ARPA-H announces medical imaging data partnership with FDA 

The joint effort seeks expert feedback on the challenges of identifying and accessing large amounts of labeled, curated, and representative medical imaging data, which is critical in developing novel diagnostic and treatment products.  

The Advanced Research Projects Agency for Health (ARPA-H) today announced a partnership with the Center for Devices and Radiological Health (CDRH) of the Food and Drug Administration (FDA), both agencies within the U.S. Department of Health and Human Services (HHS), to streamline access to affordable, high-quality medical imaging data.  

Imaging data–such as those derived from digital pathology or medical scans such as MRI, CT, PET, and SPECT–are a critical resource needed to train artificial intelligence (AI) and machine learning (ML) applications. The agencies seek to develop a medical imaging data marketplace (MIDM) to accelerate AI and ML innovation by removing barriers to obtaining data that align with regulatory quality standards and appropriately represent the relevant portions of the U.S. population.  

The MIDM is the first joint effort between ARPA-H and the FDA, bringing each agency’s unique expertise to address common barriers to innovation and efficient pre-market review of AI-enabled medical devices and Software as a Medical Device (SaMD). AI and SaMD are growing categories in an increasingly digital health care environment. The MIDM effort will include developing new tools that may give users of the marketplace greater confidence that the data accessed will adhere to the regulatory requirements for future pre-market authorizations. Within CDRH, the Office of Science and Engineering Labs (OSEL) will develop and establish the operating procedures and governance structure for the MIDM.  

“How many times have we struggled, as researchers, to locate and curate datasets that were representative of the condition we were looking to diagnose? And how many companies have had to go back to the drawing board because their FDA pre-market submission was returned with requests for broader diversity in the training and/or testing datasets?” said ARPA-H Program Manager Ileana Hancu, Ph.D. “The medical imaging data marketplace aims to make this a thing of the past by ensuring fast access to research data and facilitating more streamlined regulatory approvals.”  

"In today’s era of AI and machine learning in medical device applications, the importance of medical imaging data access cannot be understated,” said FDA Division of Imaging, Diagnostics, and Software Reliability/OSEL/CDRH Director Aldo Badano. “It is also essential that data be aligned with regulatory requirements, helping to ensure the quality and representativeness of data, to enable rigorous evaluation of the safety and effectiveness of AI devices for the intended patient populations." 

The first phase of this initiative, which is facilitated through the ARPANET-H Health Innovation Network, will deploy a survey requesting input from stakeholders on the design and development of a sustainable, affordable, diverse, and secure federated medical imaging data platform. This feedback will be used by ARPA-H as it considers a future program on a medical imaging data marketplace. 

To do so, the MIDM will leverage the nationwide network of the ARPANET-H Investor Catalyst Hub, collecting feedback about specific needs and challenges that medical imaging software and product developers, users of AI/ML medical imaging products, and private and public organizations face with utilizing data for product development and evaluation.  

“The advancement of artificial intelligence and machine-learning applications holds tremendous potential to increase the utility of medical imaging data, but ensuring this data aligns with regulatory standards and is representative of the U.S. population is a unique challenge,” said ARPA-H Deputy Director Susan Monarez. “ARPA-H is built to take on this challenge by bringing together major players across the health innovation ecosystem, including the FDA, to ensure these growing technologies have a path forward in improving health care. ARPA-H looks forward to working with the FDA to establish a medical imaging data marketplace to support the advancement of AI and ML in providing better health outcomes for all Americans.” 

The Network Survey can be found on the Investor Catalyst Hub website. All organizations with relevant experience and expertise are invited to respond, including those beyond the current Investor Catalyst Hub network.

The deadline to provide feedback is April 16, 2024. Following the feedback period, ARPA-H will aggregate the provided information and share a report summarizing the findings from the survey. The data provided will be incorporated into the program design and strategy for the future MIDM effort. 

For more information about ARPANET-H, the agency's nationwide health innovation network that connects people, innovators, and institutions to accelerate better health outcomes for everyone, visit the ARPANET-H webpage.  

For more information about the FDA CDRH and its mission to protect and promote public health by ensuring patients and providers have timely and continued access to safe, effective, and high-quality medical devices and safe radiation-emitting products, visit the CDRH webpage. 

Organizations interested in becoming a part of the Investor Catalyst Hub consortium – or spoke network – can learn more at the How to Become a Spoke webpage on the Investor Catalyst Hub website.