Evaluating Novel Diagnostic Generative AI Models for Evaluation (ENDGAME) of Pediatric Alopecia: A Comparative Study of Scarring and Non-Scarring Types
Perelman School of Medicine at the University Pennsylvania
Children’s Hospital of Pennsylvania
Overview:
Hair loss in children can be stressful and confusing for both patients and families. There are different types of hair loss, and getting the correct diagnosis is important so that children can receive the right treatment. However, this can sometimes require specialized tests or a dermatologist’s opinion. This project explores whether new computer programs powered by artificial intelligence (AI) can help doctors diagnose different types of pediatric hair loss more easily and accurately. Using two sets of scalp photos, some from published research and some from a hospital image library, we will test how well AI models can tell the difference between scarring and non-scarring types of hair loss. We will also see if the computer can suggest possible diagnoses that match those of experienced doctors. If successful, this research could help speed up diagnosis and ultimately improve care for children with skin and hair conditions.
Status:
This was funded through a 2025 PeDRA Emerging Investigator Research Grant.