Fueling HLBS Discovery with AI
The NHLBI-AI Initiative: A Transformative Opportunity for CAAI and the University of Kentucky
The NHLBI-AI Initiative: Advancing HLBS Research
The National Heart, Lung, and Blood Institute (NHLBI) has launched a landmark initiative to harness Artificial Intelligence, aiming to revolutionize discovery and clinical applications in Heart, Lung, Blood, and Sleep (HLBS) health. This strategic effort seeks to build a collaborative national ecosystem for AI-driven HLBS research.
Key Investment
Total funding over five years to catalyze AI innovation in HLBS domains.
This substantial investment underscores the national priority to integrate AI into HLBS research and care, fostering transformative breakthroughs.
Funding Allocation
Distribution of resources within the initiative.
Core Goals & Components
The initiative is designed to create a synergistic environment by integrating various components and pursuing ambitious goals:
- Transformative AI Models: Develop AI that revolutionizes HLBS discovery research.
- High-Value Datasets: Provide access to unique, machine-readable biomedical data.
- Vetted AI Tools: Offer a suite of secure and effective AI tools.
- Modular Research Programs: Support deep phenotyping and systems biology approaches.
- Collaborative Ecosystem: Foster synergy between AI/data scientists and HLBS researchers.
- Ethical AI: Ensure trustworthy, ethical, and responsible AI development.
Initiative Structure
These components will work in concert to achieve the initiative’s goals, managed under an Other Transactions (OT) Agreement with Westat, Inc. as the AICC.
University of Kentucky’s CAAI: An AI Powerhouse
The Center for Applied AI (CAAI) at the University of Kentucky is uniquely positioned to contribute significantly to the NHLBI-AI Initiative. Its mission is to apply AI to improve health outcomes, advance research, and develop the future workforce, bridging advanced technology and translational science.
CAAI’s Core Strengths for HLBS Innovation
CAAI brings a robust portfolio of technological capabilities, a proven track record in HLBS-relevant AI research, and a strong collaborative network, making it an ideal partner for the NHLBI-AI initiative.
??? Secure In-House Infrastructure
CAAI operates with significant independence, maintaining its data, computational processing, and operational services within highly secure, NIST-compliant, and HIPAA-compliant facilities, including a DGX computing cluster. This is crucial for handling sensitive HLBS data securely.
This self-operated environment offers enhanced data security and privacy assurances, a key differentiator for managing diverse and sensitive datasets.
?? Agentic AI Leadership
CAAI is at the forefront of Agentic AI development, creating autonomous systems that perceive, decide, and act. This cutting-edge specialization can drive innovation in personalized HLBS care and research automation.
?? Key AI Platforms & Services
LLM Factory
A secure, locally hosted platform for training, configuring, and deploying multi-modal Large Language Models with an OpenAI compatible API.
CLASSify
An easy-to-use web platform for training and evaluating ML classification models on tabular data, requiring no technical expertise.
SmartState
An automated protocol adherence platform using virtual agents to guide participants, send reminders, and document progress in research studies.
These platforms, along with CAT-Talk (transcription) and AI-Ready Data Trusts, form a comprehensive AI toolkit ready to support the NHLBI-AI consortium.
?? HLBS-Relevant Project Showcase
CAAI has a strong portfolio of projects directly applicable to Heart, Lung, Blood, and Sleep research, demonstrating translational impact.
HeartLens
AI for cardiovascular diagnostics, including Coronary Artery Calcification scoring from CT scans using Transformer models.
AUD in ARDS
Applying data science and LLMs to study Alcohol Use Disorder in Acute Respiratory Distress Syndrome (ARDS).
Germline Cancer Genetic Testing Automation
Automating data extraction for germline cancer genetic testing, relevant to hematologic conditions.
Intermittent Fasting & Sleep Protocol Management
Utilizing SmartState to manage research protocols related to intermittent fasting and sleep patterns.
These projects, among many others, highlight CAAI’s deep collaborations with UK HealthCare (Cardiology, Pulmonary, Markey Cancer Center) and its ability to translate AI into clinical practice.
Strategic Response: CAAI’s Hybrid Approach
CAAI is recommended to pursue a hybrid strategy, maximizing its potential impact by applying for both an AI Data Science Center (DSC) award and developing highly focused individual AI research projects. This dual approach leverages CAAI’s strengths as both an enabler and an innovator.
?? AI Data Science Center (DSC) Proposal
Leverage CAAI’s secure infrastructure, existing platforms (LLM Factory, CLASSify, SmartState), and collaborative expertise to serve as a national hub for AI tool development, data coordination, training, and ELSI oversight for the NHLBI-AI consortium.
Focus: Enablement, Integration, Leadership.
?? Individual AI Research Projects (1-2)
Develop high-impact projects in areas of unique CAAI strength, such as:
- Agentic AI for chronic HLBS disease management.
- Foundation models for multi-modal HLBS data (using LLM Factory & TOPMed/BDC).
- Advanced predictive analytics for cardiovascular or pulmonary conditions.
Focus: Transformative Discovery, Niche Innovation.
This coordinated strategy positions CAAI to significantly contribute to the NHLBI-AI initiative by providing foundational support through a DSC while also pushing the boundaries of HLBS research with targeted projects.
Roadmap to Impact: Key Milestones
A rapid and organized approach is essential to successfully respond to the NHLBI-AI-25-001 solicitation.
Finalize strategic decisions, form core writing teams, identify PIs.
Develop LOR outlines, draft content, engage collaborators.
Letter of Request (LOR) Due Date.
Notification of Meritorious LOR / Invitation for Full Application.
If invited, commence full proposal development (budgets, workplans, ELSI, data sharing).
Full Application Due Date.
Anticipated Funding Decision: December 1, 2025.
UK & CAAI: Leading the Future of AI in HLBS
With its robust infrastructure, innovative platforms, deep expertise, and strong collaborative spirit, the University of Kentucky’s Center for Applied AI is exceptionally prepared to make transformative contributions to the NHLBI-AI Enabled Precision Medicine Initiative. By pursuing a strategic hybrid approach, CAAI aims to become a cornerstone of this national endeavor, driving discovery and improving Heart, Lung, Blood, and Sleep health for all.
Learn More About CAAI