Screen a patient
Point the camera at the skin lesion, then capture.
No photo yet
Keep this photo or retake it.
My Cases
Track every case you submitted and where it is in review.
Profile & Score
How it works
- Capture 3 angles of a real skin lesion: overview, close-up, side.
- A vet checks each case in two review phases before it earns full points.
- Rare and correctly-labelled conditions earn bonus points.
Screening result
No heatmap available.
Most likely
All conditions
What is the correct condition?
Your answer trains and improves the AI.
Labeling checklist
Annotation Lab
Turn old raw images into training-ready labels.
Annotation projects
Choose the right task so the vet sees the right labels.
Project images
Select an annotation project
Search or choose a box above. The project images will appear here.
No raw photos queued
Add images to data/raw, data/to_annotate, or data/vet_clinic_photos.
needs label
100%
Layers
0
| Object | Label | Area | Source | State |
|---|
Raw image annotation
Body area
Visual signs
Class
Regions
Quality
Focused annotation
Answer the essentials, draw a box only when useful, then save and move to the next image.
Body area
Teach the AI where this finding appears
Visual signs
Add feature tags for easier recognition
Label details
Regions
QA attributes
Saved into notes
Advanced tools
Activity
0/0
Edit trail
0
No heatmap available.
AI prediction
Vet label
DUAL REVIEW REQUIRED
Similar approved cases
Previous reviews
Scoring checklist
Image Quality
25%
Label Correctness
40%
Metadata Completeness
15%
Clinical Plausibility
20%
Overall score
0%
Review access requires reviewer or admin role.
Patients
Case history
Treatments
Outcomes & follow-up
Clinic dashboard
Operational view of cases, review quality, and training readiness.
—Total cases
—Confirmed
—AI vs vet accuracy
—Ready to train
Confirmed cases per condition
Training pipeline
Clinic data asset
Confusion matrix
Improve the AI
Confirmed cases become training data.