Open-access 3D dermatology dataset
Project EquiDerm is a peer-verified repository of 3D dermatological models and dermoscopy images, contributed by clinicians and available to verified researchers worldwide.
PDF, JPG, PNG
Supports GLB, GLTF, OBJ, STL, USDZ formats
About Project EquiDerm
Project EquiDerm is a free, open-access database of digital 3D models and dermoscopy images displaying basal cell carcinomas across Fitzpatrick skin types III to VI — built to combat the inequalities in skin cancer care experienced by people of colour.
Skin cancer is the most common form of cancer worldwide, the most prevalent being basal cell carcinomas, which account for approximately 70% of all skin cancers. Primarily triggered by cumulative UV light exposure, symptoms often present on exposed regions such as the face and neck as nodular, ulcerated or superficial lesions. Diagnosis relies heavily on doctors' visual recognition, and subsequently, biopsies.
While lighter skin tones are more susceptible to developing skin cancers due to a lack of UV-protective melanin, these cancers may still manifest in individuals with darker skin. Yet dermatology education fails to provide resources representative of the diverse range of patients who may develop skin cancer, with over 70% of clinical teaching images being from lighter skin.
As a result, clinicians report significantly greater confidence in detecting skin cancer in lighter skin types, and lower confidence in darker skin. To add to this, AI models used for skin cancer diagnosis have consistently displayed poor performance on dark skin types, likely due to being trained on predominantly light-skin datasets.
Consequently, while higher Fitzpatrick skin types — a numerical classification of human skin colour from I to VI, with VI being the darkest — carry a lower risk of developing skin cancer, they are statistically more at risk of higher rates of mortality due to delayed diagnosis and treatment.
Project EquiDerm addresses the current underrepresentation of darker skin types in medical datasets, promotes equity in cancer care, and supports the rise of AI integration in medicine by enabling more accurate machine learning. AI models can be trained using our dataset — which is rich in images and models of skin cancer in dark skin types — in contrast to the datasets models are currently trained on, which have very few images of dark skin.
This lack of training data for darker skin types results in poor diagnostic accuracy for medical AI models, but Project EquiDerm attempts to directly address this issue.
The Project EquiDerm platform is connected to an encrypted cloud server for data security and is built with a login-protected admin dashboard, where moderators can review clinician submissions, data access requests and file reports, and manage the existing library to ensure quality across the database.
We have also added a clinician portal to allow doctors to upload their own 3D files and dermoscopy images of BCCs in darker skin types, enabling the continuous expansion and development of our platform over time.
Our bank of models and images of higher Fitzpatrick skin types will hopefully serve as a useful professional and educational tool — improving access to and understanding of how skin cancer can present differently in patients with darker skin, and spreading awareness of the lack of diversity in dermatological training.
As doctors will be able to use this platform as a point of reference in clinical practice, we also hope Project EquiDerm might improve doctors' confidence in recognising skin cancer in higher Fitzpatrick types, enabling earlier diagnosis and better outcomes for these patients.
Project EquiDerm's library is concentrated on Fitzpatrick Types III through VI — the range most underrepresented in existing dermatology datasets.
All lesion data is de-identified before submission. Browsing the model library is open to everyone — anyone can explore titles, metadata, and 3D previews without an account. Downloading raw model files for research or training requires a verified researcher or institutional account, granted on request.
Project EquiDerm is a research and education resource and is not a diagnostic tool. It does not replace clinical judgement, biopsy, or histopathological confirmation.
Legal
Last updated: June 2026
Project EquiDerm ("we", "us", "the platform") is an open-access research and education repository of 3D dermatological models. This policy explains what information we collect, how we use it, and the choices available to you.
When you browse the model library, we do not require an account and do not collect personal information. When you request download access or contribute a model, we collect the information you provide directly: your name, email address, institution, professional credentials (e.g. GMC, BMA, or NPI registration numbers), and a description of your research purpose.
All 3D models and associated clinical metadata submitted to the platform are de-identified prior to upload by the contributing clinician. We do not knowingly store patient names, dates of birth, contact details, or any other directly identifying information alongside model data.
Contact and credential information is used solely to verify contributor qualifications, process download access requests, and respond to enquiries. We do not sell, rent, or share your personal information with third parties for marketing purposes.
Access request records and contributor credential information are retained for as long as your account remains active, or as required to maintain the integrity of the research dataset and audit trail.
You may request access to, correction of, or deletion of your personal information at any time by contacting us at projectequiderm@gmail.com.
We may update this policy from time to time. Material changes will be reflected on this page with an updated "last updated" date.
Legal
Last updated: June 2026
By accessing or using Project EquiDerm, you agree to be bound by these Terms of Use. If you do not agree, please do not use the platform.
Project EquiDerm provides an open-access library of 3D dermatological lesion models, including Basal Cell Carcinoma (BCC) presentations across Fitzpatrick skin types III–VI, for research, education, and AI training purposes. The platform is not a diagnostic tool and must not be used as a substitute for professional medical advice, diagnosis, or treatment.
Browsing the model library is open to all users without an account. Contributing models requires verified medical credentials (e.g. registration with the GMC, BMA, AMA, or NPI). Requesting download access for research or training purposes requires submission of accurate contact and institutional information.
Clinicians contributing 3D models warrant that all submitted data has been fully de-identified, that they hold appropriate clinical authority and patient consent to share the data for research and educational purposes, and that the metadata provided (Fitzpatrick scale, lesion type, diagnosis stage, and anatomical location) is accurate to the best of their knowledge. All submissions are subject to peer review before publication.
You agree not to use the platform or any downloaded models to attempt to re-identify individuals, to misrepresent model data as a confirmed clinical diagnosis, or to use the content for any unlawful purpose.
Models and metadata contributed to Project EquiDerm remain subject to the data use terms set out separately. The Project EquiDerm name, logo, and platform design are the property of Project EquiDerm.
Content on this platform is provided "as is" for research and educational purposes. Project EquiDerm makes no warranties as to the accuracy, completeness, or clinical reliability of any model or metadata, and accepts no liability for decisions made on the basis of platform content.
We may revise these Terms of Use periodically. Continued use of the platform after changes are posted constitutes acceptance of the revised terms.
Questions about these terms can be sent to projectequiderm@gmail.com.
Legal
Last updated: June 2026
This policy describes how 3D models, dermoscopy-style metadata, and associated clinical information hosted on Project EquiDerm may be accessed and used by researchers, educators, and institutions.
The model library is browsable by anyone: titles, lesion metadata (Fitzpatrick scale, lesion type, diagnosis stage, anatomical location), and 3D previews are open access. Downloading raw model files (GLB, GLTF, OBJ, or STL formats) and associated datasets requires an approved Request Access submission, reviewed on a case-by-case basis.
Approved researchers and educators may use downloaded models and metadata for academic research, AI/ML model training and validation, clinical education and training materials, and the development of tools intended to improve equitable skin cancer detection across all skin tones.
Downloaded data may not be used to attempt to re-identify any individual, resold or redistributed as a standalone commercial dataset without written permission, represented as providing a confirmed clinical diagnosis, or used to train models intended to discriminate against any group on the basis of skin tone or other protected characteristics.
Where models or derived datasets are used in published research, presentations, or training materials, please credit "Project EquiDerm" as the data source, in addition to any contributor attribution requested at the time access was granted.
All models are de-identified by the contributing clinician prior to submission and reviewed for identifying features during the peer review process. If you believe any published model contains identifying information, please contact us immediately.
Project EquiDerm reserves the right to revoke download access for any account found to be in breach of this Data Use Policy or the Terms of Use.
For questions about data use, attribution, or to report a concern, contact projectequiderm@gmail.com.
This area is restricted to Project EquiDerm administrators. Sign in with your administrator email and password to view contribution submissions, access requests, and reports, and to upload models directly to the library.
Incorrect email or password. Please try again.