Trustworthy, clinically deployable AI for ophthalmology and broader healthcare — from medical vision–language models to the bedside.
Our work spans four directions with one shared goal — methods that survive contact with real clinical data and earn a place in the workflow, from ophthalmology to broader healthcare.
Foundation models that read medical images and clinical text together — retinal photography, OCT, MRI and reports — to learn transferable representations for diagnosis, grounding and report generation.
Fusing heterogeneous modalities — OCT, MRI, fundus, structured records and free text — under missing-modality and weak-annotation conditions to make robust, holistic clinical predictions.
Reliability you can deploy: uncertainty estimation, out-of-distribution and open-set detection, calibration and model IP protection — so models know what they don't know.
Carrying methods out of the lab and into real hospital workflows — addressing data heterogeneity, scarce labels and domain shift to deliver tools clinicians can actually use.
Meng Wang is an Assistant Professor in the Department of Ophthalmology, Yong Loo Lin School of Medicine, NUS, and a member of the Centre for Innovation & Precision Eye Health. His research develops trustworthy and clinically applicable AI for ophthalmology and broader healthcare — spanning medical vision–language models, multimodal learning and reliable AI. Prior to NUS he was a postdoctoral researcher at Harvard Medical School and a researcher at A*STAR, Singapore. His work appears in Nature Communications, Cell Reports Medicine, IEEE TMI, CVPR and MICCAI, and has received multiple international awards.
Meet the team →
PAI Lab is recruiting PhD students, postdocs, research assistants, interns and visiting students for 2027 — now open.
See open positions →PAI Lab is officially recruiting — PhD students, postdocs, research assistants/associates, interns and visiting students are all welcome. Read more →