI.2016.7493528"},{"issue":"7639","key":"59316_CR25","doi-asserted-by":"publisher","first-page":"115","DOI":"10.1038\/nature21056","volume":"542","author":"A Esteva","year":"2017","unstructured":"Esteva, A. et al. Dermatologist-level classification of skin cancer with deep neural networks. Nature 542(7639), 115\u2013118 (2017).","journal-title":"Nature"},{"key":"59316_CR26","doi-asserted-by":"crossref","unstructured":"A. N. Omeroglu, Hussein, M. A., Mohammed, E. A., Oral, S. & Aydin, A novel soft attention-based multi-modal deep learning framework for multi-label skin lesion classification. Engineering Applications of Artificial Intelligence 120, 105897 (2023).","DOI":"10.1016\/j.engappai.2023.105897"},{"key":"59316_CR27","unstructured":"Flosdorf, Carolin, Engelker, Justin, Keller, Igor & Mohr, Nicolas. Skin cancer detection utilizing deep learning: Classification of skin lesion images using a vision transformer. arXiv preprint arXiv:2407.18554 (2024)."},{"key":"59316_CR28","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2024.103188","volume":"95","author":"Ching-Hao Chiu","year":"2024","unstructured":"Chiu, Ching-Hao., Chen, Yu-Jen., Yawen, Wu., Shi, Yiyu & Ho, Tsung-Yi. Achieve fairness without demographics for dermatological disease diagnosis. Medical Image Analysis 95, 103188 (2024).","journal-title":"Medical Image Analysis"},{"key":"59316_CR29","unstructured":"Paxton, K., Aslansefat, K., Thakker, D., Papadopoulos, Y. & Maslekar, T. Enhancing fairness in skin lesion classification for medical diagnosis using prune learning. IEEE Journal of Biomedical and Health Informatics (20260."},{"key":"59316_CR30","doi-asserted-by":"crossref","unstructured":"He, K., Zhang, X., Ren, S. & Sun, J. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770\u2013778 (2016).","DOI":"10.1109\/CVPR.2016.90"},{"key":"59316_CR31","unstructured":"Chen, T., Kornblith, S., Norouzi, M. & Hinton, G. A simple framework for contrastive learning of visual representations. In International conference on machine learning, pp. 1597\u20131607. PmLR, (2020)."},{"key":"59316_CR32","unstructured":"Radford, A., et al. Learning transferable visual models from natural language supervision. In International conference on machine learning, 8748\u20138763 (PmLR, 2021)."},{"key":"59316_CR33","unstructured":"Romero, A. et al. Fitnets: Hints for thin deep nets. arXiv preprint arXiv:1412.6550 (2014)."},{"key":"59316_CR34","unstructured":"Hardt, M., Price, E. & Srebro, N. Equality of opportunity in supervised learning. Advances in neural information processing systems 29, (2016)."},{"issue":"CSCW2","key":"59316_CR35","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3555634","volume":"6","author":"Matthew Groh","year":"2022","unstructured":"Groh, Matthew, Harris, Caleb, Daneshjou, Roxana, Badri, Omar & Koochek, Arash. Towards transparency in dermatology image datasets with skin tone annotations by experts, crowds, and an algorithm. Proceedings of the ACM on Human-Computer Interaction 6(CSCW2), 1\u201326 (2022).","journal-title":"Proceedings of the ACM on Human-Computer Interaction"},{"key":"59316_CR36","unstructured":"AlKattash, J.\u00a0A. Dermaamin. https:\/\/www.dermaamin.com\/site\/ (2025)."},{"key":"59316_CR37","unstructured":"Samuel, F.\u00a0D.\u00a0S. Atlas dermatol\u00f3gico. http:\/\/atlasdermatologico.com.br\/index.jsf (2025)."},{"key":"59316_CR38","doi-asserted-by":"crossref","unstructured":"Daneshjou, R. et al. Disparities in dermatology ai performance on a diverse, curated clinical image set. Science Advances 8, (2021).","DOI":"10.1126\/sciadv.abq6147"},{"key":"59316_CR39","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1016\/j.swevo.2011.02.002","volume":"1","author":"Joaqu\u00edn Derrac","year":"2011","unstructured":"Derrac, Joaqu\u00edn, Garc\u00eda, Salvador, Molina, Daniel & Herrera, Francisco. A practical tutorial on the use of nonparametric statistical tests as a methodology for comparing evolutionary and swarm intelligence algorithms. Swarm Evol. Comput. 1, 3\u201318 (2011).","journal-title":"Swarm Evol. Comput."},{"key":"59316_CR40","unstructured":"Zong, Y., Yang, Y. & Hospedales, T. Medfair: Benchmarking fairness for medical imaging. In International Conference on Learning Representations (ICLR 2023)."},{"key":"59316_CR41","unstructured":"Zhao, H., Coston, A., Adel, T. & Gordon, G.\u00a0J. Conditional learning of fair representations. ArXiv, arXiv:1910.07162 (2019)."},{"key":"59316_CR42","doi-asserted-by":"crossref","unstructured":"Puyol-Ant\u00f3n, E., et al. Fairness in cardiac mr image analysis: An investigation of bias due to data imbalance in deep learning based segmentation. ArXiv, arXiv:2106.12387 (2021).","DOI":"10.1007\/978-3-030-87199-4_39"},{"key":"59316_CR43","doi-asserted-by":"crossref","unstructured":"Kim, B., Kim, H., Kim, K., Kim, S. & Kim, J. Learning not to learn: Training deep neural networks with biased data. 2019 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 9004\u20139012 (2018).","DOI":"10.1109\/CVPR.2019.00922"},{"key":"59316_CR44","doi-asserted-by":"crossref","unstructured":"M. Hasan Sarhan, Nassir, Navab, Abouzar, Eslami, Shadi & Albarqouni,. Fairness by learning orthogonal disentangled representations. ArXiv arXiv:2003.05707 (2020).","DOI":"10.1007\/978-3-030-58526-6_44"},{"key":"59316_CR45","doi-asserted-by":"crossref","unstructured":"Tartaglione, E., Barbano, C.\u00a0A. & Grangetto, M. End: Entangling and disentangling deep representations for bias correction. 2021 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 13503\u201313512 (2021).","DOI":"10.1109\/CVPR46437.2021.01330"}],"container-title":["Scientific Reports"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-59316-x","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-59316-x.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-59316-x.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,29]],"date-time":"2026-09-29T23:10:29Z","timestamp":1790723429000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-59316-x"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,1]]},"references-count":45,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["59316"],"URL":"https:\/\/doi.org\/10.1038\/s41598-026-59316-x","relation":{},"ISSN":["2045-2322"],"issn-type":[{"value":"2045-2322","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,1]]},"assertion":[{"value":"29 November 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"19 June 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"1 July 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors declare no competing interests.","order":1,"name":"Ethics","label":"Competing interests","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"This study did not involve any new human-subjects data collection, patient contact, or prospective clinical evaluation. All experiments were conducted on two publicly available, fully de-identified dermatology image datasets. No additional Institutional Review Board (IRB) approval or patient consent was required at the authors\u2019 institutions, in accordance with their standard policies on secondary use of public research datasets. We used both datasets strictly in accordance with the terms specified by their original publishers.","order":2,"name":"Ethics","label":"Ethics declaration","group":{"name":"EthicsHeading","label":"Declarations"}}],"article-number":"30271"}}