AGIR combines 3D motion capture with Large Language Models to assess gait impairments in neurodegenerative diseases with human-like reasoning and explanations.

• Transforms complex 3D movement data into natural language descriptions
• Provides interpretable assessments that explain reasoning, not just predictions
• Achieves clinical-grade accuracy while maintaining transparency
• Bridges the gap between technical measurements and medical decision-making

Why it matters: This approach addresses a critical limitation in current AI diagnostics by providing explainable results that clinicians can trust and understand, potentially accelerating adoption in medical settings.

AGIR: Assessing 3D Gait Impairment with Reasoning based on LLMs