Expanding Pediatric Acne Epidemiologic Research: Validation of Automated Feature Extraction from Clinical Notes
University of California, Los Angeles
University of California, Los Angeles
Overview:
Acne affects over 184 million teenagers worldwide, causing scarring, skin discoloration, and emotional distress that persist into adulthood. Studying which treatments work best requires detailed clinical information trapped in doctors’ notes. Researchers currently read these notes by hand, limiting the size and pace of pediatric acne outcomes research. We developed a computer application at UCLA that reads clinical notes and identifies acne details: lesion types, affected body areas, and lasting skin changes like scarring. This project will test the program’s accuracy against expert review by a board-certified pediatric dermatologist and evaluate whether machine learning improves performance. Our tool will allow researchers to study acne, scarring, skin discoloration, and treatment outcomes across thousands of patients. Under the mentorship of Dr. Marcia Hogeling, PeDRA EIRG support will further my development as a clinician-researcher in pediatric dermatology.
Status:
This was funded through a 2026 PeDRA Emerging Investigator Research Grant.