Skin Analysis API for Telemedicine Platforms
Triage skin concerns at the front door of your telehealth experience — so the right cases reach a clinician, and the obvious ones don't. One REST call returns lesion classification, differential, confidence and a 4-level triage recommendation.
The async-derm triage problem
Most telemedicine queues mix benign skin tags with suspicious lesions and acute rashes. Clinicians spend hours triaging — and patients wait days for the obvious cases. A pre-clinician AI screening pass, powered by the ScanSkinAI REST API, cuts that overhead without changing your clinical model. For non-engineering teams, the same triage is available as a 2-line embeddable widget.
Why telemedicine teams choose ScanSkinAI
Designed for the realities of asynchronous derm triage — not retrofitted from a consumer skin app.
Async-first design
Single REST call, structured JSON — fits any intake queue, EHR or async messaging platform.
Built for triage
4-level routing (self-monitor · GP · dermatologist · urgent) tuned to clinical workflows, not consumer scores.
Multi-lesion ready
Full-body screening returns per-lesion bounding boxes and IDs — see the multi-lesion API on our developer hub.
Multi-language outputs
Clinician summaries in EN, ZH, JA, KO, ES, FR, DE — patient-facing copy in their language, clinician copy in yours.
Multi-region residency
Optional EU / UK data residency for regulated deployments; configurable retention per partner agreement.
Clinician-friendly output
Differential diagnosis with confidence per class — so reviewers can quickly accept, override or escalate.
What the API returns
Classification
A primary class label for the lesion in the image.
Differential
Top alternative conditions to support clinical review.
Confidence + probability
Numeric scores you can threshold on for your own routing rules.
Clinical triage
Self-monitor · GP · dermatologist · urgent — a routing recommendation, not a diagnosis.
End-to-end triage flow
From patient upload to clinician queue, in under 5 seconds per case.
1. Upload
Patient submits a photo via your app — web, iOS, or Android.
2. API call
Your backend POSTs the image (or a signed URL) to /v1/analyze with a bearer token.
3. AI inference
Lesion classification + differential + confidence in ~2–4s.
4. Triage
API returns one of four routing levels you can act on programmatically.
5. Clinician
Case lands in the right queue with a structured summary attached.
For full-body screening with per-lesion bounding boxes, see Multi-Lesion Detection.
Drop it into your intake step
Send the patient photo, get a structured triage decision back. Attach the JSON to the consult record and route accordingly.
Request
curl -X POST https://api.scanskinai.com/v1/analyze \ -H "Authorization: Bearer $API_KEY" \ -H "Content-Type: multipart/form-data" \ -F "image=@lesion.jpg" \ -F "patient_age=42" \ -F "body_location=upper_back" \ -F "duration=3_months" \ -F "language=en"
Response
{
"classification": "basal_cell_carcinoma",
"confidence": 0.97,
"probability": 92,
"differential": [
{ "class": "seborrheic_keratosis", "score": 0.04 },
{ "class": "nevus", "score": 0.02 }
],
"triage": {
"level": "dermatologist",
"urgency": "high",
"summary": "Strong likelihood of BCC. Refer to dermatology."
},
"quality_flags": []
}See the engineering walkthrough: Add a skin analysis API in a day.
Routing patterns we see in production
Self-care lane
Triage = self-monitor → educational content + scheduled re-check, no clinician time used.
Async GP lane
Triage = GP → routed to your async-message queue with the AI summary attached for the clinician.
Urgent / derm lane
Triage = urgent or dermatologist → bumped to the front of the queue with high-priority flag.
Integration checklist for telemedicine vendors
What to set up before flipping production traffic on.
Webhook callback URL
HMAC-SHA256 signed events for async result delivery.
Bearer auth
Per-environment API keys; rotate via the developer portal.
PHI handling toggle
Process-and-discard, or persist for audit — per partner agreement.
Retention + residency
EU, UK or default region; configurable retention window.
Clinician summary template
Custom summary text mapped to your consult notes.
Audit logging
Every call, every override, every retry — auditable end-to-end.
Security & PHI handling
- Bearer token authentication; HMAC-SHA256 webhook signatures.
- ISO 27001, UKCA Class I, GDPR aligned and HIPAA-ready handling.
- Optional EU / UK data residency for regulated deployments.
- Configurable retention — process-and-discard, or persist for audit, per partner agreement.
Frequently asked questions
Is ScanSkinAI a medical device?
ScanSkinAI is a UKCA Class I registered software-as-a-medical-device for informational and triage support. It does not provide a diagnosis — outputs are designed to support, not replace, a qualified clinician's decision.
Can the API replace my clinicians?
No. The API is a triage assistant. It returns a structured classification, differential, confidence and triage recommendation that your clinical team can act on or override. You keep full clinical control.
How do you handle PHI / HIPAA?
ScanSkinAI is HIPAA-ready and ISO 27001 certified. You can configure process-and-discard, EU/UK data residency, and signed HMAC-SHA256 webhooks. We sign BAAs and DPAs with regulated telemedicine partners.
How do I add it to my existing patient queue?
One REST POST returns the AI assessment in seconds. Most teams call it from their intake step (before a clinician sees the case) and attach the JSON to the consult record. See our walkthrough: 'Add a skin analysis API in a day'.
What does pricing look like?
Per-call usage pricing with volume discounts. The sandbox is free with no rate caps for development. Production pricing depends on monthly call volume and data residency — talk to sales for a quote.
Continue exploring
Choose the integration depth and audience that fits your platform.
Capabilities, endpoints and pricing for the full ScanSkinAI API.
Drop-in Web Component — live in 2 minutes, no backend required.
Full-body screening with per-lesion bounding boxes and IDs.
Consumer-facing integration patterns and onboarding flows.
Feature-by-feature comparison for technical buyers.
Step-by-step engineering walkthrough with code samples.
Deployment options for insurers, employers and platforms.
Member-facing skin triage as a value-add benefit.
Add preventive skin checks to your wellbeing offering.
Start building today
Self-serve sandbox credentials in minutes. Production access reviewed within ~24 hours. Or skip the build entirely with our 2-line widget.