Medical AI
Foundation models and clinical AI for imaging and multimodal data.
Developing reliable, generalizable, and clinically meaningful AI across modalities, cohorts, and real-world settings.
Foundation models and clinical AI for imaging and multimodal data.
Understanding and improving robustness under distribution shifts and real-world change.
Evaluating and mitigating bias for fair and generalizable performance.
Applying AI to accelerate scientific and biomedical discovery.
In memory of Abdul Razaq Parsa —
a lifelong pursuer of
knowledge.
PARSA stands for Pursuing Advances in Reliable Scientific AI. We are building a research initiative focused on foundational questions at the intersection of artificial intelligence, medicine, and scientific discovery.
PARSA is named in honor of Abdul Razaq Parsa, whose lifelong pursuit of knowledge and love of books inspire our commitment to curiosity, rigor, and meaningful progress.
Our work will explore how intelligent systems can become more robust, equitable, interpretable, and useful beyond controlled benchmarks.