Healthcare
Skin Condition Classifier
Classify skin photographs across 30+ common dermatological conditions including eczema, psoriasis, acne severity grading, and lesion type pre-screening. Outputs include top-3 candidate conditions with confidence levels and recommended clinical urgency tier. All clinical outputs require GeraWitness human review before being surfaced to users.
Accuracy
87% top-1 accuracy on ISIC benchmark dataset
Latency
<1.2s
Use cases
- Telemedicine pre-consultation triage
- Dermatology waitlist priority scoring
- Pharmacy advice augmentation
- Clinical trial recruitment screening
Gera Systems integration
This capability is wired into GeraClinic + GeraWitness within the Gera Systems platform. It is available via the GeraLens API for third-party integrations.
Start using Skin Condition Classifier
GeraLens provides this capability via a simple REST API. Join the beta to access it — 1,000 free API calls per month during beta.
Join the betaSkin Condition Classifier — questions & answers
- How accurate is GeraLens Skin Condition Classifier?
- Skin Condition Classifier runs at 87% top-1 accuracy on ISIC benchmark dataset. Accuracy is tracked per class and published rather than averaged into a single headline number.
- How fast is GeraLens Skin Condition Classifier?
- Typical inference latency is <1.2s via the GeraLens REST API.
- What can I use GeraLens Skin Condition Classifier for?
- Common use cases: Telemedicine pre-consultation triage; Dermatology waitlist priority scoring; Pharmacy advice augmentation; Clinical trial recruitment screening.
- Which Gera Systems product does Skin Condition Classifier integrate with?
- Skin Condition Classifier is wired into GeraClinic + GeraWitness and is also available standalone via the GeraLens API.