Cancer Medical Imagery
- Lesion-level
- Labels
- Close-up
- Capture
- Metadata-rich
- Clinical context
- De-identified
- Privacy
We typically reply within 24 hours
Close-up dermatological images of melanoma, carcinoma, and keratosis lesions with rich clinical metadata, collected with consent and de-identified.
A consented dermatological image collection focused on melanoma, basal and squamous cell carcinoma, and keratosis lesions, captured close-up with consistent lighting and rich clinical metadata. The set is built for skin-cancer detection and triage research where labeled, well-documented imagery is scarce.
Every image is collected from consenting contributors, de-identified, and reviewed for PII before delivery. This is research and model-development data and is not a diagnostic device.
Highlights
- Melanoma, carcinoma, and keratosis lesions captured close-up
- Consistent lighting and framing for reliable comparison
- Rich metadata: lesion type, body site, and capture conditions
- Consent-led collection with de-identification
- PII reviewed before delivery
Condition coverage
Common skin-cancer and pre-cancerous lesion types. Coverage extends to specific conditions, skin tones, or capture protocols on request.
Capture and format
High-resolution close-up images with consistent lighting, captured per a documented protocol. Each image carries lesion type, body site, and capture-condition metadata.
Annotations
Lesion-type and body-site labels with capture metadata as standard, plus optional segmentation masks and severity labels on request.
Provenance
- Consenting contributors with signed consent
- Images de-identified before delivery
- PII reviewed before delivery
- Per-image audit trail and licensable usage rights
Use cases
- Skin-cancer detection and classification research
- Lesion segmentation and triage models
- Domain adaptation and fairness across skin tones
- Evaluation of dermatology vision-language models
