OAAgent: Multimodal LLM Agent for Predicting Knee Osteoarthritis Progression

Abstract

Osteoarthritis (OA) is a leading cause of disability worldwide, necessitating early prediction of its progression to enable timely interventions and personalized treatment strategies. This study introduces OAAgent, a novel AI-driven framework designed to forecast knee osteoarthritis progression by leveraging clinical and imaging data from the FNIH Biomarkers Consortium. Unlike traditional predictive models, OAAgent integrates multimodal data—including demographic variables, clinical features, and X-ray imaging biomarkers—into a reasoning based framework. This framework combines Chain-of-Thought (CoT) reasoning with task execution, enabling interpretable decision pathways that align with clinical workflows.

Ähnliche Arbeiten