Akshay Chaudhari

Timeline

1. 2012–2017

   Bioengineering degrees and MRI

   He earns a BS in bioengineering at the University of California, San Diego in 2012 and completes a PhD in Stanford’s Department of Bioengineering in 2017. His thesis is “Advances in Morphological and Quantitative Musculoskeletal MRI.” In June 2017, he continues as a postdoctoral researcher in Stanford’s Body MRI Research Group.

2. 2018

   Deep learning for MRI

   In 2018, he coauthors “Deep Learning Super-Resolution Enables Rapid Simultaneous Morphological and Quantitative Magnetic Resonance Imaging.” The method generates high-resolution images while preserving accurate cartilage T2 relaxation measurements, which help assess tissue biochemistry.

3. May 2019

   A knee MRI award

   In May 2019, he wins the International Society for Magnetic Resonance in Medicine’s W. S. Moore Young Investigator Award. The award recognizes his work on a five-minute knee MRI scan that he tests in clinical patients.

4. January 2021

   Joining Stanford’s IBIIS faculty

   In January 2021, he becomes an assistant professor in Stanford’s IBIIS division in the Department of Radiology. His work in the Body MRI Research Group develops rapid knee imaging and applies machine learning to image reconstruction.

5. July 2023

   Federated evaluation with MedPerf

   In July 2023, he coauthors “Federated benchmarking of medical artificial intelligence with MedPerf” in Nature Machine Intelligence. The open platform distributes AI models to healthcare institutions so that each institution can evaluate them on its own data.

6. July 2025

   MedVAL checks medical text

   In July 2025, he coauthors a study introducing MedVAL, which uses synthetic data to train language models to check whether AI-generated medical text agrees with its inputs. The team introduces MedVAL-Bench, containing 840 physician-annotated outputs, and evaluates models across six medical tasks. It releases code, the benchmark and the MedVAL-4B model.

7. November 2025

   Cognita Imaging joins Mosaic

   In November 2025, Mosaic Clinical Technologies acquires Cognita Imaging, where Chaudhari is a co-founder with Louis Blankemeier and Zhihong Chen. Cognita’s vision-language models analyze X-rays and head CT scans and draft preliminary results for physicians to review.

8. March 2026

   Merlin interprets CT scans

   In March 2026, he coauthors the Nature paper introducing Merlin, a three-dimensional vision-language model that learns from CT scans, electronic health records and radiology reports. The team evaluates the model on 752 tasks, including disease prediction, report generation and organ segmentation.

9. September 2026

   The STEWARD monitoring project

   In September 2026, he co-leads a Stanford team with Roxana Daneshjou and Euan Ashley that receives up to $14.9 million from ARPA-H. The team proposes STEWARD, a system to monitor autonomous AI agents used in cardiovascular care and report to clinical teams.

10. As of October 2026

    Associate professor at Stanford

    As of October 2026, he is an associate professor (research) in Stanford’s Department of Radiology and Department of Biomedical Data Science. His group develops representation learning and evaluation methods for medical AI that combines images, language and medical records. He also serves as interim chief of the Imaging Informatics research section, co-director of Clinical AI in Radiology, and associate director of research and education at the Stanford AIMI Center.