University of Utah

Self-supervised foundation model for clinical flow cytometry, producing panel-agnostic specimen-level representations from multi-panel data.

Self-supervised flow cytometry model producing tube-level representations that lightweight heads read for hematologic diagnosis and sample viability.

Generative histopathology foundation model: a diffusion transformer trained on 30M H&E tiles, conditioned on self-supervised slide embeddings.

Chest CT translation model that synthesizes a virtual expiratory scan from one inspiratory volume, so small airways disease needs no second scan.