skin disorder research project

Fostering skin and epithelial biology research.

Since 2009, the Skin Biology & Diseases Resource-Based Center (formerly the Skin Disease Research Center) has promoted outstanding bench and clinical research in cutaneous biology by providing high-quality technical services, enhancing education and communication among our members and facilitating translational research. We are funded through the NIH National Institute of Arthritis and Musculoskeletal and Skin Diseases.

Our team of expert scientists provide outstanding and cost-effective services. We are eager to partner with you in the development of your projects.”

Read her message

Enrichment Programs

Access our three shared cores, which offer  a number of resource to assist our scientists in their investigations.

Meet Our Team

We welcome your questions and comments. Get in touch with our staff, or browse our member list to get an in-depth look at our scientists' backgrounds and latest research endeavors.

Contact Us   Meet Our Members

Skin Diseases Research Center

Mission and vision.

The mission of the Skin Diseases Research Center (SDRC) is to add value to our investigators through strength of focus on uniting basic and translational skin science. The vision of the SDRC is:

• To enhance career development and to fuel the projects of new and experienced investigators with a rich matrix of resources in order to speed the progress and enhance the quality of skin diseases research
• To generate a new knowledge base that will have a significant and sustained impact on cutaneous biology aimed at improving the quality of life of patients with skin disease
• To integrate the components and activities of the Skin Diseases Research Center (SDRC)
• To manage the fiscal operations
• To coordinate the core facilities and pilot and feasibility awarded programs
• To implement an enrichment program
• To foster new SDRC initiatives
• To translate innovative basic research to patients with skin diseases
• To facilitate utilization of genomics and informatics for skin diseases

The SDRC at Case Western Reserve University / University Hospitals (UH) Health Systems promotes cooperative interaction among investigators engaged directly in skin diseases research with those conducting state-of-the-art biomedical research in other disciplines that have relevance to skin disease. The center's primary goal is to integrate groups of investigators from Case Western Reserve, UH, and other institutions into a cohesive unit that will advance research relevant to the diagnosis, pathogenesis and treatment of skin diseases. The center has attracted a number of outstanding scientists and has been instrumental in developing new technologies and encouraging their timely transfer to the clinical setting. The SDRC has also developed informative educational programs in dermatologic diseases for its investigators, physicians and the public.

Major programs include:

• a pilot and feasibility grants program,
• core facilities to support research,
• a visiting lectureship program,
• special national and regional educational programs,
• a scholars exchange program,
• and a minority student and fellow research program.

SDRC Membership

The strong record of achievement of the SDRC at Case Western Reserve and UH is due to the accomplishments of the individual investigators and the synergistic interactions fostered by the center grant. When the SDRC was originally funded in 1988, its membership was 13 (four from the Department of Dermatology, ten from five other departments). At the present time it is comprised of more than 70 members from 34 different departments and centers within Case Western Reserve, University Hospitals of Cleveland, the Cleveland VA Medical Center, the Lerner Research Institute of Cleveland Clinic and Henry Ford Hospital in Detroit. SDRC investigators in the aggregate average approximately $10 million annually to funded research projects relevant to skin.

Taken together, the research base of the SDRC at Case Western Reserve has breadth and depth, is multidisciplinary and interrelated, and provides an excellent, vibrant platform for advancing the interests of patients with skin disease through the SDRC mechanisms.

Members of our university community are encouraged to participate in the center's enrichment programs (i.e., seminar series, visiting scholars' program, P&F research-in-progress lunches, etc.) and are offered the opportunity to submit proposals to the Pilot & Feasibility Study Program (P&FS). A major benefit of membership in the SDRC is unrestricted access to the core facilities.

The primary criterion for membership in the SDRC is a commitment to research applications in skin diseases. Interdisciplinary collaborations involving skin-related research are also looked upon favorably. Membership in the SDRC requires approval by the director with the concurrence of the executive committee.

The guidelines for membership are as follows:

• A member of the Department of Dermatology who is actively engaged in research.
• A non-member of the Department of Dermatology with a funded research project directly relevant to dermatology or the SDRC's cores.
• A non-member of the Department of Dermatology who has been awarded an SDRC P&F grant.
• A non-member of the Department of Dermatology who has a productive collaboration with a member of the Department of Dermatology on a skin-related project.

Continued membership requires:

• Ongoing productivity in skin-related research leading to the publication of peer-reviewed articles and acquisition of extramural grants.
• Participation in SDRC-related activities including its cores, committees, and enrichment program activities

• Human Samples
• Human Analytic Techniques
• Grants & Outreach
• Ask a Translational Researcher

The Human Skin Disease Resource Center

Accelerating human skin disease research, we provide advice, samples and cutting edge analytic techniques to investigators at any institution with the goal of facilitating human translational skin disease research.  the advantages of harvard, without the snow..

The Human Skin Disease Resource Center is committed to supporting diversity in skin disease research. We have funds available to support individuals from groups underrepresented in medicine who wish to carry our projects through the Center.

Questions about how we can assist? Contact us!

Member Login

• Sign Up for Newsletter
• About GRIDD
• GRIDD Study
• Phase 3 and 4
• Phases & Outcomes
• Resources & Publications 
• Meet The Team
• Get Involved
• Science Behind GRIDD

GlobalSkin's GRIDD Study: A Wrap-up Report

GlobalSkin has successfully concluded its Global Research on the Impact of Dermatological Diseases (GRIDD) Study, a novel patient-led research initiative which it launched on June 5, 2023 in collaboration with Cardiff University (UK) and University Medical Centre Hamburg-Eppendorf (Germany).

The GRIDD Study gathered global data about the real impact of conditions of the skin, hair, nails, and mucosa. The objective of collecting these insights is to help reshape global perspectives on dermatological diseases, stimulate research efforts, optimize healthcare system spending, and inform future policies for improved patient outcomes. At the heart of the study was the new Patient-Reported Impact of Dermatological Diseases (PRIDD) measure, a tool we developed with more than 6000 patients over the past five years to capture and quantify patient experiences.

• Privacy Policy
• Terms of Service
• Social Media Disclaimer

Copyright 2018- 2024 International Alliance of Dermatology Patient Organizations. All right reserved.

• Vision & Mission
• Collaborations
• Board & Committees
• Partnerships
• Member List
• Communities
• Member Impact Fund
• Patient Expert Scholarship
• World Atopic Eczema Day Fund
• Rare Disease Day Fund
• Publications
• Events Listing
• Workshop Series
• GlobalSkin 2019 Conference
• GlobalSkin 2021 Conference
• GlobalSkin 2023 Conference
• GlobalSkin 2025 Conference

  Login

Sign in to your account

• Forgot your username?
• Forgot your password?

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Journal Proposal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Recommended Articles
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

Machine learning methods in skin disease recognition: a systematic review.

1. Introduction

2. skin lesion datasets and image preprocessing, 2.1. skin lesion datasets, 2.2. image preprocessing, 2.3. segmentation and classification evaluation metrics, 3. skin lesion segmentation methods, 3.1. traditional segmentation methods, 3.2. dl skin lesion segmentation methods, 4. skin lesion classification, 4.1. feature extraction and selection, 4.2. dl-based feature extract and selection methods, 4.3. traditional ml models for skin disease classification, 4.4. deep learning models for skin disease classification, 5. current status, challenges, and outlook, 5.1. current research publication status, 5.2. challenges and outlooks, 5.2.1. macroscopic images with robust diagnosis, 5.2.2. racial and geographical biases in public datasets, 5.2.3. dataset characteristics and dl methods, 6. conclusions, author contributions, institutional review board statement, informed consent statement, data availability statement, conflicts of interest.

• ALKolifi-ALEnezi, N.S. A Method Of Skin Disease Detection Using Image Processing And Machine Learning. Procedia Comput. Sci. 2019 , 163 , 85–92. [ Google Scholar ] [ CrossRef ]
• Skin Disorders: Pictures, Causes, Symptoms, and Treatment. Available online: https://www.healthline.com/health/skin-disorders (accessed on 21 February 2023).
• ISIC Archive. Available online: https://www.isic-archive.com/#!/topWithHeader/wideContentTop/main (accessed on 20 February 2023).
• Sun, J.; Yao, K.; Huang, K.; Huang, D. Machine learning applications in scaffold based bioprinting. Mater. Today Proc. 2022 , 70 , 17–23. [ Google Scholar ] [ CrossRef ]
• Haenssle, H.A.; Fink, C.; Schneiderbauer, R.; Toberer, F.; Buhl, T.; Blum, A.; Kalloo, A.; Hassen, A.B.H.; Thomas, L.; Enk, A.; et al. Man against machine: Diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists. Ann. Oncol. 2018 , 29 , 1836–1842. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Rotemberg, V.; Kurtansky, N.; Betz-Stablein, B.; Caffery, L.; Chousakos, E.; Codella, N.; Combalia, M.; Dusza, S.; Guitera, P.; Gutman, D.; et al. A patient-centric dataset of images and metadata for identifying melanomas using clinical context. Sci. Data 2021 , 8 , 34. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Melanoma Skin Cancer Rreport. Melanoma UK. 2020. Available online: https://www.melanomauk.org.uk/2020-melanoma-skin-cancer-report (accessed on 20 February 2023).
• Mendonça, T.; Ferreira, P.M.; Marques, J.S.; Marcal, A.R.; Rozeira, J. PH 2-A dermoscopic image database for research and benchmarking. In Proceedings of the 2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Osaka, Japan, 3–7 July 2013; pp. 5437–5440. [ Google Scholar ]
• Tschandl, P.; Rosendahl, C.; Kittler, H. The HAM10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions. Sci. Data 2018 , 5 , 1–9. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Combalia, M.; Codella, N.C.; Rotemberg, V.; Helba, B.; Vilaplana, V.; Reiter, O.; Carrera, C.; Barreiro, A.; Halpern, A.C.; Puig, S.; et al. Bcn20000: Dermoscopic lesions in the wild. arXiv 2019 , arXiv:1908.02288. [ Google Scholar ]
• Dermnet. Kaggle. Available online: https://www.kaggle.com/datasets/shubhamgoel27/dermnet (accessed on 20 February 2023).
• Giotis, I.; Molders, N.; Land, S.; Biehl, M.; Jonkman, M.F.; Petkov, N. MED-NODE: A computer-assisted melanoma diagnosis system using non-dermoscopic images. Expert Syst. Appl. 2015 , 42 , 6578–6585. [ Google Scholar ] [ CrossRef ]
• Yap, J.; Yolland, W.; Tschandl, P. Multimodal skin lesion classification using deep learning. Exp. Dermatol. 2018 , 27 , 1261–1267. [ Google Scholar ] [ CrossRef ] [ Green Version ]
• Dermofit Image Library Available from The University of Edinburgh. Available online: https://licensing.edinburgh-innovations.ed.ac.uk/product/dermofit-image-library (accessed on 20 February 2023).
• Gutman, D.; Codella, N.C.; Celebi, E.; Helba, B.; Marchetti, M.; Mishra, N.; Halpern, A. Skin lesion analysis toward melanoma detection: A challenge at the international symposium on biomedical imaging (ISBI) 2016, hosted by the international skin imaging collaboration (ISIC). arXiv 2016 , arXiv:1605.01397. [ Google Scholar ]
• Codella, N.C.; Gutman, D.; Celebi, M.E.; Helba, B.; Marchetti, M.A.; Dusza, S.W.; Kalloo, A.; Liopyris, K.; Mishra, N.; Kittler, H.; et al. Skin lesion analysis toward melanoma detection: A challenge at the 2017 international symposium on biomedical imaging, hosted by the international skin imaging collaboration. In Proceedings of the 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018), Washington, DC, USA, 4–7 April 2018; pp. 168–172. [ Google Scholar ]
• Codella, N.; Rotemberg, V.; Tschandl, P.; Celebi, M.E.; Dusza, S.; Gutman, D.; Helba, B.; Kalloo, A.; Liopyris, K.; Marchetti, M.; et al. Skin lesion analysis toward melanoma detection 2018: A challenge hosted by the international skin imaging collaboration (isic). arXiv 2019 , arXiv:1902.03368. [ Google Scholar ]
• ISIC Challenge. Available online: https://challenge.isic-archive.com/landing/2019/ (accessed on 21 February 2023).
• Kawahara, J.; Daneshvar, S.; Argenziano, G.; Hamarneh, G. Seven-point checklist and skin lesion classification using multitask multimodal neural nets. IEEE J. Biomed. Health Inform. 2019 , 23 , 538–546. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Alahmadi, M.D.; Alghamdi, W. Semi-Supervised Skin Lesion Segmentation With Coupling CNN and Transformer Features. IEEE Access 2022 , 10 , 122560–122569. [ Google Scholar ] [ CrossRef ]
• Abhishek, K.; Hamarneh, G.; Drew, M.S. Illumination-based transformations improve skin lesion segmentation in dermoscopic images. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, Seattle, WA, USA, 14–19 June 2020; pp. 728–729. [ Google Scholar ]
• Oliveira, R.B.; Mercedes Filho, E.; Ma, Z.; Papa, J.P.; Pereira, A.S.; Tavares, J.M.R. Computational methods for the image segmentation of pigmented skin lesions: A review. Comput. Methods Programs Biomed. 2016 , 131 , 127–141. [ Google Scholar ] [ CrossRef ] [ Green Version ]
• Hameed, N.; Shabut, A.M.; Ghosh, M.K.; Hossain, M.A. Multi-class multi-level classification algorithm for skin lesions classification using machine learning techniques. Expert Syst. Appl. 2020 , 141 , 112961. [ Google Scholar ] [ CrossRef ]
• Toossi, M.T.B.; Pourreza, H.R.; Zare, H.; Sigari, M.H.; Layegh, P.; Azimi, A. An effective hair removal algorithm for dermoscopy images. Ski. Res. Technol. 2013 , 19 , 230–235. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Zhang, C.; Huang, G.; Yao, K.; Leach, M.; Sun, J.; Huang, K.; Zhou, X.; Yuan, L. A Comparison of Applying Image Processing and Deep Learning in Acne Region Extraction. J. Image Graph. 2022 , 10 , 1–6. [ Google Scholar ] [ CrossRef ]
• Shetty, B.; Fernandes, R.; Rodrigues, A.P.; Chengoden, R.; Bhattacharya, S.; Lakshmanna, K. Skin lesion classification of dermoscopic images using machine learning and convolutional neural network. Sci. Rep. 2022 , 12 , 18134. [ Google Scholar ] [ CrossRef ]
• Gulzar, Y.; Khan, S.A. Skin Lesion Segmentation Based on Vision Transformers and Convolutional Neural Networks—A Comparative Study. Appl. Sci. 2022 , 12 , 5990. [ Google Scholar ] [ CrossRef ]
• Vesal, S.; Ravikumar, N.; Maier, A. SkinNet: A deep learning framework for skin lesion segmentation. In Proceedings of the 2018 IEEE Nuclear Science Symposium and Medical Imaging Conference Proceedings (NSS/MIC), Sydney, Australia, 10–17 November 2018; pp. 1–3. [ Google Scholar ]
• Al-Masni, M.A.; Al-Antari, M.A.; Choi, M.T.; Han, S.M.; Kim, T.S. Skin lesion segmentation in dermoscopy images via deep full resolution convolutional networks. Comput. Methods Programs Biomed. 2018 , 162 , 221–231. [ Google Scholar ] [ CrossRef ]
• Ünver, H.M.; Ayan, E. Skin lesion segmentation in dermoscopic images with combination of YOLO and grabcut algorithm. Diagnostics 2019 , 9 , 72. [ Google Scholar ] [ CrossRef ] [ Green Version ]
• Thapar, P.; Rakhra, M.; Cazzato, G.; Hossain, M.S. A novel hybrid deep learning approach for skin lesion segmentation and classification. J. Healthc. Eng. 2022 , 2022 , 1709842. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ali, A.R.; Li, J.; Trappenberg, T. Supervised versus unsupervised deep learning based methods for skin lesion segmentation in dermoscopy images. In Proceedings of the Advances in Artificial Intelligence: 32nd Canadian Conference on Artificial Intelligence, Canadian AI 2019, Kingston, ON, Canada, 28–31 May 2019; pp. 373–379. [ Google Scholar ]
• Gupta, S.; Panwar, A.; Mishra, K. Skin disease classification using dermoscopy images through deep feature learning models and machine learning classifiers. In Proceedings of the IEEE EUROCON 2021–19th International Conference on Smart Technologies, Lviv, Ukraine, 6–8 July 2021; pp. 170–174. [ Google Scholar ]
• Kassem, M.A.; Hosny, K.M.; Fouad, M.M. Skin lesions classification into eight classes for ISIC 2019 using deep convolutional neural network and transfer learning. IEEE Access 2020 , 8 , 114822–114832. [ Google Scholar ] [ CrossRef ]
• Khan, M.A.; Javed, M.Y.; Sharif, M.; Saba, T.; Rehman, A. Multi-model deep neural network based features extraction and optimal selection approach for skin lesion classification. In Proceedings of the 2019 International Conference on Computer and Information Sciences (ICCIS), Aljouf, Saudi Arabia, 3–4 April 2019; pp. 1–7. [ Google Scholar ]
• Khan, M.A.; Sharif, M.; Akram, T.; Bukhari, S.A.C.; Nayak, R.S. Developed Newton-Raphson based deep features selection framework for skin lesion recognition. Pattern Recognit. Lett. 2020 , 129 , 293–303. [ Google Scholar ] [ CrossRef ]
• Nock, R.; Nielsen, F. Statistical region merging. IEEE Trans. Pattern Anal. Mach. Intell. 2004 , 26 , 1452–1458. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Hojjatoleslami, S.; Kittler, J. Region growing: A new approach. IEEE Trans. Image Process. 1998 , 7 , 1079–1084. [ Google Scholar ] [ CrossRef ] [ Green Version ]
• Levner, I.; Zhang, H. Classification-driven watershed segmentation. IEEE Trans. Image Process. 2007 , 16 , 1437–1445. [ Google Scholar ] [ CrossRef ]
• Maeda, J.; Kawano, A.; Yamauchi, S.; Suzuki, Y.; Marçal, A.; Mendonça, T. Perceptual image segmentation using fuzzy-based hierarchical algorithm and its application to dermoscopy images. In Proceedings of the 2008 IEEE Conference on Soft Computing in Industrial Applications, Muroran, Japan, 25–27 June 2008; pp. 66–71. [ Google Scholar ]
• Devi, S.S.; Laskar, R.H.; Singh, N.H. Fuzzy C-means clustering with histogram based cluster selection for skin lesion segmentation using non-dermoscopic images. Int. J. Interact. Multimed. Artif. Intell. 2020 , 6 , 1–6. [ Google Scholar ] [ CrossRef ]
• Bhati, P.; Singhal, M. Early stage detection and classification of melanoma. In Proceedings of the 2015 Communication, Control and Intelligent Systems (CCIS), Mathura, India, 7–8 November 2015; pp. 181–185. [ Google Scholar ]
• Henning, J.S.; Dusza, S.W.; Wang, S.Q.; Marghoob, A.A.; Rabinovitz, H.S.; Polsky, D.; Kopf, A.W. The CASH (color, architecture, symmetry, and homogeneity) algorithm for dermoscopy. J. Am. Acad. Dermatol. 2007 , 56 , 45–52. [ Google Scholar ] [ CrossRef ]
• Soyer, H.P.; Argenziano, G.; Zalaudek, I.; Corona, R.; Sera, F.; Talamini, R.; Barbato, F.; Baroni, A.; Cicale, L.; Di Stefani, A.; et al. Three-point checklist of dermoscopy. Dermatology 2004 , 208 , 27–31. [ Google Scholar ] [ CrossRef ]
• Argenziano, G.; Catricalà, C.; Ardigo, M.; Buccini, P.; De Simone, P.; Eibenschutz, L.; Ferrari, A.; Mariani, G.; Silipo, V.; Sperduti, I.; et al. Seven-point checklist of dermoscopy revisited. Br. J. Dermatol. 2011 , 164 , 785–790. [ Google Scholar ] [ CrossRef ]
• Oliveira, R.B.; Pereira, A.S.; Tavares, J.M.R. Computational diagnosis of skin lesions from dermoscopic images using combined features. Neural Comput. Appl. 2019 , 31 , 6091–6111. [ Google Scholar ] [ CrossRef ] [ Green Version ]
• Zakeri, A.; Hokmabadi, A. Improvement in the diagnosis of melanoma and dysplastic lesions by introducing ABCD-PDT features and a hybrid classifier. Biocybern. Biomed. Eng. 2018 , 38 , 456–466. [ Google Scholar ] [ CrossRef ]
• Dalal, N.; Triggs, B. Histograms of oriented gradients for human detection. In Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05), San Diego, CA, USA, 20–26 June 2005; Volume 1, pp. 886–893. [ Google Scholar ]
• Jie, S.; Hong, G.S.; Rahman, M.; Wong, Y. Feature extraction and selection in tool condition monitoring system. In Proceedings of the AI 2002: Advances in Artificial Intelligence: 15th Australian Joint Conference on Artificial Intelligence, Canberra, Australia, 2–6 December 2002; pp. 487–497. [ Google Scholar ]
• Akram, T.; Lodhi, H.M.J.; Naqvi, S.R.; Naeem, S.; Alhaisoni, M.; Ali, M.; Haider, S.A.; Qadri, N.N. A multilevel features selection framework for skin lesion classification. Hum. Centric Comput. Inf. Sci. 2020 , 10 , 1–26. [ Google Scholar ] [ CrossRef ] [ Green Version ]
• Amin, J.; Sharif, A.; Gul, N.; Anjum, M.A.; Nisar, M.W.; Azam, F.; Bukhari, S.A.C. Integrated design of deep features fusion for localization and classification of skin cancer. Pattern Recognit. Lett. 2020 , 131 , 63–70. [ Google Scholar ] [ CrossRef ]
• Hegde, P.R.; Shenoy, M.M.; Shekar, B. Comparison of machine learning algorithms for skin disease classification using color and texture features. In Proceedings of the 2018 International Conference on Advances in Computing, Communications and Inform. (ICACCI), Bangalore, India, 19–22 September 2018; pp. 1825–1828. [ Google Scholar ]
• Yu, Z.; Jiang, X.; Zhou, F.; Qin, J.; Ni, D.; Chen, S.; Lei, B.; Wang, T. Melanoma recognition in dermoscopy images via aggregated deep convolutional features. IEEE Trans. Biomed. Eng. 2018 , 66 , 1006–1016. [ Google Scholar ] [ CrossRef ]
• Kassem, M.A.; Hosny, K.M.; Damaševičius, R.; Eltoukhy, M.M. Machine learning and deep learning methods for skin lesion classification and diagnosis: A systematic review. Diagnostics 2021 , 11 , 1390. [ Google Scholar ] [ CrossRef ]
• Hameed, N.; Shabut, A.; Hossain, M.A. A Computer-aided diagnosis system for classifying prominent skin lesions using machine learning. In Proceedings of the 2018 10th Computer Science and Electronic Engineering (CEEC), Essex, UK, 19–21 September 2018; pp. 186–191. [ Google Scholar ]
• Yao, K.; Huang, K.; Sun, J.; Jude, C. Ad-gan: End-to-end unsupervised nuclei segmentation with aligned disentangling training. arXiv 2021 , arXiv:2107.11022. [ Google Scholar ]
• Yao, K.; Huang, K.; Sun, J.; Jing, L.; Huang, D.; Jude, C. Scaffold-A549: A benchmark 3D fluorescence image dataset for unsupervised nuclei segmentation. Cogn. Comput. 2021 , 13 , 1603–1608. [ Google Scholar ] [ CrossRef ]
• Yao, K.; Sun, J.; Huang, K.; Jing, L.; Liu, H.; Huang, D.; Jude, C. Analyzing cell-scaffold interaction through unsupervised 3d nuclei segmentation. Int. J. Bioprinting 2022 , 8 , 495. [ Google Scholar ] [ CrossRef ]
• Yao, K.; Huang, K.; Sun, J.; Hussain, A.; Jude, C. PointNu-Net: Simultaneous Multi-tissue Histology Nuclei Segmentation and Classification in the Clinical Wild. arXiv 2021 , arXiv:2111.01557. [ Google Scholar ] [ CrossRef ]
• Ali, M.S.; Miah, M.S.; Haque, J.; Rahman, M.M.; Islam, M.K. An enhanced technique of skin cancer classification using deep convolutional neural network with transfer learning models. Mach. Learn. Appl. 2021 , 5 , 100036. [ Google Scholar ] [ CrossRef ]
• Shahin, A.H.; Kamal, A.; Elattar, M.A. Deep ensemble learning for skin lesion classification from dermoscopic images. In Proceedings of the 2018 9th Cairo International Biomedical Engineering Conference (CIBEC), Cairo, Egypt, 20–22 December 2018; pp. 150–153. [ Google Scholar ]
• Ichim, L.; Popescu, D. Melanoma detection using an objective system based on multiple connected neural networks. IEEE Access 2020 , 8 , 179189–179202. [ Google Scholar ] [ CrossRef ]
• Patil, R.; Bellary, S. Machine learning approach in melanoma cancer stage detection. J. King Saud-Univ.-Comput. Inf. Sci. 2022 , 34 , 3285–3293. [ Google Scholar ] [ CrossRef ]
• Sudre, C.H.; Li, W.; Vercauteren, T.; Ourselin, S.; Jorge Cardoso, M. Generalised dice overlap as a deep learning loss function for highly unbalanced segmentations. In Proceedings of the Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support: Third International Workshop, DLMIA 2017, and 7th International Workshop, ML-CDS 2017, Held in Conjunction with MICCAI 2017, Québec City, QC, Canada, 14 September 2017; pp. 240–248. [ Google Scholar ]
• Yeung, M.; Sala, E.; Schönlieb, C.B.; Rundo, L. Unified focal loss: Generalising dice and cross entropy-based losses to handle class imbalanced medical image segmentation. Comput. Med. Imaging Graph. 2022 , 95 , 102026. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Bertels, J.; Eelbode, T.; Berman, M.; Vandermeulen, D.; Maes, F.; Bisschops, R.; Blaschko, M.B. Optimizing the dice score and jaccard index for medical image segmentation: Theory and practice. In Proceedings of the Medical Image Computing and Computer Assisted Intervention–MICCAI 2019: 22nd International Conference, Shenzhen, China, 13–17 October 2019; pp. 92–100. [ Google Scholar ]
• van Beers, F.; Lindström, A.; Okafor, E.; Wiering, M.A. Deep Neural Networks with Intersection over Union Loss for Binary Image Segmentation. In Proceedings of the ICPRAM, Prague, Czech Republic, 19–21 February 2019; pp. 438–445. [ Google Scholar ]
• Ho, Y.; Wookey, S. The real-world-weight cross-entropy loss function: Modeling the costs of mislabeling. IEEE Access 2019 , 8 , 4806–4813. [ Google Scholar ] [ CrossRef ]
• Zhou, Q.; Shi, Y.; Xu, Z.; Qu, R.; Xu, G. Classifying melanoma skin lesions using convolutional spiking neural networks with unsupervised stdp learning rule. IEEE Access 2020 , 8 , 101309–101319. [ Google Scholar ] [ CrossRef ]
• Alsaade, F.W.; Aldhyani, T.H.; Al-Adhaileh, M.H. Developing a recognition system for diagnosing melanoma skin lesions using artificial intelligence algorithms. Comput. Math. Methods Med. 2021 , 2021 , 1–20. [ Google Scholar ] [ CrossRef ]
• Yao, K.; Su, Z.; Huang, K.; Yang, X.; Sun, J.; Hussain, A.; Coenen, F. A novel 3D unsupervised domain adaptation framework for cross-modality medical image segmentation. IEEE J. Biomed. Health Inform. 2022 , 26 , 4976–4986. [ Google Scholar ] [ CrossRef ]
• Yao, K.; Gao, P.; Yang, X.; Sun, J.; Zhang, R.; Huang, K. Outpainting by queries. In Proceedings of the Computer Vision–ECCV 2022: 17th European Conference, Tel Aviv, Israel, 23–27 October 2022; pp. 153–169. [ Google Scholar ]

Click here to enlarge figure

Share and Cite

Sun, J.; Yao, K.; Huang, G.; Zhang, C.; Leach, M.; Huang, K.; Yang, X. Machine Learning Methods in Skin Disease Recognition: A Systematic Review. Processes 2023 , 11 , 1003. https://doi.org/10.3390/pr11041003

Sun J, Yao K, Huang G, Zhang C, Leach M, Huang K, Yang X. Machine Learning Methods in Skin Disease Recognition: A Systematic Review. Processes . 2023; 11(4):1003. https://doi.org/10.3390/pr11041003

Sun, Jie, Kai Yao, Guangyao Huang, Chengrui Zhang, Mark Leach, Kaizhu Huang, and Xi Yang. 2023. "Machine Learning Methods in Skin Disease Recognition: A Systematic Review" Processes 11, no. 4: 1003. https://doi.org/10.3390/pr11041003

Article Metrics

Article access statistics, further information, mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals

• Introduction to Genomics
• Educational Resources
• Policy Issues in Genomics
• The Human Genome Project
• Funding Opportunities
• Funded Programs & Projects
• Division and Program Directors
• Scientific Program Analysts
• Contacts by Research Area
• News & Events
• Research Areas
• Research Investigators
• Research Projects
• Clinical Research
• Data Tools & Resources
• Genomics & Medicine
• Family Health History
• For Patients & Families
• For Health Professionals
• Jobs at NHGRI
• Training at NHGRI
• Funding for Research Training
• Professional Development Programs
• NHGRI Culture
• Social Media
• Broadcast Media
• Image Gallery
• Press Resources
• Organization
• NHGRI Director
• Mission and Vision
• Policies and Guidance
• Institute Advisors
• Strategic Vision
• Leadership Initiatives
• Diversity, Equity, and Inclusion
• Partner with NHGRI
• Staff Search

NIH scientists find treatment for rare genetic skin disorder

• Share on Facebook
• Submit to Reddit
• Share on LinkedIn

Genome sequencing reveals genetic basis for disabling pansclerotic morphea, a severe inflammatory disease.  

Researchers at the National Institutes of Health and their colleagues have identified genomic variants that cause a rare and severe inflammatory skin disorder, known as disabling pansclerotic morphea, and have found a potential treatment. Scientists discovered that people with the disorder have an overactive version of a protein called STAT4, which regulates inflammation and wound healing. The work also identified a drug that targets an important feedback loop controlled by the STAT4 protein and significantly improves symptoms in these patients. The results were published in the New England Journal of Medicine .

The study was led by researchers at the National Human Genome Research Institute (NHGRI), part of NIH, in collaboration with researchers from the University of California, San Diego (UCSD) and the University of Pittsburgh. Researchers from the National Institute of Arthritis and Musculoskeletal and Skin Diseases and the National Institute of Allergy and Infectious Diseases, both part of NIH, also participated in the study.

Only a handful of patients have been diagnosed with disabling pansclerotic morphea, a disorder first described in the medical literature around 100 years ago. The disorder causes severe skin lesions and poor wound healing, leading to deep scarring of all layers of the skin and muscles. The muscles eventually harden and break down while the joints stiffen, leading to reduced mobility. Because the disorder is so rare, its genetic cause had not been identified until now.

“Researchers previously thought that this disorder was caused by the immune system attacking the skin,” said Sarah Blackstone, a predoctoral fellow within NHGRI's Inflammatory Disease Section, a medical student at the University of South Dakota and co-first author of the study. “However, we found that this is an oversimplification, and that both skin and the immune system play an active role in disabling pansclerotic morphea.”

The researchers used genome sequencing to study four individuals with disabling pansclerotic morphea and found that all four have genomic variants in the STAT4 gene. The STAT4 gene encodes a type of protein that helps turn genes on and off, known as a transcription factor. The STAT4 protein not only plays a role in fighting infections but also controls important aspects of wound-healing in the skin.

The findings of this study open doors for JAK inhibitors to be a potential treatment for other inflammatory skin disorders or disorders related to tissue scarring, whether it is scarring of the lungs, liver or bone marrow.

The scientists found that the STAT4 genomic variants result in an overactive STAT4 protein in these four patients, creating a positive feedback loop of inflammation and impaired wound-healing that worsens over time. To stop this harmful feedback loop, they targeted another protein in the inflammatory pathway that interacts with the STAT4 molecule and is called Janus kinase, also known as JAK. When the researchers treated the patients with a JAK-inhibiting drug called ruxolitinib, the patients’ rashes and ulcers dramatically improved. 

“So far, there has not been a standard treatment for this disorder because it’s so rare and not well-understood. However, our study gives an important new treatment option for these patients,” said Blackstone.  

Existing treatments for disabling pansclerotic morphea are designed to halt the progression of the disorder, but previous therapies have been mostly ineffective, often with severe side effects. People with the disorder typically don’t live more than 10 years after their diagnosis.

The study suggests that ruxolitinib could be an effective treatment for patients with this disorder. Ruxolitinib is part of a broader class of drugs called JAK inhibitors, which are commonly used to treat arthritis, eczema, ulcerative colitis and other chronic inflammatory diseases. 

“The findings of this study open doors for JAK inhibitors to be a potential treatment for other inflammatory skin disorders or disorders related to tissue scarring, whether it is scarring of the lungs, liver or bone marrow,” said Dan Kastner, M.D., Ph.D. , an NIH distinguished investigator, head of NHGRI’s Inflammatory Disease Section and a senior author of the paper. 

“We hope to continue studying other molecules in this pathway and how they are altered in patients with disabling pansclerotic morphea and related conditions to find clues to understanding a broader array of more common diseases,” said Lori Broderick, M.D., Ph.D., a senior author of the paper and an associate professor at UCSD. 

About NHGRI and NIH

About the National Human Genome Research Institute (NHGRI):  At NHGRI, we are focused on advances in genomics research. Building on our leadership role in the initial sequencing of the human genome, we collaborate with the world's scientific and medical communities to enhance genomic technologies that accelerate breakthroughs and improve lives. By empowering and expanding the field of genomics, we can benefit all of humankind. For more information about NHGRI and its programs, visit  www.genome.gov . About the National Institutes of Health (NIH):  NIH, the nation's medical research agency, includes 27 Institutes and Centers and is a component of the U.S. Department of Health and Human Services. NIH is the primary federal agency conducting and supporting basic, clinical, and translational medical research, and is investigating the causes, treatments, and cures for both common and rare diseases. For more information about NIH and its programs, visit  www.nih.gov .

Press Contact

Last updated: June 1, 2023

Recent and Current Projects

With the exciting advances that are continuously being made in dermatology, there is increasing need to understand the multiple components of dermatologic disease to maximize benefit to patients.  The Vashi Lab’s continued mission is to apply the highest standards of care and rigorous evaluation to questions in dermatology.  We combine clinical expertise with analytical approaches to understand the skin and dermatologic disease in order to improve patient outcomes while advancing healthcare delivery.

Dr. Vashi’s research interests include a wide variety of topics related to both medical and cosmetic dermatology.  A few of her recent projects are described below.

Societal obsession with beauty is deeply engrained in our past, with the appreciation of human aesthetics dating back to early Greek civilization.  Both personal preferences and cultural standards influence our ideas on beauty, and there is substantial agreement as to what constitutes human beauty within a society at any given point in time. In the study below, Dr. Vashi examined how our societal perceptions of beauty have changed over the past 27 years using People Magazine’s World’s Most Beautiful lists from 1990 and 2017.

Maymone MBC, Neamah HH, Secemsky EA, Kundu RV, Saade D, Vashi NA. The Most Beautiful People: Evolving Standards of Beauty. JAMA Dermatol. Published online October 11, 2017. doi:10.1001/jamadermatol.2017.3693

Dr. Vashi had over 100 media exposures including but not limited to NBC News, NewsWeek, MSN News, USNews, Yahoo News, GoodHousekeeping, ABC News, Bazaar, Cosmopolitan, and Chicago Tribune in reference to this study.  With an international presence, it had translation and media exposures in over 20 different countries and languages.  In addition, it was rated the #2 “Most Talked About Article of 2017” by JAMA Dermatology .

See  NBC News’ discussion of the findings of Dr. Vashi’s study in the article “ What Makes Someone ‘Most Beautiful’ Is Changing, Study Says .”

Sun Protection

Hyperpigmentation, a common issue seen by dermatologists, can worsen when exposed to the sun. The study below explores the different ways that patients with hyperpigmentation protect themselves from the sun’s harmful UV rays.

Maymone M, Neamah HH, Wirya SA, Patzelt NM, Zancanaro PQ, Vashi NA. Sun protective behaviors in patients with cutaneous hyperpigmentation: A cross-sectional study. J Am Acad Dermatol. 2017;76(5):841–846.e2.

In April 2017, Yahoo! News published the article “ How Hyperpigmentation Patients Shield Themselves from the Sun ” describing Dr. Neelam Vashi’s findings.

Melasma is a common disorder of hyperpigmentation that can worsen when exposed to the sun and is often difficult to treat. Thus, it is important to know the extent of disease to provide proper patient counseling and treatment guidance. Dr. Neelam Vashi researched different techniques as aids for diagnosing disease extent.

Wirya SA, Maymone MBC, Widjajahakim R, Vashi NA. Subclinical melasma: Determining disease extent. J Am Acad Dermatol. 2017;77(2):e41-e42.

Dr. Neelam Vashi was interviewed on this subject by WCVB-TV, Channel 5.

Aging of the skin is clinically described by wrinkles, sunspots, uneven skin color, and sagging skin; however, these signs vary across ethnicity. This article looks at how variations in cutaneous aging are related to differences in skin structure and function.

Vashi NA, Maymone M, Kundu RV. Aging differences in ethnic skin. J Clin Aesthet Dermatol. 2016;9(1):31-38.

Dr. Neelam Vashi appeared in the article “ Outsmart Aging. Your ethnicity plays a major role in how your skin matures. Face down our challenge with a personalized plan. ” featured in Dr. Oz’s The Good Life magazine.

TRANSFORMING DERMATOLOGY: BENEFITS OF ARTIFICIAL INTELLIGENCE ON DIAGNOSING AND MANAGING SKIN DISEASES IN SCHOOL CHILDREN

• This person is not on ResearchGate, or hasn't claimed this research yet.

Discover the world's research

• 25+ million members
• 160+ million publication pages
• 2.3+ billion citations

No full-text available

To read the full-text of this research, you can request a copy directly from the authors.

• Khatiya Cheldize
• Danielle Brown
• Esther E. Freeman

• Nanjappachetty Govindarajan

• Manisha Balai
• Ashok Kumar Khare

• Bhushan Kumar
• Recruit researchers
• Join for free
• Login Email Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google Welcome back! Please log in. Email · Hint Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google No account? Sign up

An official website of the United States government

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.

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.

• Publications
• Account settings

The PMC website is updating on October 15, 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Indian Dermatol Online J
• v.5(1); Jan-Mar 2014

Skin diseases and conditions among students of a medical college in southern India

Nitin joseph.

Department of Community Medicine, Kasturba Medical College, Manipal University, Mangalore, India

Ganesh S Kumar

1 Department of Community Medicine, JIPMER, Puducherry, India

Maria Nelliyanil

2 Department of Community Medicine, A. J. Institute of Medical Sciences, Mangalore, India

Introduction:

Skin diseases are a common problem among young adults. There is paucity of data about it among medical students. This study aimed to find out the pattern of skin disorders and to describe their association with various socio-demographic factors among medical students.

Materials and Methods:

This cross-sectional study was conducted in June 2011 in a medical college in Mangalore, Karnataka. Two-hundred and seventy eight medical students were chosen from the 4 th , 6 th and 8 th semester through convenient sampling method. Data on hair and skin morbidities suffered over past 1 year and its associated factors were collected using a self-administered questionnaire.

Most of the participants 171 (61.5%) were of the age group 20-21 years and majority were females 148 (53.2%). The most common hair/skin morbidities suffered in the past one year were acne 185 (66.6%), hair loss 165 (59.3%), and sun tan 147 (52.9%). Fungal infection ( P = 0.051) and severe type of acne ( P = 0.041) were seen significantly more among males while hair morbidities like hair loss ( P = 0.003), split ends of hairs ( P < 0.0001) and dandruff ( P =0.006) were seen significantly more among female students. Patterned baldness ( P = 0.018) and sun tan ( P < 0.0001) were significantly more among non-Mangalorean students than native Mangaloreans. Presence of dandruff was significantly associated with hair loss ( P = 0.039) and usage of sunscreen was found to protect from developing sun tans ( P = 0.049).

Conclusion:

Skin disorders, particularly the cosmetic problems are very common among medical students. Gender and place of origin were found to significantly influence the development of certain morbidities.

INTRODUCTION

Skin diseases are a major health problem affecting a high proportion of the population in India.[ 1 ] Skin diseases can place a heavy emotional and psychological burden on patients that may be far worse than the physical impact.[ 2 ] Increased consciousness especially among the youth of their body and beauty further aggravates their anxiety.[ 3 ]

Many factors determine the pattern and prevalence of cutaneous diseases among the youth such as gender, race, personal hygiene, quality of skin care, environmental milieu and diet.[ 4 ] In some instances, patients appear to produce their skin lesions as an outlet for nervous tensions arising from interpersonal conflicts and/or unresolved emotional problems.[ 5 ]

Even though dermatology is characterized by an enormous range of disease/reaction patterns, prevalence surveys suggest that the bulk of skin diseases belong to fewer than ten categories.[ 6 ] Such observations are useful in developing educational and preventive health programs for the benefit of university students. Their proper management at earlier stages with education of students is important to prevent disfiguring complications and psychological sequelae later in life.[ 3 ]

However, very few studies have been carried out in India to find out the problem of skin diseases and that especially among the medical students. The reason for this negligence could be the low mortality rate of the majority of skin diseases in comparison with other diseases. This has also resulted in international health policy makers and local decision makers to make dermatological morbidities a low priority.[ 7 ] Another concern is that the benefits of public health interventions in reducing the prevalence, morbidity and mortality of skin diseases may be underestimated.[ 8 ] Thus there is a need for more studies with respect to dermatological morbidities in a developing country like India. With this background, this study was carried out to find out the pattern and severity of skin disorders and to describe their association with various socio-demographic factors among medical students of a private medical college in Mangalore city of south India.

MATERIALS AND METHODS

This cross-sectional study was done in June 2011. The ethical approval for conducting this study was obtained from institutional ethics clearance committee. A sample size of 278 was determined using a confidence level of 95%, with 15% degree of precision of the expected proportion and an estimated minimum prevalence of 40%. These students were chosen from the 4 th , 6 th and 8 th semester through convenient sampling method so that the sample will have a balanced representation of 2 nd , 3 rd and final phase medical students of the institution.

The students were briefed about the objective of the study and written informed consent was taken for participation. A pre-tested self-administered semi-structured questionnaire was used for data collection. The face validity of this questionnaire was done by an expert in dermatology who reviewed the contents of the questionnaire. The questionnaire was subjected to a pilot trial on 10 students before it was distributed in its final form. Reliability of the questionnaire was assessed using Cronbach's Alpha the value of which was 0.82 indicating good internal consistency. Questions on the presence of any skin morbidities suffered by the student participants in the past 1 year were asked.

Additionally questions like frequency of face wash in a day, usage of facial cleansing products, frequency of head and body bath in a week, frequency of usage of hair shampoo in a week, usage of sunscreen lotions, moisturizers or cosmetics, frequency of changing into new clothes, habit of sharing linen with friends and promptness in seeking dermatologist consultation for skin ailments were asked to assess the quality of skin care.

Life style habits were assessed based on amount of water consumed in a day, frequency of eating fatty or oily food stuffs in a week, frequency of consumption of fruits and vegetables in a week, smoking habits and recreation habits like swimming.

Each response for the question meant to assess quality of skin care and life style habits were given scores from 0 to 2. Scores from 0 to 11 for questions deciding quality of skin care meant poor, 12-22 meant good level of skin care. Similarly scores from 0 to 5 for questions deciding life style meant poor and 6-10 meant good level of lifestyle habits.

The data entry and analysis were done using Statistical Package for Social Sciences software package (SPSS Inc., Chicago, IL) version 16. Chi-square test was used to find out the association of socio-demographic variables with the presence of skin morbidities, quality of skin care and life style habits P < 0.05 was taken as statistically significant association.

Mean age of participants was 20.35 ± 1.23 years [ Table 1 ].

Age, gender and place distribution of students

Of the 278 students, 69 (24.8%) had fair skin, 120 (43.2%) had wheatish skin, 74 (26.6%) had brown skin and 15 (5.4%) had dark skin. The one- year-period prevalence of various skin morbidities showed acne to be the commonest skin morbidity in 185 (66.5%) cases followed by sun tan in 147 (52.9%) cases. Among the hair morbidities commonest was hair loss seen in 165 (59.3%) cases followed by dandruff seen in 129 (46.4%) cases [ Table 2 ].

Association between various hair/skin morbidities among students with gender

Fungal infection was seen significantly among a greater proportion of males while among females the significant morbidities were hair loss, split end of hairs and dandruff [ Table 2 ].

Patterned baldness and sun tan were seen significantly more among greater proportion of non-Mangaloreans than native Mangaloreans [ Table 3 ].

Association between various hair/skin morbidities among students with place of origin ( n =278)

White/black heads were seen significantly more among females while papular and pustular types of acne were seen significantly more among a greater proportion of males. The proportion of cases with pustular type of acne was 30 (10.8%) [ Table 4 ].

Association between gender with type and duration of acne

Of the 278 students with morbidities, 236 (84.9%) had good quality skin care and the rest had poor quality skin care. 108 (83.1%) males and 128 (86.5%) females reported good quality skin care ( P = 0.428). Among the participants with good quality skin care, 161 (68.2%) reported presence of morbidities whereas among participants with poor quality skin care, 24 (57.1%) reported presence of morbidities ( P = 0.161).

Of the 278 students with morbidities, 236 (84.9%) had good life style habits and the rest had poor life style habits. One hundred and seven (82.3%) males and 129 (87.2%) females reported good life style habits ( P = 0.259). Among the participants with good life-style habits, 162 (68.6%) reported presence of morbidities, whereas among participants with poor life-style habits, 23 (54.8%) reported presence of morbidities ( P = 0.079). Out of 129 cases with history of dandruff, hair loss was present in 85 (65.9%) cases ( P = 0.039).

Usage of sunscreen in hot sun was associated with significant reduction in proportion of cases with sun tan among the participants [ Table 5 ].

Association between presence of sun tan with usage of sun screen among students

It has been found that one- fourth of us (or more) suffer from at least one skin disease, a situation that constitutes a significant global burden of disease.[ 9 ] Economic burden of skin diseases is enormous and added to this easy visibility of dermatological illness has led to deterioration in the quality of life resulting in social handicap.[ 10 , 11 ]

In certain parts of the world, it was observed that the mortality rate and disability-adjusted life years due to skin diseases were at par with certain communicable and non-communicable diseases.[ 7 ] In a regression model, skin diseases as well as rheumatism was more strongly associated with feeling depressed than asthma, diabetes and angina pectoris.[ 12 ] Considering their significant impact on the individual, the family, the social life of patients and their heavy economical burden, the public health importance of these diseases is underappreciated.[ 8 ] This study too has shown that various types of skin morbidities are common among medical students. It has been reported that younger adults suffer more social problems as a result of skin problems than older adults.[ 12 ] Thus control of skin morbidities will definitely lead to improvement in the quality of life of young adults. In this study the most common morbidity reported was acne followed by hair loss which was also supported by other studies.[ 3 , 13 ]

Acne has been incriminated with sweating and hot weather, which is very compatible with the hot and humid climatic conditions prevailing in Mangalore.[ 14 ] The proportion of severe acne cases in this study was 10.8% which was more than the observation of 5.4% made in the Sindh based study.[ 3 ] Studies carried out in other countries have found that acne is a disfiguring disease and it should not be looked at as trivial,[ 15 ] as it may seriously affect the patient's life.[ 16 ] Screening adolescents for conditions like acne may be of great importance because it affects their image in the society and because of the wide armamentarium of therapy which is available.[ 17 ]

Hair loss was the next most common problem, which is very much global in nature. The true magnitude of problem is difficult to establish from this study as the data on the hair density and thickness in our subjects was lacking. There was significant association of dandruff as a risk factor for hair loss in this study which was similar to the findings of other studies.[ 3 , 18 ] However, in the absence of any apparent systemic or local cause for generalized hair loss, it can be assumed that constitutional factors or micro-deficiency of iron, vitamins and proteins may be the cause of hair loss in these subjects.[ 19 , 20 ]

Hair loss culminating in baldness is another sensitive issue among adolescents as they are invariably sensitive regarding their external features and thus may be easily withdrawn psychologically and avoid social activities due to androgenetic alopecia and this tends to affect girls more than boys.[ 21 ] In this study almost a quarter of students had baldness with greater proportion observed among males.

Increased tanning of skin was the third most common morbidity. This was understandable as 68% of the participants had fair or wheatish skin. This skin type is prone to tanning on sun exposure. Being less aware of the tanning effect of sun light and not using personal protective measures while outdoors must have promoted tanning and darkening in these subjects.[ 22 ]

Fungal infections were reported by more than a third of our participants in the past 1 year. Previous studies have reported that periods of high humidity (50-80%) and elevated temperatures reaching up to 35°C are ideal for fungal infections.[ 17 ] This probably could explain the reason behind a number of cases with fungal infections among students in Mangalore.

In a study carried out among university students in Sindh, Pakistan acne was seen in 59.5%, hair loss in 59%, pigmentary disorders in 36.3%, dandruff in 26.1% and fungal infection in 4.9% of the cases. All these observations made were lower than our findings. The study also found pruritis among 2.3% of the cases and eczema among 2.1% of the cases.[ 3 ] In another study carried out among 1279 university medical students by Roodsari et al ., 91.7% students had skin morbidities. Here acne was seen in 56%, hair loss (evaluated only in females) in 14%, dandruff in 11%, hand eczema in 10%, seborrheic dermatitis in 9% and pityriasis versicolor in 8% cases.[ 13 ] But for acne which is easily identifiable, the other skin morbidities were higher in this study than ours probably because disease identification in the former study was done by dermatologists unlike our study where it was self-reported by students. An Icelandic study found that the prevalence of urticaria was significantly higher among the medical students and was seen in 41% of students.[ 23 ] These variations in morbidities among students of same age group in different parts of the world could be due to racial, genetic and environmental variations.

In this study acne was found to be slightly more and hair problems was seen significantly more among females, which was similar to the findings of a study done among university students in Lebanon where both acne and hair problems were significantly more among females.[ 17 ]

Although there was no significant difference between the proportion of males and females with acne in the present study, the type of acne differed significantly between the two groups. White/black heads were seen significantly more among females while papule and pustule were seen significantly more males. This was similar to the observation made in another study carried out in New Zealand where severe type of acne was seen more among males.[ 24 ] Severity of this condition among males could be because of hormonal factors.[ 25 ]

Fungal infection seen significantly more among males in this study could be due to their lesser quality of skin care and life style habits in comparison to females. Other cutaneous disorders like pyoderma, folliculitis, scabies and pediculosis were not seen in this study. The reason for absence of these bacterial and parasitic infections could probably be that very few participants in this study had poor quality of skin care or hygiene. No cases of eczema, hyper pigmentary lesions like melasma, hypopigmentary lesions like vitiligo, nail disorders or skin cancers were reported by any of the participants.

Sun tans were seen significantly more among a greater proportion of non-Mangaloreans than native Mangaloreans. This could probably be explained by the non-adjustment to the hot and humid conditions of Mangalore among the outstation students. It was also observed that the users of sunscreen had significantly less cases of sun tans compared to non-users, signifying the importance of spreading awareness about the usage of such protective methods.

Limitations

The present study may not be generalized to other population groups because of different factors associated with different skin morbidities. It may not reveal the true burden of skin disorders among young adults as much as a population-based study. Also as these morbidities were self-reported there may be a possibility of recall bias. In this study, quality of skin care was assessed based on frequency of activities like face wash or body bath or based on the frequency of usage of hair shampoo or sunscreen lotions or moisturizers or cosmetics. Since the quality of these activities or products as well as its proper application on the body was not enquired, it could be a limitation in estimating the true quality of skin care.

Moreover it was difficult to differentiate between the physiological and pathological conditions in hair loss. The most important drawback of this study was that few skin morbidities might have been diagnosed by medical students themselves without actually consulting a dermatologist leading to inaccurate self-reported diagnosis. Hence more of such studies from a broader socioeconomic spectrum are required, which need to be suitably supported with dermatological examination of study subjects.

From the findings of one- year- period prevalence of various skin disorders we conclude that skin morbidities are very common among medical students, particularly cosmetic problems like acne, hair loss and skin tan. Severe types of acne and fungal infections were significantly more among males whereas hair morbidities were significantly more among females. Patterned baldness and sun tans were seen significantly more among non-Mangalorean students than native Mangaloreans. This emphasizes the need to popularize the importance of personal protective measures like usage of sun screens among students. Establishment of registries for specific skin diseases, particularly for those with a high disease burden will also help in good case accountability stressing importance to dermatological public health.

ACKNOWLEDGMENTS

The authors of this study would like to thank M.B.B.S students, Ms. Monica N, Mr. Ishan Parashar, Ms. Hemashri, Ms. Supraja Subramanian, Ms. Liya Susan Peter, Ms. Anupriya Dalmiya and Ms. Akanksha Bansal of K.M.C Mangalore for their help in data collection. We also thank Dr. Mohan Kudur, Associate Professor, Department of Dermatology, Venereology and Leprology, Srinivas Institute of Medical sciences and Research Centre, Mangalore for his help and support.

Source of Support: Nil

Conflict of Interest: None declared

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 08 February 2024

Optimizing skin disease diagnosis: harnessing online community data with contrastive learning and clustering techniques

• Yue Shen 1   na1 ,
• Huanyu Li 2   na1 ,
• Can Sun 3 ,
• Hongtao Ji 4 ,
• Daojun Zhang 5 ,
• Yiqi Tang 2 ,
• Yu Chen 1 ,
• Zikun Wei   ORCID: orcid.org/0009-0000-7231-3696 2 &
• Junwei Lv   ORCID: orcid.org/0000-0001-7278-6864 2  

npj Digital Medicine volume  7 , Article number:  28 ( 2024 ) Cite this article

2032 Accesses

1 Citations

3 Altmetric

Metrics details

• Physical examination

Skin diseases pose significant challenges in China. Internet health forums offer a platform for millions of users to discuss skin diseases and share images for early intervention, leaving large amount of valuable dermatology images. However, data quality and annotation challenges limit the potential of these resources for developing diagnostic models. In this study, we proposed a deep-learning model that utilized unannotated dermatology images from diverse online sources. We adopted a contrastive learning approach to learn general representations from unlabeled images and fine-tuned the model on coarsely annotated images from Internet forums. Our model classified 22 common skin diseases. To improve annotation quality, we used a clustering method with a small set of standardized validation images. We tested the model on images collected by 33 experienced dermatologists from 15 tertiary hospitals and achieved a 45.05% top-1 accuracy, outperforming the published baseline model by 3%. Accuracy increased with additional validation images, reaching 49.64% with 50 images per category. Our model also demonstrated transferability to new tasks, such as detecting monkeypox, with a 61.76% top-1 accuracy using only 50 additional images in the training process. We also tested our model on benchmark datasets to show the generalization ability. Our findings highlight the potential of unannotated images from online forums for future dermatology applications and demonstrate the effectiveness of our model for early diagnosis and potential outbreak mitigation.

Similar content being viewed by others

Efficient diagnosis of psoriasis and lichen planus cutaneous diseases using deep learning approach

A patient-centric dataset of images and metadata for identifying melanomas using clinical context

A dataset of skin lesion images collected in Argentina for the evaluation of AI tools in this population

Introduction.

Skin diseases pose a significant challenge in China. There are up to 240 million dermatological visits per year, among which 80% are for skin diseases beyond melanoma. Nevertheless, the uneven distribution of healthcare resources and a shortage of dermatologists can lead to misdiagnosis and rising medical costs 1 . To improve the prognosis and reduce social costs, accurate and convenient diagnosis of skin diseases is critical. Artificial intelligence (AI) has shown great potential in dermatology due to the widespread use of photography in diagnosis 2 , 3 . However, most AI applications focus on benign and malignant lesion diagnosis 4 , leaving the potential of AI for a broader range of skin diseases largely unexplored.

Diagnosing skin diseases with machine learning methods 5 , 6 , 7 and its deep learning branch using convolutional neural networks (CNNs) 8 , 9 , 10 , 11 based on photographs has received much attention. While high-quality images are critical for training AI models, the labor costs associated with collecting these images can be prohibitively expensive. Fortunately, recent advancements in self-supervised contrastive learning offer a solution. These methods enable the pre-training of models using vast amounts of unlabeled or non-strictly labeled images and have shown outstanding performance in various tasks 12 , 13 , 14 , 15 . There are prior works exploring contrastive learning in dermatological diagnosis, emphasizing its capability to extract consistent representations and enhance generalizability and diagnostic accuracy 16 , 17 . For instance, FairDisCo 18 applied contrastive learning with additional network branches to enhance fairness across different ethnics. Ref. 19 introduced federated contrastive learning for dermatological disease diagnosis via on-device learning. Other studies have indicated that utilizing contrastive learning methods to combine multi-level features of skin lesion images can enhance the accuracy of diagnosing skin diseases 20 , 21 . However, most of these models are trained and tested on professional benchmark datasets. This can pose challenges when applying models trained on professional images to non-professional ones 22 , 23 . Benchmark datasets are typically captured in controlled medical research settings and focus on diseases with high medical significance, such as malignancies. Consequently, there’s a significant gap between these dataset’s distributions and the prevalence of common skin diseases in daily life. Also, the diversity in image capture settings restricts the generalizability of these models in society. To bridge this gap from the source data, the abundant unlabeled and coarse-labeled skin image data from online forums has come to our view. Traditionally, their unscreened and unannotated nature renders them unsuitable for traditional AI training, which calls for further exploration.

In this study, we present a deep-learning framework that leverages vast amounts of unannotated and coarse-labeled dermatology images from online sources. We employ a three-stage classification algorithm based on contrastive learning. In the pre-training stage, the model learns feature representations from unlabeled images. Our pre-trained model can be fine-tuned to downstream tasks with better performance compared with baseline models trained on general dataset. To reduce the effect of incorrect labels in the fine-tuning using Internet-sourced images, we propose a filtering approach using features extracted with unsupervised model and clustering approaches. This approach not only reduces training costs but also improves the model’s generalization ability for online diagnosis scenarios. Additionally, our method allows for easy fine-tuning on novel categories with limited standardized images, reducing data collection time and labor costs. We demonstrate this with an early warning system for monkeypox. Also, we have tested the performance of our model on benchmark datasets to show the generalize ability of our model when facing images from different ethnicities. In summary, our work reveals a new direction for dermatology AI research, leveraging unannotated and coarse-labeled internet-derived image data and contrastive learning to develop deep learning models for skin disease diagnosis. This approach has the potential to revolutionize dermatology, offering a more efficient and cost-effective method for diagnosing a wide range of skin diseases and ultimately improving patient outcomes.

Evaluation of pre-trained models

To evaluate the performance of our pre-trained model, we fine-tuned it using the entire coarse-labeled training set of 0.13 million images. To assess the efficacy of our model, we compared our results (denoted as ‘Derm’) with those obtained using a published pre-trained model trained with ImageNet (denoted as ‘ImageNet’). At the same time, we note that some work has pointed out that knowledge from ImageNet can speed up convergence, improve the generalization ability and performance when facing new problem domains 17 , 24 . Therefore, we designed two additional pre-training experiments to explore the role of ImageNet in understanding online dermatology images. The first experiment involved mixing the images from ImageNet and our dermatology images for pre-training (denoted as ‘ImageNet+Derm’). The second experiment pre-trained the model on dermatology images initialized with self-supervised ImageNet weights (denoted as ‘ImageNet→Derm’). All four pre-trained models were fine-tuned under the same setting with 0.13 million coarse-labeled images. The results are presented in Table 1 . Our top-1 diagnostic accuracy on the test set increased from 42.05% to 45.05% when pretrained sorely on dermatology images, indicating a notable improvement in performance. These findings suggest that unlabeled skin disease data available on the Internet, even without standardized sampling and labeling processes, holds great potential in the field of skin disease diagnosis. It is noteworthy that simply combining ImageNet and dermatology image data for pre-training shows only a marginal increase in top-1 accuracy for dermatology classification. Initializing with ImageNet model weights brings greater gains than simply mixing the two datasets. However, it should also be noted that these pre-training approaches incur a greater training cost than solely working on the dermatology dataset or the ImageNet dataset. The increase in training cost primarily arises from the dataset expansion. We conducted pre-training for each method using 64 RTX 3090 GPUs. In the ‘ImageNet+Derm’ configuration, the training duration (approximately 12 h per 100 epochs) almost doubles compared to the ‘Derm’ setup (approximately 5 h and 45 min per 100 epochs). While in this study, we utilized publicly available models trained on ImageNet, resulting in a comparable computational cost for ‘ImageNet→Derm’ as with the ‘Derm’ setup, it’s crucial not to overlook the additional time required to acquire model weights trained on ImageNet if different network architectures were to be employed. We attribute the modest gains from the ‘ImageNet→Derm’ approach to catastrophic forgetting, where the model loses previously acquired knowledge from ImageNet when exposed to new, unlabeled data. Additionally, the high prevalence of label noise within internet-sourced dermatology images likely hinders the model’s ability to learn accurate representations in fine-tuning without selective filtering.

Effect of filtering coarse-labeled data

We acknowledge the challenge posed by noisy labeling in these coarse-labeled images during the fine-tuning stage. To address this, we filtered the training set using a validation set of 20 images per disease based on feature distance obtained by pre-trained model. This approach reduced the number of training images from 0.13 million to approximately 30000, but our model’s top-1 diagnostic accuracy improved from 45.05% to 46.61%, and top-3 accuracy increased from 65.13% to 68.48%, which also surpass the gains brought by pre-training on models initialized with ImageNet model weights, indicating the necessity of our filtering methodology. These findings indicate that using a larger validation set to obtain a more comprehensive description of the clusters per disease may lead to more effective filtering results.

To develop a flexible model adaptable to various diseases, reducing the amount of labeled data can significantly decrease training time and costs. However, using too little data may not effectively capture the feature clusters of a disease. To further explore this issue, we randomly selected subsets of 20, 30, 40, 50, 60, 70, and 80 images from the 80 validation images collected for each disease to examine the impact of the size of the validation set on filtering the coarse-labeled training data and the model’s performance. To ensure test reproducibility, we conducted three trials using different random seeds to select subsets of the validation dataset and fine-tune the model. The final top-k diagnostic accuracy is presented in Fig. 1a . The average top-1 accuracy after filtering images based on 20, 30, 40, and 50 validation samples over the three trials was 46.61%, 47.77%, 48.32%, and 49.64% respectively, indicating significant improvement compared to the baseline of 42.05%. The ROC curve, as Fig. 1b shows, also indicates an improvement in the performance when the validation samples increased. By ANOVA, we are unable to statistically consider the data from the three trials to be significantly different ( p  = 0.77), while statistically indicating that the model performance is significantly higher than the baseline of 42.05% (all p -values much less than 0.01) as shown in Fig. 1c . Furthermore, performance improved gradually as the number of validation samples increased, likely due to a more precise description of each cluster center with a larger validation set. However, the improvement in subsequent models was relatively low when the number of validation samples exceeded 50. As shown in the Fig. 1d , the average top-1 accuracy of filtering images based on 60,70,80 validation samples were 49.61%, 49.77%, and 49.79% respectively, while the rest of the top-k accuracy also remained almost the same. We also used our proposed filtering approach with 50 validation images per disease on top of the ‘ImageNet->Derm’ pretrained model and achieved an average top-1 accuracy of 50.44%. Compared with fine-tuning with the whole coarse-labeled training set (46.13%), our filtering approach gave a 4.31% increase, which further proved the effectiveness of our filtering strategy across different pre-training baselines. We counted the average number of images per category after filtering with different number of validation images as Fig. 1e shows. Generally, the number of remaining training sets did not vary too much, especially when the number of validation images reached 50 per category. Intuitively, we think that the estimated cluster centers differ more from the actual cluster centers when there are fewer validation images, thus causing greater bias when filtering the training set by Euclidean distance. Therefore, the estimated center of each cluster tends to be stable with more validation images.

a , b show the top-k diagnosis accuracy and ROC curve of our model. We pre-trained our model using unannotated images collected from the Internet and then fine-tuned it on the full coarse labeled training set. Our top-1 diagnostic accuracy on the test set increased from 42.05% to 45.05% and the AUC of the ROC curve increased from 0.859 to 0.872. After filtering potential noisy labels using validation images, the performance improved as the number of validation images increased. When there were 50 validation images per category, the top-1 accuracy reached 49.64%. c Boxplot showing the performance of three trials using different subset of validation images. Boxes represent the median costs and interquartile range. Whiskers extend to the farthest data points. ANOVA analysis showed that our model’s performance was significantly better than the baseline, and that different validation sets used for filtering did not produce statistically significant differences ( p  = 0.77). d Top-k diagnosis accuracy improvement of our model saturates when the number of validation images reaches 50 per category, suggesting that 50 validation images per category are sufficient for the filtering process. e Number of images after filtering averaged over three trials did not vary too much when changing the number of validation images, especially when the number of validation images reached 50 per category, indicating the estimated center of each cluster tends to be stable with more validation images.

To illustrate the effectiveness of our filtering method, we used t-SNE to generate a scatter plot of the cluster distribution of the remaining training set after filtering with 50 validation images per category. We also randomly sampled the same number of images from the original training set to draw a scatter plot for comparison, as shown in Fig. 2a . While t-SNE may cause some deformation in the appearance and distances of clusters, it still allows for a rough idea of the relative position and coverage of each disease in the feature space. Our results indicate that the selected training set after clustering and filtering using features obtained by the pre-trained model displays clearer boundaries for each cluster, and the relative location of each cluster corresponds to dermatologists’ knowledge. For example, in the upper right corner of the scatter plot, two isolated clusters representing androgenic alopecia and alopecia areata can be observed, which are similar yet distinct from other diseases. The discernible clustering of diverse skin conditions indicates that the features employed in our analysis can capture distinct attributes that are relevant to each disorder. In Fig. 2b we mapped the average top-1 specificity and sensitivity of each disease to the scatter plot. All the specificity were over 0.94, but the sensitivities displayed considerable variation. Generally, diseases with clusters relatively far from the center of other diseases and with fewer surrounding clusters exhibited higher sensitivity. For instance, the sensitivity of androgenetic alopecia, alopecia areata, acne, and melasma were 0.87, 0.82, 0.85, and 0.77, respectively. These diseases are empirically more typical and are easier for the physician to diagnose based on the image alone. Conversely, diseases with clusters closer to the overall center of other diseases and with more surrounding clusters demonstrated poorer sensitivity. For example, lupus erythematosus and eczema dermatitis only got 0.04 and 0.20. These diseases often lack typical lesion characteristics, and the physicians also require additional information to make accurate diagnoses. We present the expression levels of certain host features in Fig. 2c to provide a better understanding of the selected training set. The spatial distribution of selected features highlights the likelihood that these characteristics are linked to specific types of skin diseases and affected areas of the skin. When we randomly selected several images from both the filtered training set and the excluded images and mapped them onto the scatter plots, it can be observed that the retained images generally exhibit typical skin lesion characteristics of their respective diseases, while the excluded images tend to be farther from the cluster centers and are mostly identifiable as label errors. It is important to note, however, that there are three scenarios in which images may have been excluded from the training set. Firstly, images containing more than one skin disease, with the coarse label failing to become the primary focus. Secondly, atypical skin lesions, such as alopecia areata on the eyebrows, also have a high chance of being far away from the typical cluster. And thirdly, skin diseases under treatment, where recovery or medication will also change the appearance of lesions. These exclusions could potentially lead to a decreased recognition capability of our model for atypical skin lesions, even though they only account for a small proportion of the collected skin disease images. While it is generally believed that more labeled data leads to better model performance, our experiments demonstrate that images with correct knowledge and distinct features are more likely to help the model learn diagnostic criteria than a large amount of data with ambiguous or incorrect labels.

a t-SNE plots of the filtered training set with 50 validation images per category colored by disease categories, demonstrating the effectiveness of our approach in capturing the distinct features of each disease. For comparison, a subset of images of the same number is randomly selected from the coarse labeled training set. b Average sensitivity and specificity for each disease of 3 trials. Specificity of each category are all over 0.94 but sensitivity varied considerably. Generally, sensitivity would be higher if the cluster was relatively far from the center of other diseases with fewer surrounding clusters. c Expression levels of some host features. The levels of these characteristics are associated with the types of skin diseases and affected areas of the skin.

Transfer learning to monkeypox detection

Figure 1d may suggest that for a new disease, 50 cases may be needed to describe the clusters based on the pre-trained contrastive model provided. Besides, at the top-10 confidence level, the diagnosis accuracy already reached 90%. Therefore, adapting our model to an early warning system for a new disease may no longer require a large amount of image data specific to that disease, resulting in a significant improvement in both training speed and overall cost. We expect the model to identify potential risk images from a large number of uploaded data by internet users and provide early warning signals for emerging diseases. We demonstrate this in the experiment of monkeypox, a rare disease primarily affecting dark-skinned individuals. We fine-tuned our model on 50 monkeypox images added to the selected coarse labeled training set and tested it on a mixed set of 2146 common skin disease images and 170 monkeypox images collected from the internet. For images with low diagnostic confidence, we classified them as ‘others’. In order to ensure suspected patients receive proper diagnostic evaluation, it is important to have a low false negative rate in a primary hospital or internet scenario. Therefore, including a high ratio of monkeypox images in the warning system while blocking unrelated diseases is crucial. Relevant results are shown in Fig. 3 . We projected the image features of monkeypox onto the feature representation map of the previously analyzed 22 skin diseases as Fig. 3a shows. Due to the limited amount of image data included, the resulting cluster appears relatively loose. Nevertheless, it is evident that, apart from a few cases that may be attributed to mislabeled images or distinct location-specific features (such as skin lesions near hair), most monkeypox image projections are located near viral warts and tinea. This observation aligns with our understanding of the characteristics associated with monkeypox. The ROC curve for monkeypox is shown in Fig. 3b , with an AUC of 0.982. We counted the images that were diagnosed as monkeypox along with the confidence level as Fig. 3c shows. Our experiments show that when the confidence level is ranked in the top 10, 95.29% of monkeypox images are diagnosed, with suspected images accounting for only 6.99% of the total images, the confusion matrix of which is shown in Fig. 3d . The sensitivity was 0.953 and specificity was 0.997. Furthermore, our model detected 61.76% of monkeypox images at top-1 level, which surpasses 15 of the 22 regular skin diseases. This suggests that the model has promising performance in identifying rare and emerging diseases while utilizing a small volume of training data. In addition, we observed that images classified as ‘others’ in our experiments, mostly consisted of skin diseases that were not included in the training process like fungal skin infections. This may be due to the highly distinctive nature of these skin lesions, which neither resemble typical monkeypox lesions nor bear similarities to any of the 22 common skin diseases analyzed.

a Scatter of monkeypox images. most monkeypox image projections are located near viral warts and tinea. b ROC curve of the fine-tuned model. We fine-tuned the model with the filtered training set and 50 extra monkeypox images to act as a warning system for monkeypox, testing on 170 monkeypox images and 2146 skin disease-related images. c Performance of the monkeypox warning system at each confidence level. Mp refers to monkeypox images. Our model detected 61.76% of monkeypox images at top-1 level. At the top-10 confidence level, 95.29% of monkeypox images were successfully diagnosed, with highly suspected images accounting for only 6.99% of the total images. d Confusion matrix at the top-10 confidence level (Sensitivity = 0.953, Specificity = 0.997).

Development of online diagnosis app

We developed a prototype of an online dermatology diagnosis system implemented as a WeChat-based app named ‘Huifu’ following the JAMA CLEAR dermatology guidelines 25 , which could be used on smartphones, as shown in Fig. 4 . We have included a completed CLEAR Derm checklist in Supplementary Table 1 to show our adherence to the guidelines. This app enables patients to receive diagnostic advice from our model by directly taking and uploading images from their smartphones. After collecting basic information from patients to build a health record, Huifu combined a survey and picture-taking of a skin lesion. The process begins by inquiring about the location of the lesions and their symmetry. Based on this information, the software guides the patient through different image collection protocols. We employ a pre-processing module to assess image clarity, lighting conditions, and camera distance to ensure that the images meet the collection standards. Subsequently, our model detects and segments the skin lesion. If monkeypox does not rank within the top 10 confidence levels, we proceed to request additional basic information related to the skin disease from the patient, such as whether it is accompanied by itching, pain, etc., to provide a more accurate diagnosis. However, if monkeypox does appear within the top 10 confidence levels, it will be considered a high-risk case. In such instances, we will pose specific questions, such as ‘Have you recently been in an area where monkeypox is prevalent?’ For cases with a high suspicion of monkeypox, a doctor will be assigned for a possible online consultation.

Starting from the first line, patients information is collected to build a personal health record. Then, images of the lesion area are required to be uploaded. Our model is used to give a diagnosis based on the image uploaded. We use a follow-up system to further help improve diagnosis accuracy. Finally, the recommended diagnosis is presented.

At current stage, our app has been served as a decision support system for clinicians to offer valuable suggestions and enhance the efficiency of doctors’ work. Since November 2021, we have been piloting the app with physicians in 18 tertiary hospitals, and it has assisted 186 doctors in making more informed diagnostic decisions. Lesion images are taken using smartphones, either by physicians or under their guidance, whereupon the images are uploaded for AI analysis. The final diagnostic judgment, however, rests with the physicians. We monitored backend usage data and requested feedback from doctors on the consistency of the model’s results with their own judgment. As shown in Supplementary Fig. 1 , backend data was used to illustrate the consistency rate of top-2 diagnostic suggestions with physicians’ opinions, and the average using time during our test respectively. Based on our records up to November 2022, our app has been utilized in 26,676 patient encounters, with 21,288 completing the full diagnostic process. The average usage time per encounter stood at 107 s. Notably, the adoption rate of the app’s top-1 diagnosis by doctors is 63.04%. This is a significant metric for us, indicating that patients have achieved results at a faster pace than what would typically require more time from a doctor. It reflects the clinical utility and acceptance of our AI-driven diagnoses. In addition, the encouraging diagnostic performance, especially for emerging diseases such as monkeypox, has indicated the potential to collaborate with public health authorities to assist in internet-based screening efforts for such conditions.

Performance on benchmark datasets

Similar to the case of monkeypox, our model holds promise for early malignancy detection, potentially leading to significant improvements in patient outcomes. However, in Chinese online forums there are few images related to malignancies (as discussed in Supplementary Information ), potentially due to the relatively low incidence rate and limited awareness among Chinese Internet users 26 . Besides, it is also difficult to collect standard validation or test malignant images from our collaboration hospitals. Therefore, in this part we consider two benchmark datasets including related data: Fitzpatrick17k 27 and diverse dermatology images (DDI) 28 datasets. At the same time, we notice that our model was initially trained and tested on photos predominantly featuring East Asian individuals, potentially introducing some bias. These two benchmark datasets, gathered from countries beyond China, allow us to assess our model’s generalization across diverse races and skin tones.

We began by consolidating specific labels from benchmark datasets to align with our predefined labels. Given our primary focus on detecting malignant diseases, particularly melanoma as the most severe form of skin cancer, we isolated melanoma from the broader malignancy category, establishing two distinct classes: malignancies and melanoma. Consequently, we formulated a classifier encompassing 24 classes. Images that didn’t match these 24 labels were categorized as ‘others’. Supplementary Table 2 contains detailed information on the label merging rules and dataset sizes. Due to the inadequacy of the DDI dataset to form a new class for training, we opted to combine the two datasets together for this experiment. Classes with fewer than 10 images in each dataset are not considered, as such a small sample size can significantly skew the testing performance.

Firstly, we fine-tuned our model incorporating 50 images each from melanoma and malignancies into the filtered coarse labeled training set. Of these, 10 were sourced from DDI, and 40 from Fitzpatrick17k. Then we tested it on the remaining images, detailing the top-1 diagnosis accuracy for the classes in Table 2 . Note that due to the dataset imbalance between the two benchmarks, our model demonstrated a tendency to perform better on Fitzpatrick17k. Although achieving a top-1 accuracy of 49.47% on our test set and showcasing a degree of general diagnostic capability on the benchmark dataset, the overall performance was underwhelming. The two newly introduced malignant classes exhibited relatively high diagnostic accuracy (12.50% and 16.79% on DDI, and 55.26% and 43.22% on Fitzpatrick17k). We attribute this disparity mainly to our model’s training on predominantly East Asian images, lacking exposure to diverse skin tones or races, while benchmark datasets are mainly non-East Asian images.

To address the knowledge gap within the model, we augmented the filtered training set with 20 images from each category, excluding malignancies and melanoma, sourced from the benchmark dataset. Due to data volume imbalances, we specifically added images from classes containing more than 50 images to ensure adequate numbers for testing. The diagnosis accuracy for the remaining benchmark images is detailed in Table 2 . Remarkable enhancements were observed across nearly all classes. Overall performance on the DDI dataset increased from 19.45% to 33.69%, and on the Fitzpatrick dataset, it surged from 24.95% to 35.57%.

We attach the subgroup performance on both datasets in Supplementary Tables 3 and 4 . On the Fitzpatrick17k dataset, we noted a gradual improvement in the model’s performance as skin tones deepened. Initially, the model achieved an overall performance of 33.13% for Fitzpatrick Scale I and 37.46% for Fitzpatrick Scale VI. We attributed this outcome to differences in the proportion and diagnostic accuracy of ‘others’ across varied skin tones. After excluding this class, the model’s overall performance elevated to 41.28%, demonstrating consistent performance across different skin tones (38.85%, 40.60%, 40.80%, 44.03%, 46.70% and 40.86% for Fitzpatrick scale I to VI). This improved performance across various skin tones suggests that our pre-trained model has captured key dermatological features, showcasing robust generalization ability. Additionally, its adaptability to malignant skin diseases underscores the potential for enhancing early intervention possibilities.

Specifically, the performance in identifying malignancies improved even without additional training data. On the Fitzpatrick dataset, the model’s accuracy in detecting melanoma reached 59.96% (sensitivity=0.59), matching other deep learning studies’ performance on the Fitzpatrick dataset 29 . The accuracy in identifying malignancies increased to 48.86%. This lower performance in malignancy identification might be because melanoma, as a specific disease, exhibits more distinct and concentrated features, while other malignancies present varied appearances. However, as malignancies is still quite a different class from the common diseases with our initial 22 classes, the performance is still outstanding considering the small amount of data needed, which indicates the potential in the early intervention of malignancies.

Additionally, we observed a significant decline in diagnosing acne compared to our test set. Upon image review, we believe this might be due to substantial morphological differences in acne across diverse skin tones and ethnicities, creating a gap that cannot be merely addressed by grouping it as the same ‘acne’ as present in our dataset. Similar to melanoma, if we extract 50 cases of acne from benchmark datasets as a new category, our model’s diagnostic accuracy on these acne images reaches 67.95%. Unfortunately, due to limitations in the quantity of benchmark datasets, it is challenging to further increase the number of images in the training set to explore the model’s diagnostic potential. However, our results emphasize that by considering the substantial disparities between the benchmark dataset and our training data, our model can be easily adapted to new downstream tasks with small number of training data.

Bias discussion

Pre-training on images sourced from the Internet can carry and amplify social biases. In this section, we will provide an initial analysis of the biases present in our model. We address this issue from two perspectives: the data distribution in the training set and the performance across subgroups in the test set.

While naturally diverse, gathering metadata from the online images used to train the model is often challenging. To handle this, we conducted random sampling and manual labeling of the training set to showcase potential biases within our training dataset. We sampled 500 images from both the filtered and discarded coarse-labeled training sets, having trained dermatologists identify gender, age, and lesion area information in the corresponding images. Since labeling skin tones without professional training is difficult, we followed 27 to compute the Individual Typology Angle (ITA) with YCbCr masks for skin tone estimation and derived Fitzpatrick scales. The unannotated data used for pre-training mirrors the unfiltered coarse-labeled dataset, a blend of these two subsets. The distribution of the sampled images is presented in Table 3 . Remarkably, in the filtered image set, the percentage of ‘unknown’ for all subgroups, except for skin tone, was notably lower compared to the discarded image set. The proportion of unknown gender decreased from 47.80% to 29.34%, and unknown age reduced from 63.60% to 35.93%. In the filtered dataset, female account for the majority of images with explicit gender information, which may be because women are more active in seeking advice for skin diseases on online forums. The dataset predominantly concentrates on lighter skin tones, in line with the skin color distribution among East Asians. Additionally, compared to the discarded images, the proportions of the very light skin tone (I) and the dark skin tone (VI) decreased in the filtered images, indicating our filtering possibly eliminated images taken in extreme lighting conditions or through camera filters. The decrease in images with unknown information or under extreme lighting conditions partially indicates that our filtering approach retained images with more diagnostic information, potentially contributing to our model’s enhanced diagnostic performance.

As we collected patient metadata along with the validation and test set images from offline hospitals, we showcase subgroup performance on our test set in Supplementary Table 5 . Specifically, we compared the performance of three models under distinct settings. Model A: pretrained on ImageNet dataset and finetuned with whole coarse-labeled set. Model B: pretrained on the dermatology dataset and finetuned with whole coarse-labeled set. Model C: pretrained on the dermatology dataset and finetuned with filtered coarse-labeled set with 50 validation images.

Notably, Model C outperforms both Model A and Model B. In subgroup analysis by lesion area, Model C significantly improved the diagnosis of conditions impacting the upper extremities, hands, lower extremities, and feet compared to Models A and B, while the improvement for head/neck and torso conditions was moderate. Across genders, all three models exhibited slightly better performance for females. The performance difference between males and females in the three models is 3.33%, 4.57%, and 3.72%, respectively. This indicates that Model C, without amplifying gender differences, exhibited an overall performance increase, emphasizing the filtering strategy’s capacity to generalize and provide balanced performance across genders.

Regarding skin tones, although a small amount of test images was categorized as Fitzpatrick scale V or VI based on their ITA scores, our test set comprises typical images of individuals from the Chinese population. Hence, rather than evaluating darker skin tone subgroup performance, we’re presenting the model’s performance facing East Asian skin that looks darker. We observed that Model C displayed relatively higher accuracy for skin types commonly found in East Asian populations, where the bias was most pronounced across all three models. Model C shows good consistency across Fitzpatrick scales II to V (52.58%, 49.40%, IV 57.98%, and V 51.98%, respectively). However, accuracy for the lightest (I) and darkest (VI) skin tones is comparatively lower, at 48.66% and 42.62%, respectively. Nonetheless, compared to Model A, the accuracy for skin classification in types I and VI is still higher by 6.87% and 4.05%, respectively. This might indicate that due to the uniqueness of the Chinese internet, the model’s diagnostic abilities for skin diseases among populations with the lightest and darkest skin tones are somewhat limited. It’s crucial to note that the difference in performance across various skin tone subgroups primarily stems from the inherent characteristics of the Chinese internet itself rather than from model training. Most internet users contributing to the dataset are East Asian residing with minimal immigration diversity in urban areas of China. While there are ethnic minorities in China with darker or lighter skin tones, these populations are primarily concentrated in border regions, resulting in lower representation within the dataset. Consequently, the dataset tends to represent individuals predominantly falling within Fitzpatrick skin types II to V, with significantly fewer representations of skin types I or VI. Our experiments on benchmark datasets further confirm this.

In conclusion, our study has revealed the value of unlabeled data on the internet for learning dermatological diagnoses using contrastive learning to pre-train a model. Although the improvement in diagnostic performance for 22 common skin diseases after fine-tuning was only 3% compared with the baseline model pre-trained on ImageNet, this should not be interpreted as a lack of value in the data from the Internet. The main reason for the limited improvement is that there are many labeling errors in the coarse annotations of data used during fine-tuning, which can greatly affect the model’s performance if not addressed properly. Therefore, we have established a training framework that uses the validation images as calibration to clean the coarse-labeled training set and maximize the value of the Internet data, resulting in a further significant improvement in the diagnostic performance of the model by a maximum of 4.6%, while reducing the computational costs required. It should be noticed that, although integrating ImageNet with unlabeled skin images offers a slight improvement in the pre-training stage, our filtering strategy can offer more performance gains, and emphasizes the necessity of a filtering approach when dealing with noisy datasets from the Internet. We demonstrate the ability of our model to transfer to downstream tasks with emerging diseases through the experiment of monkeypox. We also tested our pre-trained model on benchmark datasets to supplement missing out-of-distribution results in our test set and to verify our model’s generalization across malignant diseases and skin tones of different ethnicities. Our model can also distinguish malignancies with small amount of training data added. Due to utilizing data from the Chinese Internet, our model was trained predominantly on images of Asian individuals. While its initial performance on benchmark datasets primarily featuring other ethnicities may not be impressive, with the addition of a small number of images for guidance, it demonstrates excellent generalization on these new datasets as well. These results suggest that our framework has the potential for widespread use in the diagnosis of other skin diseases, particularly those for which labeled data is scarce.

Based on the prevalence of skin diseases and the pressure on medical resources, developing an effective AI-assisted diagnosis system for dermatological diseases can have significant value. Such a system can provide primary dermatologists and general practitioners with diagnostic expertise that is equivalent to that of top dermatologists, thereby bridging the gap between primary and advanced dermatology. A cross-sectional study 30 has demonstrated that most dermatologists are willing to adopt AI tools to enhance time efficiency, diagnostic accuracy, and patient management. Additionally, for ordinary patients, online forums provide a platform to discuss their health concerns, including skin diseases. Using appropriate AI tools based on images allows patients to detect and treat potential diseases early. Timely diagnosis and treatment of rare and infectious skin diseases have significant clinical value, as they can encourage dermatologists to initiate appropriate treatment plans, improve patient experiences, reduce the risk of long-term sequelae, and reduce the incidence and mortality rates associated with severe skin adverse reactions or invasive skin cancer. At the macro level, it will contribute to the optimal utilization of medical resources, including targeted treatment and appropriate referrals to specialist physicians. This can alleviate the pressure on the healthcare system and minimize the waste of medical resources. Additionally, AI diagnostic information can be more directly and systematically integrated into other systems, providing information for public health interventions, policymaking, and resource allocation.

However, online skin disease diagnosis will face complex skin disease classification tasks with multiple disease subtypes and complex pathogenesis. Traditional supervised methods require a large amount of annotated data and may even involve human evaluation in the training process 31 , which is not feasible in the Internet context. Meanwhile, contrastive learning offers a potent tool for the automated diagnosis of skin diseases. Past uses of contrastive learning in the field of dermatology primarily involve structured and high-quality but limited-scale benchmark datasets. These datasets, geared for medical research, often concentrate on diseases with higher medical value and acquire images under stringent uniform capture requirements. Despite significant progress achieved by previous models in these specialized domains, the challenge arises for daily use to capture images that meet these stringent requirements and use these models 19 , 20 , 21 . For example, capturing images akin to the ISIC dataset using mobile devices can be quite challenging. Furthermore, these specialized datasets limit the types of diseases these models can cover, while skin diseases in daily life often exhibit a ‘long-tail’ distribution, demanding higher levels of coverage and generalization from models 32 , 33 .

Unlike the work in ref. 34 that improved the loss function to enhance the learning capability of contrastive learning for out-of-distribution (OOD) data representations, our work focuses on the effective utilization of web data right from the data source. These data sources often possess quality issues, making the research quite challenging. However, the inherent diversity of the internet data aligns better with regular people’s everyday scenarios, broadening the scope of our downstream tasks to support a more diverse set of common diseases than traditional contrastive learning. Through our filtering approach, our model achieved a significant improvement in diagnostic accuracy, demonstrating its effectiveness. Furthermore, experiments on benchmark datasets indicated the remarkable generalization ability of our model, swiftly adapting to and handling unknown data. In practical application, our research covers not only common skin diseases but also emerging ones, as showcased by our successful application of the model for early detection of monkeypox. Merely by incorporating an additional 50 images of monkeypox into the training process, our model attained an impressive 61.76% top-1 accuracy, showcasing its outstanding adaptability to emerging and evolving health challenges.

Considering that physicians have limited opportunities to diagnose based solely on image information, it would be unfair to directly compare the performance in this scenario. However, it is possible to assess the value of our work by comparing the results with similar studies. A previous study 3 reported diagnostic accuracy rates of dermatologists, primary care physicians (PCPs), and nurse practitioners (NPs) as 63%, 44%, and 40%, respectively. The performance of our model is approximately comparable to that of PCPs. It is worth noting that the dermatology images sourced from the Internet encompass not only common skin diseases. While our target diseases share a substantial similarity with the aforementioned study, they also include less commonly seen diseases such as systemic lupus erythematosus, lichen planus, and blue nevus, with diagnosis accuracies below 40%. These diseases are more challenging to diagnose or resemble malignant skin diseases, often receiving more attention and discussion on the internet. If we solely consider skin diseases included in the aforementioned study, our average diagnosis accuracy reaches 52.57%. On one hand, the existence of images depicting rare or atypical skin diseases underscores the value of internet skin image data. On the other hand, it signifies the potential for monitoring rare or emerging epidemic skin diseases on the internet. Although the experiments indicate that due to data biases, our model exhibits higher performance among East Asian populations, given the severe shortage of dermatologists and PCPs in China, our ‘Huifu’ software holds substantial value. Our app aligns with the current shift towards patient-centered diagnostic approaches. Smartphone cameras evidently provide a convenient and swift means of dermatological consultation for a larger number of patients. This work showcases the potential of online health forums as valuable resources for medical research and the development of AI-driven diagnostic tools, paving the way for more inclusive and accurate healthcare solutions.

AI-assisted diagnosis of skin diseases has immense potential for further advancement, especially with recent breakthroughs in multimodal language models like ChatGPT 35 . These models possess the capability to process both image and text information, enabling quick access to accurate information about skin diseases for patients and healthcare providers, along with personalized responses to users’ inquiries and descriptions. Furthermore, the integration of text-based patient symptoms with image-based skin lesions, gathered from internet forums and other user data sources, can further enhance diagnostic accuracy. However, it is important to note that dermatological conditions often encompass numerous atypical cases. Relying solely on images for a preliminary diagnosis and then mechanically asking questions to differentiate them from common diagnoses would be time-consuming and may not yield accurate results. Building upon our work, it may be possible to improve the efficiency of consultations by employing a common language model that can better target differential diagnoses sharing typical lesion features. This approach could optimize the efficiency of consultations. For skin lesions located far away from clusters of other diseases, the reliability of the diagnosis would increase, requiring only a few follow-up questions, thereby further optimizing efficiency. With the advent of Internet hospitals, these models can even identify similarities between patients with similar symptoms or conditions. Thus, incorporating internet data can lead to more effective and efficient development of skin disease diagnostic models. As AI technologies continue to advance, we can expect even more exciting possibilities for AI-assisted diagnosis of skin diseases in the future.

The overview of our framework is shown in Fig. 5 .

Our approach is designed for online dermatological diagnosis scenarios and makes full use of the image data from the Internet. Starting from the top, firstly, images were collected from various Internet resources. Several screenings were implemented to pre-process the collected images. Besides 1.18 million images without annotations, 0.13 million images with coarse labels were obtained by matching keywords from topics and merged based on published standards. Secondly, a model was pre-trained on the unlabeled images using contrastive learning approaches, and it acted as a feature extractor later. Thirdly, features were extracted from both the coarse-labeled training set and the validation set. Clustering and filtering were performed to discard potential incorrect labels. Finally, the filtered training set was used to fine-tune the model, and a self-adaptive threshold was adopted to handle out-of-distribution images. The approach also allows a small number of images from new categories to be added for transfer to new downstream tasks, and it shows good generalization ability.

Data description

Significant progress has been made in the past few years based on large-scale annotated image datasets in object classification. The ImageNet Large Scale Visual Recognition Challenge (ILSVRC) 36 has emerged as a central testbed for object classification research and has showcased landmark achievements in machine learning. For our experiment, we referred to the number of images used in the pre-training and fine-tuning of popular unsupervised learning methods trained on ImageNet and collected data from various sources on the Internet. The Medical Ethics Committee of the Third Affiliated Hospital of CQMU provided ethical review and approval for this study. Collaborating doctors obtained additional written informed consent using EC-approved forms from patients during the collection of validation and test images for the study. Our data collection process through the ‘Huifu’ app adhered to the user and content security solutions provided by the WeChat Mini Program, ensuring in-app informed consent was secured via a checkbox when gathering backend data.

The training images are collected from search engines, public forums, online doctor consultation platforms, dermatologists’ personal channels and other publicly available resources in China. We conducted searches based on specific keywords, allowing for a maximum of 3000 images up to date per search. This collection process was completed by the middle of 2021. Initially, we collected over 3 million images related to skin disease. To balance categories, we removed some samples from keyword subsets with sufficient image representation.

In the pre-process stage, to clean the data, we go through several screenings. We first apply a skin segmentation module using a U-Net 37 to filter images without enough skin. We follow the experiment in ref. 38 and set the threshold as 0.75. We also drop images that are too small (less then 224 × 224), have extreme aspect ratio (over 3:1) or of low-quality as in ref. 39 . At last, we use center-crop and resize all the images into 512 × 512 to create the training set of 1.18 million dermatosis-related skin images without annotations.

Note that annotation is unnecessary in the unsupervised pre-training phase of the model. However, labels are indispensable for fine-tuning it as a classifier. The skin disease images were annotated using two methods. Firstly, skin diseases or related terminologies were used as coarse labels for the images in self-organized communities and forums related to skin diseases in China. Secondly, skin disease images were directly searched on search engines, and databases with coarse labels created on the web were incorporated. To standardize and merge certain coarse labels, we relied on physician experience and dataset characteristics. We adopted industry-standard specifications such as CDISC 40 and ICD10 41 to transform large sections of unstructured text into standardized data through natural language processing techniques. For example, ‘whelks’, ‘pimples’, and ‘comedo’ were combined into the general category ‘acne’. We also merged labels that contained skin diseases that could not be reliably distinguished without consultation, such as ‘viral warts’ and those that contained only partial images that did not allow for identification of the lesion site, such as ‘tinea manuum’ and ‘tinea pedis’ into the broader category of ‘tinea’. Finally, we selected the 22 most representative skin diseases of the highest proportion, each with at least 800 images, for model construction, as listed in Table 4 . The details of selecting these 22 diseases can be seen in Supplementary Information . In total, we collected 0.13 million dermatology-related skin images with coarse label information for fine-tuning.

To test the performance of our model, extra image dataset consisting of high-quality skin lesion images was obtained under standardized conditions and filtered to ensure quality to be used as validation and test set. Our validation and test set were gathered through collaboration with doctors in offline hospitals. Physicians physically examined the patients, assessed skin lesions, conducted interviews, and performed essential pathological tests to ascertain diagnostic results. After that, physicians captured the images using a smartphone in natural light or simulated natural light without any shadows and maintained a focal length of 1 while focusing on the center of the lesion. We disabled the camera’s beauty, whitening, smoothing, and filtering features during this process. For facial lesions, frontal and lateral images were captured at a 45° angle, while for symmetrical lesions on the extremities, two images of the affected area were combined. Asymmetrical lesions on the extremities required a complete image of the affected area. All images were required to ensure that the lesion covered 80% of the image area, and 20–30% of the surrounding skin was captured. In addition to this, physicians included essential meta-data (gender and lesion area) while photographing the skin lesions. The diagnosis for each case was determined by a dermatologist with at least 15 years of clinical experience in dermatology, and the dataset was reviewed by two dermatologists before inclusion. The dataset was created by 33 mid-to senior-level dermatologists. We prospectively collected 80 and 150 cases of each of the 22 most prevalent dermatological diseases respectively from 15 tertiary hospitals to serve as the validation and test set in the following experiments.

Unsupervised encoder

We employed the formulation of swapping assignments between multiple Views (SwAV) presented in ref. 15 to learn the feature representations of skin diseases in an unsupervised manner using the online clustering method. SwAV encodes two different augmented views of the same image into features z t and z s respectively. Then a set of trainable code vectors q t and q s are computed by matching these features to a set of K prototypes \(\left\{{c}_{1},\cdots ,{c}_{K}\right\}\) . The similarity between these representations is formulated as a swapped prediction problem between positive pairs, whereby feature vectors from one view are forced to match the cluster’s code from the other view. The loss function is expressed with Eq. ( 1 ),

Notably, unlike other instance-based methods, SwAV does not employ negative pairs explicitly. Instead, the representation is prevented from collapsing through the batch-wise online code computations.

Data cleansing based on distance

We train SwAV on images scraped from the Internet without annotation and use the model as a feature extractor for those images with coarse labels, which were subsequently used in the fine-tuning stage. The labels and the features of original training dataset are denoted as Y train,i and \({Z}_{{\rm{train}},i}\in {{\mathbb{R}}}^{1\times d}\left(i=\mathrm{1,2},\cdots ,{N}_{{\rm{train}}}\right)\) , with d dimensions for N train samples. For the total M classes, we use a small subset of clean annotated data from doctor diagnosis as the validation set with N val images for each class, the labels of which are denoted as \({Y}_{{\rm{val}},i}\) and use the pre-trained model to extract features \({Z}_{{\rm{val}},i}\in {{\mathbb{R}}}^{1\times d}\left(i=\mathrm{1,2},\cdots ,{N}_{{\rm{val}}}M\right)\) of them. These validation images have obvious inter-class distinctions can be served as a calibration for those images with label noise. Based on the assumption that each cluster for different diseases is a convex packet 42 , our strategy is to estimate the center of each cluster for every class based on the validation images.

We calculate the Euclidean distance between the image in the training set and the center of the M clusters.

We discard the training images whose closest cluster does not match its coarse label. That is, we only retain images that meet Eq. ( 5 ).

For dimension reduction, a principal component analysis is implemented. We project the feature representation of both the training set Z train and the validation set Z val to the vector space expanded by the first 100 principal components of the validation set.

Transfer learning to downstream task with self-adaptive threshold

Collecting data and obtaining annotations from experienced physicians for rare diseases is a challenging task. In order to demonstrate the efficacy of our approach in handling rare diseases with small sample sizes, we gathered a dataset of 220 monkeypox images from various public websites on the internet. Out of these, 50 cases from literature case reports and the European CTC official website were selected as the training set, while the remaining 170 were used as the test set. To account for variations in ethnicity and skin color, we also collected 2146 skin images from English websites using relevant keywords such as Dermatosis, Skin diseases, Rash, Ringworm, Dermatitis, etc. These images were also included in the test set. Note that the recent monkeypox outbreak were first reported in 2022, which ensured these images will not be included in our pre-training dataset or coarse-labeled dataset. Also, manual screening ensured that the training and test sets of monkeypox images had no data overlap. The diverse skin disease images used were deliberately sourced from the Internet that dated after 2022. As a result, we expect no data leakage in this transfer learning experiment.

Since class categories or diagnoses that are not included in the algorithm’s training data are common in the online scenario, we needed to deal with images that belonged to out-of-distribution (OOD) classes. To address this, we employed a self-adaptive threshold to allocate images with low classification confidence to the ‘other’ class. After fine-tuning, we extracted the feature of the training set and calculated the distance among cluster centers of different classes.

Taking the idea that the more distant the clusters, the easier it is to distinguish them, our approach considered the distance between the nearest and farthest clusters, with the threshold set as (7) to distinguish them effectively 43 .

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

The entire dataset is not available due to privacy restrictions. However, part of the data that supports the findings could be available upon reasonable request from the corresponding author. The benchmark datasets used in this study could be found as follows: DDI ( https://ddi-dataset.github.io ), fitzpatrick17k ( https://github.com/mattgroh/fitzpatrick17k ).

Code availability

Relevant code used in this study can be found on github https://github.com/shenyue-98/SwAVDerm .

Ao, Y., Feng, Q., Zhou, Z., Chen, Y. & Wang, T. Resource allocation equity in the china’s rural three-tier healthcare system. Int. J. Environ. Res. Public Heal . 19 , https://doi.org/10.3390/ijerph19116589 (2022).

Liu, Y. et al. A deep learning system for differential diagnosis of skin diseases. Nat. Med. 26 , 900–908 (2020).

Article   ADS   CAS   PubMed   Google Scholar  

Dildar, M. et al. Skin cancer detection: a review using deep learning techniques. Int. J. Environ. Res. Public Heal . 18 , https://doi.org/10.3390/ijerph18105479 (2021).

Parikh, K. S., Shah, T. P., Kota, R. & Vora, R. Diagnosing common skin diseases using soft computing techniques. Int. J. Bio-Sci. Bio-Technol. 7 , 275–286 (2015).

Article   Google Scholar  

Ahmed, K. et al. Early prevention and detection of skin cancer risk using data mining. Int. J. Comput. Appl. 62 , 1–6 (2013).

ADS   Google Scholar  

Chan, S. et al. Machine learning in dermatology: current applications, opportunities, and limitations. Dermatol. Ther. 10 , 365–386 (2020).

Dildar, M. et al. Skin cancer detection: a review using deep learning techniques[J]. Int. J. Environ. Res. Public Health 18 , 5479 (2021).

Article   PubMed   PubMed Central   Google Scholar  

Aractingi, S. & Pellacani, G. Computational neural network in melanocytic lesions diagnosis: artificial intelligence to improve diagnosis in dermatology? Eur. J. Dermatol. 29 , 4–7 (2019).

Article   PubMed   Google Scholar  

Allugunti, V. R. A machine learning model for skin disease classification using convolution neural network. Int. J. Comput. Program. Database Manag. 3 , 141–147 (2022).

Wu, Z. et al. Studies on different cnn algorithms for face skin disease classification based on clinical images. IEEE Access 7 , 66505–66511 (2019).

Benzekry, S. Artificial intelligence and mechanistic modeling for clinical decision making in oncology. Clin. Pharmacol. Ther. 108 , 471–486 (2020).

Schmarje, L., Santarossa, M., Schröder, S.-M. & Koch, R. A survey on semi-, self-and unsupervised learning for image classification. IEEE Access 9 , 82146–82168 (2021).

Chen, T., Kornblith, S., Norouzi, M. & Hinton, G. A simple framework for contrastive learning of visual representations. In III, H. D. & Singh, A. (eds.) Proceedings of the 37th International Conference on Machine Learning , 119 of Proceedings of Machine Learning Research, 1597–1607 (2020).

van den Oord, A., Li, Y. & Vinyals, O. Representation learning with contrastive predictive coding. Preprint at https://arxiv.org/abs/1807.03748 (2019).

Caron, M. et al. Unsupervised learning of visual features by contrasting cluster assignments. In Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M. & Lin, H. (eds.) Advances in Neural Information Processing Systems , 33, 9912–9924 (2020).

Huang, S.-C. et al. Self-supervised learning for medical image classification: a systematic review and implementation guidelines. NPJ Digit. Med. 6 , 74 (2023).

Azizi, S. et al. Big self-supervised models advance medical image classification. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) , 3478–3488 (2021).

Du, S. et al. (eds.) Computer Vision – ECCV 2022 Workshops , 185–202 (2023).

Wu, Y. et al. Federated contrastive learning for dermatological disease diagnosis via on-device learning (invited paper). In 2021 IEEE/ACM International Conference On Computer Aided Design (ICCAD) , 1–7, (2021).

Hsu, B. W.-Y. & Tseng, V. S. Hierarchy-aware contrastive learning with late fusion for skin lesion classification. Comput. Methods Prog. Biomed. 216 , 106666 (2022).

Wang, Z., Zhang, L., Shu, X., Wang, Y. & Feng, Y. Consistent representation via contrastive learning for skin lesion diagnosis. Comput. Methods Prog. Biomed. 242 , 107826 (2023).

Albarqouni, S. et al. Aggnet: deep learning from crowds for mitosis detection in breast cancer histology images. IEEE Trans. Med. Imaging 35 , 1313–1321 (2016).

Rawat, W. & Wang, Z. Deep convolutional neural networks for image classification: a comprehensive review. Neural Comput. 29 , 2352–2449 (2017).

Article   MathSciNet   PubMed   Google Scholar  

Yun, S. et al. Cutmix: Regularization strategy to train strong classifiers with localizable features. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2019).

Daneshjou, R. et al. Checklist for evaluation of image-based artificial intelligence reports in dermatology: CLEAR derm consensus guidelines from the international skin imaging collaboration artificial intelligence working group. JAMA Dermatol 158 , 90–96 (2022).

Wu, Y. et al. Burden of melanoma in china, 1990–2017: Findings from the 2017 global burden of disease study. Int. J. Cancer 147 , 692–701 (2020).

Article   CAS   PubMed   Google Scholar  

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 (CVPR) Workshops , 1820–1828 (2021).

Daneshjou, R. et al. Disparities in dermatology ai performance on a diverse, curated clinical image set. Sci. Adv. 8 , eabq6147 (2022).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Aggarwal, P. Performance of artificial intelligence imaging models in detecting dermatological manifestations in higher fitzpatrick skin color classifications. JMIR Dermatol. 4 , e31697 (2021).

Jain, A. et al. Development and assessment of an artificial intelligence–based tool for skin condition diagnosis by primary care physicians and nurse practitioners in teledermatology practices. JAMA Netw. Open 4 , e217249–e217249 (2021).

Razeghi, O., Qiu, G., Williams, H. & Thomas, K. Skin lesion image recognition with computer vision and human in the loop. https://api.semanticscholar.org/CorpusID:15228838 (2012).

Kim, C. et al. Fostering transparent medical image AI via an image-text foundation model grounded in medical literature. medRxiv https://doi.org/10.1101/2023.06.07.23291119 (2023).

Li, C.-X. et al. Artificial intelligence in dermatology: past, present, and future. Chin. Med. J. 132 , 2017–2020 (2019).

Guha Roy, A. et al. Does your dermatology classifier know what it doesn’t know? detecting the long-tail of unseen conditions. Med. Image Anal. 75 , 102274 (2022).

Ouyang, L. et al. Training language models to follow instructions with human feedback. In Koyejo, S. et al. (eds.) Advances in Neural Information Processing Systems , 35, 27730–27744 (2022).

Russakovsky, O. et al. Imagenet large scale visual recognition challenge. Int. J. Comput. Vis. 115 , 211–252 (2015).

Article   MathSciNet   Google Scholar  

Ronneberger, O., Fischer, P. & Brox, T. U-net: Convolutional networks for biomedical image segmentation. In Navab, N., Hornegger, J., Wells, W. M. & Frangi, A. F. (eds.) Medical Image Computing and Computer-Assisted Intervention –MICCAI 2015 , 234–241, https://doi.org/10.1007/978-3-319-24574-4_28 (2015).

Topiwala, A., Al-Zogbi, L., Fleiter, T. & Krieger, A. Adaptation and evaluation of deep learning techniques for skin segmentation on novel abdominal dataset. In 2019 IEEE 19th International Conference on Bioinformatics and Bioengineering (BIBE) , 752–759, https://doi.org/10.1109/BIBE.2019.00141 (2019).

Choi, M. G., Jung, J. H. & Jeon, J. W. No-reference image quality assessment using blur and noise. Int. J. Comput. Sci. Eng. 3 , 76–80 (2009).

Google Scholar  

National Health Commission of the PRC. Health data element dictionary. http://www.nhc.gov.cn/fzs/s7852d/202310/b02b9e310c25477faeff44b2be8ef1a1.shtml (2023).

DiSantostefano, J. International classification of diseases 10th revision (icd-10). J. Nurse Pract. 5 , 56–57 (2009).

Sun, D., Toh, K.-C. & Yuan, Y. Convex clustering: Model, theoretical guarantee and efficient algorithm. J. Mach. Learn. Res . 22 (2021).

Holger, D., Janßen, A., Resnick, S. I. & Wang, T. On a minimum distance procedure for threshold selection in tail analysis. SIAM J. Math. Data Sci . 2 , 75–102, (2020).

Download references

Acknowledgements

We are very grateful for the support from the third affiliated hospital of CQMU and cooperating hospitals for data collection and annotation. Yue Shen was supported by the World-leading Innovative Graduate Study Program in Proactive Environmental Studies (WINGS-PES) at the University of Tokyo.

Author information

These authors contributed equally: Yue Shen, Huanyu Li.

Authors and Affiliations

Simulation of Complex Systems Lab, Department of Human and Engineered Environmental Studies, Graduate School of Frontier Sciences, The University of Tokyo, Chiba, Japan

Yue Shen & Yu Chen

Shanghai Beforteen AI Lab, Shanghai, China

Huanyu Li, Kun Hu, Yiqi Tang, Zikun Wei & Junwei Lv

Institution of Aix-marseille, Wuhan University of Technology WHUT, Wuhan City, China

Shanghai Business School No. 6333, Oriental Meigu Avenue, Shanghai, China

The third affiliated hospital of CQMU, Chongqing, China

Daojun Zhang

You can also search for this author in PubMed   Google Scholar

Contributions

Yue Shen, Huanyu Li, and Junwei Lv, initiated the project and provided guidance on the concept and design. Yue Shen, Hongtao Ji, and Can Sun wrote the manuscript. Daojun Zhang provided clinical expertise, and guidance on the study design, collected and analyzed the data. Zikun Wei, Yiqi Tang, and Kun Hu developed the network architecture and helped with data collection. Yue Shen implemented data/modeling infrastructure, set up raining and testing, created the figures, wrote the methods, and performed additional analysis. Huanyu Li, Zikun Wei, and Junwei Lv supervised the project. All authors discussed the results and reviewed the manuscript. All authors approve the submitted version.

Corresponding authors

Correspondence to Zikun Wei or Junwei Lv .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information, reporting summary, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Shen, Y., Li, H., Sun, C. et al. Optimizing skin disease diagnosis: harnessing online community data with contrastive learning and clustering techniques. npj Digit. Med. 7 , 28 (2024). https://doi.org/10.1038/s41746-024-01014-x

Download citation

Received : 29 May 2023

Accepted : 18 January 2024

Published : 08 February 2024

DOI : https://doi.org/10.1038/s41746-024-01014-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

• Campus News
• Student News
• UK HealthCare
• UK Happenings
• Arts & Culture
• Professional News

MAMMA Alliance supports 4 innovative mobile health projects

LEXINGTON, Ky. (Sept. 24, 2024) — In August 2020, the University of Kentucky College of Medicine Office of Research launched 18  Alliance  Research Initiative teams from each of the five research priority areas — substance use disorder, cardiovascular, diabetes and obesity, neuroscience, and cancer — and other important emerging areas of science. Since then, some alliances have completed their research projects or transitioned to larger funding mechanisms due to their success and several new alliances have been added to the active roster.  

One of the new alliances is the mHealth Application Modernization and Mobilization Alliance (MAMMA). mHealth, also called mobile health, refers to the utilization of mobile devices to aid in health care delivery. Established in 2023, MAMMA has been making significant strides in the realm of mobile health application research, leveraging the power of technology to conduct social and biomedical research via smartphone apps. 

The overarching goal of MAMMA is to help researchers successfully integrate mobile technologies into their research methodologies, leveraging the power of technology to reach broader audiences, enhance data collection, and better inform patient care.

Mikhail Koffarnus, Ph.D., an associate professor of family and community medicine, and Carolyn Lauckner, Ph.D., an associate professor of behavioral science in the Center for Health Equity Transformation (CHET), serve as co-directors of the alliance. They support a large team of interdisciplinary researchers from the Colleges of Medicine, Health Sciences , Arts and Sciences , Fine Arts , Law , and Agriculture, Food, and Environment . 

“One of the values of this alliance is bringing everybody together to share their expertise on this common problem,” said Koffarnus. “Without it, these individuals normally wouldn’t interact because they’re in completely different parts of the university.”

Monthly meetings serve as a cornerstone for knowledge sharing and collaboration within the alliance. These sessions provide a platform for researchers to present their progress, seek feedback, and discuss common challenges, such as regulatory hurdles, buying cell phones or navigating data privacy concerns. 

Since its inception, the alliance has sponsored four significant pilot studies, each addressing different health challenges through innovative mobile solutions. 

One of MAMMA’s first pilot awardees was Yuyao Sun, M.D., a clinical neurologist, who developed a project using a mobile app to track fat mass in patients with amyotrophic lateral sclerosis (ALS). The app, connected to a Bluetooth smart scale, allows for remote monitoring of patient’s symptoms that are critical for disease management between regular doctor visits. 

“She developed this idea in response to a need that she saw in her practice,” said Lauckner. “It’s really patient-responsive.”  

Another alliance project, led by Jami Warren, Ph.D., associate professor in the UK College of Health Sciences, targets postural orthostatic tachycardia syndrome (POTS) patients by utilizing an app connected to a smartwatch to monitor and track their symptoms. By empowering patients with quantifiable data, the app enhances their ability to manage and communicate symptoms effectively with health care providers, which could potentially accelerate their diagnosis and treatment.

MAMMA awarded two additional pilot awards, co-sponsored by CHET, to projects utilizing mHealth applications to address health disparities. 

Carrie Oser, Ph.D., a University Research Professor and the Di Silvestro Endowed Professor in Sociology, is working with MAMMA’s programmer to develop an app to support an intervention for people in recovery from substance use disorders who live in rural areas and may be experiencing challenges to their recovery. The app hopes to monitor symptoms, track social interactions, and help identify stressors that may trigger a return to use. It is especially beneficial to patients in rural areas who may lack regular access to supportive services or treatment. 

Justin Huber, M.D., an assistant professor of physical medicine and rehabilitation, was inspired to pursue his project after witnessing a need among his patients. With an emphasis on helping patients regain mobility following a stroke, Huber was already working with patients in his lab to capture clips of their movement on video to analyze their recovery progress using a custom algorithm. Now, by utilizing an mHealth application, he hopes to make that level of care more accessible to other providers and to his patients. Using smartphones, the ultimate goal is to allow patients to record their movements at home, allowing Huber to digitally track their symptoms and recovery progression. 

“We’re hopeful that this could have a real impact,” said Lauckner, “especially for rural patients who need rehabilitative services but may lack the means to come to the hospital frequently or have mobility issues that make traveling difficult.”

In addition to granting pilot funding awards, MAMMA offers recipients the invaluable resource of an in-house full-time application developer, Mitchell Embry, to help develop and maintain their mobile health applications over time. 

This innovative approach not only helps jump-start researchers’ projects but also addresses a common barrier among physicians and biomedical researchers — access to coding and programming expertise — which is essential for developing effective and user-friendly mobile health interventions.

“Providing access to a full-time programmer for our researchers is basically unheard of,” said Lauckner. “But it’s been absolutely invaluable to our alliance.”

Looking ahead, the alliance aims to not only expand its impact through ongoing and future research projects but also advocate for more streamlined regulatory pathways and sustainable funding models to support researchers interested in mobile health research. 

“The MAMMA Alliance has made tremendous strides in a short time,” said Becky Dutch, Ph.D., vice dean for research. “They are opening up new avenues of research for faculty across campus. Providing community, expertise and access to an experienced programmer for those awarded pilots allow the alliance to rapidly jump-start a new research program, apply for new funding and, most importantly, work to improve health across the Commonwealth and the nation.” 

By championing these efforts, MAMMA is poised to support mobile health research innovation, ensuring that cutting-edge mHealth technologies benefit both researchers and the communities they serve. 

Learn more about the Alliance Research Initiative . 

UK HealthCare is the hospitals and clinics of the University of Kentucky. But it is so much more. It is more than 10,000 dedicated health care professionals committed to providing advanced subspecialty care for the most critically injured and ill patients from the Commonwealth and beyond. It also is the home of the state’s only National Cancer Institute (NCI)-designated Comprehensive Cancer Center, a Level IV Neonatal Intensive Care Unit that cares for the tiniest and sickest newborns, the region’s only Level 1 trauma center and Kentucky’s top hospital ranked by U.S. News & World Report.

As an academic research institution, we are continuously pursuing the next generation of cures, treatments, protocols and policies. Our discoveries have the potential to change what’s medically possible within our lifetimes. Our educators and thought leaders are transforming the health care landscape as our six health professions colleges teach the next generation of doctors, nurses, pharmacists and other health care professionals, spreading the highest standards of care. UK HealthCare is the power of advanced medicine committed to creating a healthier Kentucky, now and for generations to come. 

Latest Stories

Uk researchers help understand sex-specific differences in aortic aneurysms, 3 steps to take if you receive an academic alert, recovery programs help kentuckians with substance use disorder, ‘behind the blue’: 2024 great teachers spotlight — eric thomas weber.

• Emergency information
• Increase Font Size
• Decrease Font Size

Northern NSW Local Health District

Finalist: Excellence in Aboriginal Healthcare​ Award

The Rheumatic Health Disease (RHD) project is a collaborative project by Northern NSW Local Health District (NNSWLHD) with Aboriginal Medical Services (AMS). The project aims to address the significant public health issue of Acute Rheumatic Fever (ARF) and RHD among Aboriginal communities in Northern NSW.

Initiated by a process undertaken by AMS doctors identified many undiagnosed RHD cases, leading to the development of a comprehensive healthcare strategy.

This project aims to:

• Reduce incidences of RHD.
• Develop educational resources to improve staff and community knowledge in partnership with local AMS’.
• Improve identification and treatment of Group A Streptococcus and ARF.
• Development and implementation of a new Clinical Pathway to guide NSW Health Clinicians.

The NNSWLHD project team surveyed to assess staff knowledge. This informed the development of tailored educational resources and a training package, that was successfully implemented. The Clinical Pathway was developed in collaboration with NNSWLHD Senior Medical Staff.

We acknowledge the contributions from Mr Scott Monaghan for his support and advocacy and his medical team, NNSWLHD staff Emma-Jane Davies (lead), Dr Tim Williams, Kirsty Glanville, Robyn Auld, Sally Adams and Daniel Ashton.

The initiative has been shared with other local health districts and recognised at various healthcare forums. The RHD project showcases the collaboration and commitment to culturally safe healthcare, aligning with NSW Health's mission to support Aboriginal health.​​

​​​​​​​​​​​​​​​​​​​​​

Related links

Meet the other finalists for the Excellence in Aboriginal Healthcare Award.

• Culture Leading the Way
• Daalbirrwirr Gamambigu (Safe Children)