AI-Enabled Tools, Projects and Capabilities

The HEAL Core Lab uses AI-enabled tools for radiomic integration with exposomic risk, AI-enabled spatial phenotyping to study neighborhood-level risks and prediction tools, large language models (LLM) and vision language models (VLM)-based risk prediction from heterogenous patient data and knowledge graph-leveraged AI-assisted data extraction, synthesis and dashboard generation.

CARE-GAP AI Initiatives

There are currently many gaps in translating evidence to improving outcomes at the individual level. AI with human-in-the-loop designs can augment and improve implementation gaps in known hard to resolve failure points.

Technical Capabilities and Resources

At the forefront of the convergence between data, intelligence and care, our team brings deep expertise across the full continuum of biomedical data and AI integration. We harness rich multimodal datasets, including EHR data, advanced imaging, EKG and device-derived signals from clinical and personal monitoring sources to build meaningful intelligence pipelines. Our services encompass custom machine learning, LLM incorporation, and advanced reasoning frameworks that accelerate insight generation from complex biomedical data. Through robust integration tools and clinical decision support systems, we enable seamless consumption of these AI capabilities within existing workflows, transforming data into actionable knowledge and advancing precision health at scale.

• Expertise in Imaging and Data Sources - EHR (Epic), Imaging, EKG, Devices, Personal Monitoring Devices
• Services - Custom ML/LLM pipelines, NLP agents
• Acceleration - Biomedical LLMs and VLMs, ML models, custom fine-tuned open-source models
• Consumption - Integration tools, clinical decision support