PRIMED-AI: Data-to-Model Academic-Industrial Partnerships (D2M-AIP) for Precision Medicine with AI: Integrating Imaging with Multimodal Data (UG3/UH3 Clinical Trial Optional) | RFA RM 27 012

Opportunity Information: Apply for RFA RM 27 012

PRIMED-AI: Data-to-Model Academic-Industrial Partnerships (D2M-AIP) for Precision Medicine with AI: Integrating Imaging with Multimodal Data is a National Institutes of Health (NIH) cooperative agreement funding opportunity that aims to stand up the broader PRIMED-AI Program. The program is centered on building practical, trustworthy clinical decision support (CDS) tools that use artificial intelligence to combine clinical imaging with other kinds of health data, with the end goal of improving patient care across a wide range of diseases and care settings. A key emphasis is moving beyond single-modality AI (for example, imaging-only models) toward multimodal systems that can synthesize information from images together with complementary clinical signals, so the resulting CDS tools can support more personalized and effective medical decisions.

The main purpose of this specific NOFO is to accelerate the development and testing of AI-enabled, image-centered multimodal CDS tools that are being advanced as Software as a Medical Device (SaMD). In other words, applicants are expected to be building software that is intended for medical use and is being developed with a level of rigor consistent with that kind of product pathway, rather than a purely exploratory research prototype. NIH is looking for projects with a strong chance of producing a measurable, real-world benefit, such as better patient outcomes, improved care delivery, or more efficient clinical processes. The program also explicitly focuses on adoption into clinical workflows, which signals that reviewers will care about usability, integration into existing care environments, and the practical steps needed for implementation, not just model performance on retrospective datasets.

This opportunity uses the UG3/UH3 structure and is listed as Clinical Trial Optional. In general, that kind of phased mechanism is designed to support a milestone-driven progression from an initial stage focused on building, refining, and demonstrating feasibility (UG3) into a later stage focused on expanded development and more advanced testing or validation (UH3), with clear go/no-go decision points. Because clinical trials are optional, teams can propose studies that do or do not meet NIHs definition of a clinical trial, depending on what is necessary to validate the CDS tool and demonstrate its impact. The funding instrument is a cooperative agreement, which typically means NIH expects to have substantial programmatic involvement during the project period, often through close coordination around milestones, data/resource sharing expectations, and alignment with broader program goals.

The target outputs are multimodal AI-based CDS tools that integrate imaging with additional health data types. While the NOFO summary does not enumerate every modality, multimodal health data commonly includes elements such as electronic health record variables, laboratory results, pathology data, genomics or other omics, physiologic signals, medications, demographics, clinical notes, and longitudinal outcomes. The intent is that imaging remains central, but the model and the resulting decision support are strengthened by incorporating the broader clinical context. NIH also highlights reliability, cost-effectiveness, and sustainability, suggesting that proposals should address issues like model robustness across sites and populations, drift monitoring, maintainability, compute and deployment costs, and plans for updating models over time in realistic healthcare environments.

Eligibility is broad and spans many organization types, including state, county, and local governments; special districts; independent school districts; public and state-controlled universities; private institutions of higher education; federally recognized tribal governments and other tribal organizations; public housing authorities/Indian housing authorities; nonprofit organizations (with or without 501(c)(3) status); for-profit organizations (including those other than small businesses); small businesses; and other entities as allowed by NIH policy. The opportunity also explicitly permits international participation: non-U.S. entities (foreign organizations) may apply, non-U.S. components of U.S. organizations are eligible, and foreign components are allowed as defined by the NIH Grants Policy Statement. That makes it feasible to propose cross-border academic-industrial collaborations, multinational validation efforts, or inclusion of international clinical datasets, provided the application meets NIH requirements and can justify the value of the foreign component.

Administratively, this is an NIH discretionary funding opportunity (CFDA 93.310) under Funding Opportunity Number RFA-RM-27-012. NIH anticipates making around 8 awards. The original closing date listed is 2026-10-19, with a creation date of 2026-06-30. The ceiling amount is not specified in the summary provided, so applicants would need to consult the full NOFO for budget limits, project period expectations, and any program-specific caps or milestone-related funding constraints.

At a high level, the projects that fit best are those that can credibly connect the full chain from data to model to clinical use: assembling or accessing appropriate imaging-plus-multimodal datasets, building and validating models that are reliable and generalizable, packaging the result into a CDS tool suitable for real clinical environments, and generating evidence that the tool improves decisions and care processes. The program framing makes clear that NIH is not only funding algorithm development, but also pushing toward deployable, clinically meaningful AI that can realistically be adopted in workflow and sustained over time.

• The National Institutes of Health in the health sector is offering a public funding opportunity titled "PRIMED-AI: Data-to-Model Academic-Industrial Partnerships (D2M-AIP) for Precision Medicine with AI: Integrating Imaging with Multimodal Data (UG3/UH3 Clinical Trial Optional)" and is now available to receive applicants.
• Interested and eligible applicants and submit their applications by referencing the CFDA number(s): 93.310.
• This funding opportunity was created on 2026-06-30.
• Applicants must submit their applications by 2026-10-19. (Agency may still review applications by suitable applicants for the remaining/unused allocated funding in 2026.)
• The number of recipients for this funding is limited to 8 candidate(s).
• Eligible applicants include: State governments, County governments, City or township governments, Special district governments, Independent school districts, Public and State controlled institutions of higher education, Native American tribal governments (Federally recognized), Public housing authorities/Indian housing authorities, Native American tribal organizations (other than Federally recognized tribal governments), Nonprofits having a 501 (c) (3) status with the IRS, other than institutions of higher education, Nonprofits that do not have a 501 (c) (3) status with the IRS, other than institutions of higher education, Private institutions of higher education, For-profit organizations other than small businesses, Small businesses, Others.

Apply for RFA RM 27 012

[Watch] Creating a grant proposal using the step-by-step wizard inside the applicant portal: