James S. Duncan

Also known as Jim Duncan

James Duncan is a biomedical engineer at Yale whose model-based methods for analyzing medical images have grown into deep learning for brain imaging. He worked on infrared image processing at Hughes Aircraft from 1973 to 1983, joined Yale's faculty in 1983, and from 1987 used parameterized shape models with Bayesian reasoning in image segmentation. In the MICCAI community he chaired the 1997 Information Processing in Medical Imaging meeting, was general chair of MICCAI 2005, became a MICCAI Fellow in 2010, and received the society's Enduring Impact Award in 2017. His later work includes a graph neural network pipeline for autism biomarkers in task-fMRI in 2019, the AdaBelief optimizer in 2020 and BrainGNN in 2021. He chairs Yale's Department of Biomedical Engineering, a post he has held since 2022, and holds the Hunt professorship in Biomedical Engineering, Radiology and Biomedical Imaging.

Now at: Yale University
Born: 1951
Badges: Bridge
Loyal
Pioneer

Years link to their sources.

Timeline

1973–1983

Engineer at Hughes Aircraft

In 1973, he joins the Electro-Optical and Data Systems Group at Hughes Aircraft Company, where he works on signal and image processing for forward-looking infrared imaging systems until 1983. Along the way he earns an M.S. from UCLA in 1975 and a Ph.D. in electrical engineering from the University of Southern California in 1982.
1980s

Joining Yale and modeling shape

He joins the faculty of Yale University in 1983. From 1987, he and his research team use parameterized global shape models with a Bayesian reasoning strategy to bring prior object shape into image segmentation, and from the late 1980s he uses shape features on the inner and outer surfaces of the heart wall as material tags for tracking left ventricular motion.
1997–2005

Chairing meetings and NIH panels

In 1997, he chairs the 15th international meeting on Information Processing in Medical Imaging. From 1999 to 2003, he is a charter member of the NIH Diagnostic Imaging Study Section, which he chairs from 2001 to 2003, and in 2005 he is general chair of the MICCAI meeting. He is also one of the founding co-Editors-in-Chief of the journal Medical Image Analysis.
2010

A MICCAI Society fellowship

The MICCAI Society names him a Fellow in 2010 for outstanding scientific contributions to biomedical image analysis and for promoting the field through conference organization, journal editorship and service to the Society.
2014

Connecticut science academy election

In 2014, he is elected to the Connecticut Academy of Science & Engineering.
2017

The MICCAI Enduring Impact Award

In 2017, he receives the MICCAI Society's Enduring Impact Award, which goes to a senior researcher whose work has made an enduring impact on medical image computing and computer assisted interventions.
2019–2021

Deep learning for brain imaging

In 2019, he coauthors a graph neural network pipeline that uses task-fMRI to find autism biomarkers, with Xiaoxiao Li, Nicha Dvornek, Yuan Zhou, Juntang Zhuang and Pamela Ventola. The AdaBelief optimizer, coauthored in 2020, adapts stepsizes by the belief in observed gradients and is tested on image classification and language modeling. In 2021, he contributes to BrainGNN, an interpretable brain graph neural network for fMRI analysis published in Medical Image Analysis.
As of October 2026

Hunt Professor and department chair

He is the Ebenezer K. Hunt Professor of Biomedical Engineering, Radiology and Biomedical Imaging, a title he has held since 2007, and has chaired Yale's Department of Biomedical Engineering since 2022. In April 2026, he gives a UT Southwestern seminar on neuroimage analysis in autism, from model-based estimation to data-driven learning.