AI-Driven Multimodal Risk Inference Engine
At the heart of ActualHealth™ is an AI-powered risk inference engine that combines data from wearables, self-reported health histories, demographics, social determinants of health (SDOH), and clinical literature to provide a comprehensive health overview. Its main goal is to deliver continually updated, personalized disease risk trajectories, moving beyond static assessments to track changes over time to the initial chronic conditions of diabetes, cardiovascular disease, and depression.
Technical Innovation
AI-enabled adaptive data fusion to handle uncertainty across diverse datasets
Using clinical evidence as computable priors for up-to-date risk calculations
Longitudinal modeling with SDOH for dynamic projections
Key innovations include:
ActualHealth™ starts with a deterministic, rules-based system using peer-reviewed disease prevention data. It is evolving into a self-learning probabilistic inference model that updates with real-world multimodal data and translates clinical findings into structured, machine-readable risk models for more tailored predictions.
The system's ability to generalize across additional disease states remains under evaluation:
System Architecture and Evolution
Cardiometabolic
Musculoskeletal, chronic kidney disease, and others
Autoimmune
Lupus and arthritis variants
Oncologic conditions