Reliability across clinical settings
Auditing medical foundation and vision-language models under domain shift, with attention to localized reliability, calibration, and high-confidence errors.
About Me
Hello! I am Pengyang Yu, a PhD student in the School of Computer Science at University College Dublin.
My research focuses on auditing and monitoring the reliability and fairness of medical foundation models across clinical settings.
For academic inquiries, please contact me at pengyang.yu@ucdconnect.ie.
Research Highlights
Fairness in medical AI
Studying dual fairness in federated medical imaging: fairness across collaborating institutions and fairness across patient groups.
Context-aware deployment
Developing monitoring, calibration, and fair deferral strategies that help clarify when, where, and for whom medical AI can be trusted.
Education
Dublin City University
MSc in Artificial Intelligence
2024.09 – 2025.09
Selected Publications
Published · 2026-07-05
Fairness in federated medical imaging: a systematic review through the dual fairness lens
Artificial Intelligence Review · 2026 · Journal article
DOI: 10.1007/s10462-026-11632-4Published · 2025-12-25
MedLiteNet: Lightweight Hybrid Medical Image Segmentation Model for the Medical Internet of Things (MIoT)
IEEE Journal of Selected Areas in Sensors · 2026 · 3:78–89 · Published 2025-12-25
DOI: 10.1109/JSAS.2025.3648685