ML Healthcare Virtual Meet-Up

​This session continues the CT foundation models series with a paper focused on abdominal CT.
Where prior work like CT-RATE/CT-CLIP targets chest CT, Merlin tackles abdominal CT, a modality with more anatomical variety and stronger clinical demand given radiologist shortages. It's built on a multistage pretraining framework that adds electronic health record data alongside CT scans and radiology reports, trained on over 6 million images and 1.8 million diagnosis codes.

The paper: Merlin: a computed tomography vision–language foundation model and dataset (Nature, 2026) evaluates across 752 tasks zero-shot classification, phenotype prediction, 5-year disease risk, report generation, and 3D segmentation with external validation on 44,098 scans from 3 independent sites.

Link: https://www.nature.com/articles/s41586-026-10181-8

Come ready to discuss multimodal grounding (EHR + imaging + text), external validation rigor, and how this compares to chest-focused CT foundation models.


Join our sessions to catch up on all things  ML Healthcare research, led by Leema Krishna, Anas Zafar and Oumayma Essarhi. Learn more and check out recent events from this group on their minisite.

​For the full experience, connect with the community on Discord. Learn more about the Cohere Labs ML research open science community, and apply to join.