Welcome to the March 2026 issue of the
Medical Imaging and Data Resource Center (MIDRC) newsletter!

Two New MIDRC Datasets Now Available:
Spine CT and Shoulder MRI

The Medical Imaging & Data Resource Center (MIDRC) has released two new large-scale, expertly annotated imaging datasets: Spine CT and Shoulder MRI.

These datasets were created to help the research community tackle clinically important and technically challenging problems. Both include data from multiple institutions and scanners, with high-quality expert annotations—making them well suited for training, validation, and benchmarking medical imaging AI models.

Why these datasets matter

• Focus on common but complex clinical imaging problems

• Include expert annotations for reliable model development

• Use data from multiple institutions to support generalizable research

Why they’re exciting

• Spine CT enables research on injury, degeneration, and spine structure

• Shoulder MRI supports challenging soft-tissue analysis

Both datasets are available through the MIDRC data portal. Watch the seminar video to hear directly from the team about how the datasets were built and how they can be used.

MIDRC “by the numbers” highlights the scale, diversity, and accessibility of its imaging datasets. Explore the data and see how it can support your research!

Our Biomedical Imaging Hub

The MIDRC BDF Imaging Hub (BIH) MIDRC's federated platform within the ARPA-H Biomedical Data Fabric, makes it easier than ever to discover biomedical imaging data across multiple repositories. Rather than storing imaging files itself, BIH aggregates structured metadata and provides unique identifiers that allow researchers to seek data directly from the original source nodes. With thoughtful validation and documentation, BIH offers a powerful, transparent foundation for reproducible imaging research - check it out!

MIDRC Awarded $1.5M Grant to
Expand Global AI-Ready Data Commons

The Medical Imaging and Data Resource Center (MIDRC) was awarded a generous $1.5 million grant (grant #: 2024-354607) from the Chan Zuckerberg Initiative DAF, an advised fund of Silicon Valley Community Foundation. This grant is fueling the development of MIDRC-Global, a groundbreaking initiative aimed at building and expanding an AI-ready data commons to transform medical imaging research on a global scale.

To enhance the diversity and global relevance of our dataset, MIDRC-Global is expanding our data collection efforts beyond the United States. We are expanding our portfolio to include global partners, sites in Kenya and Ghana, who are contributing imaging and metadata, as well as expertise, to MIDRC. The enhanced data commons will empower AI researchers to mitigate biases as well as design AI algorithms for specific populations. The ability to share and reuse MIDRC’s data has a big impact on both AI/ML research, at scale (through the use of the open commons), and on the validation and translation of developed algorithms (through the use of the sequestered commons).

This funding will directly support:

• Expanding data collection efforts to create a more comprehensive research resource.

• Enhancing data organization and processing to streamline AI development.

• Improving the MIDRC user portal to enable broader access and collaboration.

By enriching MIDRC’s data commons with high-quality medical imaging datasets from across the globe, AI researchers will have unprecedented opportunities to improve the reliability, accuracy, and applicability of machine learning algorithms.

Through open-access resources and innovative data-sharing mechanisms, MIDRC-Global will help drive groundbreaking advancements in AI-powered medical research, fostering new discoveries that will benefit patients and healthcare providers around the world.

Stay tuned for more updates as we embark on this exciting journey to shape the future of AI in medical imaging!

Introducing MIDRC-Learn:
An Education Hub for Interoperable Medical Imaging Data

The Medical Imaging and Data Resource Center (MIDRC) has launched MIDRC-Learn, a centralized learning portal designed to educate diverse audiences on interoperable medical imaging data, data commons, and AI-ready research infrastructure. The platform, which is being developed throughout this 2026 year, will translate complex technical concepts into accessible, role-specific resources that support both education and hands-on research.

MIDRC-Learn aims to offers tiered content for multiple audiences, including the general public, data contributors, multidisciplinary researchers, and highly technical AI developers. Materials will span training videos, infographics, detailed documentation, Jupyter notebooks, hands-on exercises, and interactive workspaces. Core topics include common data elements (CDEs), data models, interoperability, privacy-preserving record linkage, and federated learning.

The MIDRC-Learn portal is supported through the National AI Research Resource (NAIRR) Secure Pilot, co-led by the National Institutes of Health and the U.S. Department of Energy. As part of this effort, MIDRC-Learn provides educational resources that demonstrate how secure, privacy-preserving interoperable imaging and clinical data can accelerate AI development while maintaining rigorous security standards.

By combining conceptual education with practical, reproducible tools, MIDRC-Learn supports workforce development, lowers barriers to entry, and enables researchers to more effectively leverage interoperable data for high-impact biomedical discovery!

MIDRC Featured at NIH ODSS
Data Science Town Hall

The Medical Imaging and Data Resource Center (MIDRC) was invited by the NIH Office of Data Science Strategy (ODSS) to present at the January ODSS Data Science Town Hall, held on January 22, 2026. MIDRC served as the sole featured presenter for the January Town Hall, highlighting the program’s significance to NIH-wide data science efforts.

MIDRC delivered a presentation outlining the program’s mission, infrastructure, and scientific impact, followed by a Q&A session with ODSS and NIH data science leadership. The session opened and closed with remarks from ODSS leadership, including Dr. Susan Gregurick.

Although the January Town Hall was a closed session, a public, condensed version of the presentation will be shared with the broader community during the March MIDRC Seminar on March 17th at 2:00 PM CT.

The MIDRC Seminar Series provides an opportunity for the medical imaging and research community to hear directly from MIDRC investigators on new datasets, methods, and advances, followed by a live Q&A.

This virtual event is free and open to all! Register here!

MIDRC at SPIE Medical Imaging 2026!
Vancouver, Canada from February 15–19, 2026

MIDRC Paper Receives Award at SPIE Medical Imaging Conference!

We are proud to announce that a research paper supported by the Medical Imaging and Data Resource Center (MIDRC) received a Runner-Up Award for Best Paper at the SPIE Imaging Informatics Conference, held last week as part of the SPIE Medical Imaging meeting.

The paper, “Large-scale data harmonization of radiology studies via multigranular vision–language alignment,” was authored by Zihan Li and collaborators, part of the MIDRC work by Dr. Paul Kinahan. The work presents a new approach for harmonizing large, multi-institutional radiology datasets by leveraging vision–language models, helping address inconsistencies in imaging metadata across sites.

This recognition highlights the impact of MIDRC-supported research in advancing scalable, data-driven solutions for medical imaging and informatics.

MIDRC at RSNA Annual Meeting 2025!
Chicago, IL from November 30 – December 4, 2025

Should you have any questions or wish to discuss facilitating a data contribution from your institution in support of MIDRC's mission, please contact us here

MIDRC's open source web-based tools can be accessed here

MIDRC's open source downloadable tools can be found here

Our growing data portal is live and can be accessed here

Thank you again for your interest in and support of MIDRC!