Model-to-Clinic (M2C) for Precision Medicine with AI: Integrating Imaging with Multimodal Data (PRIMED-AI) (UG3/UH3, Clinical Trial Optional)

Companion Funding Opportunity

RFA-RM-27-011 , U01 Research Project (Cooperative Agreements)
RFA-RM-27-012 , UG3/ UH3 Phase 1 Exploratory/Developmental Cooperative Agreement/Exploratory/Developmental Cooperative Agreement Phase II
RFA-RM-27-014 , U54 Specialized Center (Cooperative Agreements)
RFA-RM-27-015 , U24 Resource-Related Research Project (Cooperative Agreements)

The Precision Medicine with AI: Integrating Imaging with Multimodal Data (PRIMED-AI) NOFOs seek to spur on the development of innovative, reliable, cost-effective, and sustainable multimodal AI-based clinical decision support (CDS) tools with the potential for transformational impact. Overall, the PRIMED-AI Program catalyzes the integration of clinical imaging with other types of multimodal health data that informs CDS tool development and testing in clinical workflows, which serves to enhance patient care for a wide range of health conditions.

The Model-to-Clinic (M2C) NOFO uniquely seeks projects that take CDS tools developed as Software as a Medical Device (SaMD), from validated prototypes to clinical applications. 

M2C projects are structured as phased innovation awards (UG3/UH3) designed to advance multimodal AI-based CDS tools through progressive stages of development and validation. 

UG3 Phase (Exploratory/Developmental): This initial phase focuses on establishing the technical and operational foundation for clinical translation. Projects will develop and refine AI models, regulatory preparations, clinical implementation infrastructure, deployment strategies, and demonstrate preliminary feasibility in target clinical settings. Milestones must be achieved before transitioning to the UH3 phase.

UH3 Phase (Advanced Development): Following successful completion of the UG3 phase, the UH3 phase emphasizes CDS model clinical implementation and validation. Projects will deploy CDS tools in clinical workflows; conduct prospective clinical studies to evaluate performance; demonstrate potential for impact and integration with existing health IT systems; and generate evidence for broader adoption and sustainability. Projects must demonstrate measurable potential for clinical impact and pathway(s) to sustained use beyond the award period.

The combined UG3/UH3 award period cannot exceed 5 years total. Transition from UG3 to UH3 requires achieving specified milestones and NIH approval.

M2C projects are expected to have high potential for demonstrable, positive impact on patient outcomes and/or healthcare processes.

Key Terms used in PRIMED-AI Program

• Clinical decision support (CDS) tool.  A type of software, computational model or digital system that is incorporated into clinical workflows to assist in determining a course of action related to patient care. 
• Clinical imaging.  Any FDA-approved imaging modality used in patient care, including radiologic (e.g., radiographic, computed tomographic, magnetic resonance, molecular, radionuclide imaging), ophthalmologic (e.g., Optical Coherence Tomography), endoscopic, and dermatologic imaging, and video.  Clinical imaging of human participants is intended to be the anchor data type that multimodal data are integrated with in the PRIMED-AI Program, which will form the basis for AI algorithm development and testing of CDS tools.  
• DICOM standard.  Digital Imaging and Communications in Medicine (DICOM) standard, the most widely used by the community to address interoperability challenge, is strongly encouraged but not required.  Inclusion of non-DICOM standard clinical imaging must include a plan to develop standards in conjunction with the PRIMED-AI community if none currently exist.
• Harmonization. The process of bringing together data from different sources and ensuring that it is consistent, comparable, and compatible. This involves standardizing data formats, structures, and definitions so that data from various sources can be integrated and analyzed together effectively.
• Interoperability.  The ability for AI models and associated data and metadata to be understood and work across different AI platforms and have the potential to be used consistently across different health systems.
• Multimodal data (MMD). Representing different types of data and information from multiple sources that may include multiple clinical imaging modalities and non-imaging health data (e.g., electronic health records, EEG, EKG, laboratory test results (-omics), wearable sensor data, medical reports).  Multiscale data are encouraged; however, microscopy-based imaging of biospecimens ex vivo (e.g., digital pathology) cannot represent the sole imaging data type.  Although non-human imaging and/or MMD data may have assisted in development of an AI-model, overt representation and reliance on data derived from non-human sources for CDS tool development, testing, and validation will be given low programmatic priority.
• Playbook.   A collection of actionable guidelines, standardized protocols, and/or standardized operating procedures for the reliable and effective development and deployment of multimodal clinical decision support tools. The Playbook is a collection of frameworks.
• Precision Medicine. Sometimes called personalized medicine or individualized medicine, refers to a healthcare approach that uses information based on a patient's individual characteristics such as health measures, genotype, phenotype, environment, and lifestyle information to guide, tailor, and optimize decisions related to their medical care and management.
• PRIMED-AI Consortium. The consortium constitutes members of PRIMED-AI excluding NIH program staff.  PRIMED-AI Program is an umbrella term encompassing the consortium, NIH staff, and overall programmatic objectives. 
• Uncertainty Quantification.Measuring or quantifying the impact of uncertainties in complex systems, including quantifying the confidence in outcomes predicted by multimodal AI models. 
• Validation. Validation exists on a continuum in the PRIMED-AI Program.  Analytical or technical validation is based on the evaluation of algorithmic performance and the ability of a multimodal AI model to make accurate predictions.  Initially, a model or algorithm can meet expected performance on retrospective and/or entirely new clinical datasets within the confines of a specific hospital or healthcare system. It is useful locally (internally) but is not yet applicable (generalizable) to the wider real-world population. Subsequently, for clinical validation, a model or algorithm can be tested (externally) on new wider real-world population datasets to predict a meaningful outcome and meet regulatory criteria for the claimed use case.  The PRIMED-AI Program anticipates validation of projects along this continuum as outlined in the NOFOs.
• Verification. The process by which data integrity and construction of models is assessed for appropriateness within the context of use or intended purpose.