Problem
Research questions and friction points this paper is trying to address.
gait analysis
post-stroke rehabilitation
multimodal data integration
clinical reporting
cognitive workload
OGA-AID: Clinician-in-the-loop AI Report Drafting Assistant for Multimodal Observational Gait Analysis in Post-Stroke Rehabilitation
🤖 AI Summary
This study addresses the clinical challenges in post-stroke gait rehabilitation—namely, the reliance on time-consuming, multi-source data and the absence of structured reporting—by proposing a novel approach that integrates a multi-agent large language model with a clinician-in-the-loop mechanism. The system jointly processes patient motion videos, kinematic trajectories, and clinical text to generate automated, structured gait assessment reports through multimodal fusion. By innovatively combining a multi-agent architecture with real-time therapist feedback, the method significantly outperforms existing single-pass multimodal baselines on real patient data. Furthermore, incorporating brief therapist notes enhances assessment accuracy, thereby strengthening human–AI collaborative decision-making in clinical practice.
📝 Abstract
Gait analysis is essential in post-stroke rehabilitation but remains time-intensive and cognitively demanding, especially when clinicians must integrate gait videos and motion-capture data into structured reports. We present OGA-AID, a clinician-in-the-loop multi-agent large language model system for multimodal report drafting. The system coordinates 3 specialized agents to synthesize patient movement recordings, kinematic trajectories, and clinical profiles into structured assessments. Evaluated with expert physiotherapists on real patient data, OGA-AID consistently outperforms single-pass multimodal baselines with low error. In clinician-in-the-loop settings, brief expert preliminary notes further reduce error compared to reference assessments. Our findings demonstrate the feasibility of multimodal agentic systems for structured clinical gait assessment and highlight the complementary relationship between AI-assisted analysis and human clinical judgment in rehabilitation workflows.
Innovation
Methods, ideas, or system contributions that make the work stand out.
clinician-in-the-loop
multimodal gait analysis
multi-agent LLM
structured clinical reporting
post-stroke rehabilitation
🔎 Similar Papers
No similar papers found.
