Data-centric AI for Healthcare and Medicine
The Tushar Lab is a newly established research group at the University of Arizona advancing data-centric AI for healthcare and medicine through data discovery, intelligent tools, and the integration of clinical, simulated, and synthetic datasets.
Department of Radiology and Imaging Sciences · University of Arizona
About the Lab
Our Mission
We are building a research program around a central question: how do we develop AI for healthcare and medicine that is trustworthy, rigorously evaluated, and grounded in clinical needs?
Our work combines data-centric AI, intelligent tools, and clinical, simulated, and synthetic data to move from data to discovery in ways that better support healthcare professionals, reduce workload, and strengthen reliability in practice.
As a newly established and growing lab, we welcome collaboration with researchers, students, and partners who share an interest in trustworthy, reproducible, and clinically grounded AI.
Recently Shared Data
✦ View All DataiTrialSpace
Cohort-by-specification engine for lung-CT AI: declare a virtual trial by prevalence, nodule size, lobe distribution, and demographics; synthesize anatomically grounded CTs; evaluate CADe/CADx/VLMs under controlled conditions. 13 trial modes, 44K+ synthetic volumes, and a zero-shot VLM benchmark.
HAID
A curated dataset for health AI research and development, enabling training and evaluation of medical imaging models.
News
✦ View All News
Tushar Lab Launches at the University of Arizona
The Tushar Lab has officially launched at the University of Arizona in the Department of Radiology and Imaging Sciences, with a research focus on data-centric AI for healthcare and medicine.
✦ Read More Introducing iTrialSpace — the first public release from Tushar Lab
An engine that flips lung-CT AI evaluation from "find more data" to "specify the data you need": declare a cohort by prevalence, nodule size, lobe distribution, and demographics, and the engine manufactures anatomically grounded synthetic CTs to match.
✦ Read More
Paper accepted at SASHIMI 2026 (MICCAI 2026)
Our collaborative work with the Duke Center for Virtual Imaging Trials has been accepted at the SASHIMI 2026 Workshop, held in conjunction with MICCAI 2026 in Strasbourg.
✦ Read More