Piloted 2024
Skin Allergy Detection
This project solves slow and subjective allergy testing by automatically analyzing skin test images. It provides faster, accurate, and consistent results using a simple, low-cost digital tool.
Outcome Metrics
- Tested on 50+ patient images with automated wheal detection and measurement.
- Achieved approximately 95% accuracy with low measurement error (~0.48 mm).
- Reduced diagnosis time and eliminated observer variability in skin prick test interpretation.