GAITGen: Disentangled Motion-Pathology Impaired Gait Generative Model -- Bringing Motion Generation to the Clinical Domain

2025-03-28

AI Summary

Clinical gait analysis for movement disorders like Parkinson’s disease (PD) suffers from scarce ground-truth annotations, poor model generalizability, and bias propagation. Method: We propose the first pathology-severity-conditioned, high-fidelity gait sequence generation framework. It innovatively decouples motor dynamics from pathological features via a conditional residual vector-quantized VAE (C-RVQ-VAE) paired with a mask-residual Transformer, jointly regulated by fine-grained pathology severity embeddings. Contribution/Results: On the real-world PD-GaM clinical dataset, our method surpasses state-of-the-art methods in both reconstruction fidelity and synthetic gait quality. Clinical experts validate the generated gaits as highly realistic and pathologically consistent. Downstream PD severity estimation achieves significant performance gains, establishing a novel paradigm for low-resource clinical AI modeling.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Intelligent Robots: Motion and Path Planning
• Natural Language Processing: Generation

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Web Mining and Content Analysis: Web data generation and simulation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Gait analysis is crucial for the diagnosis and monitoring of movement disorders like Parkinson's Disease. While computer vision models have shown potential for objectively evaluating parkinsonian gait, their effectiveness is limited by scarce clinical datasets and the challenge of collecting large and well-labelled data, impacting model accuracy and risk of bias. To address these gaps, we propose GAITGen, a novel framework that generates realistic gait sequences conditioned on specified pathology severity levels. GAITGen employs a Conditional Residual Vector Quantized Variational Autoencoder to learn disentangled representations of motion dynamics and pathology-specific factors, coupled with Mask and Residual Transformers for conditioned sequence generation. GAITGen generates realistic, diverse gait sequences across severity levels, enriching datasets and enabling large-scale model training in parkinsonian gait analysis. Experiments on our new PD-GaM (real) dataset demonstrate that GAITGen outperforms adapted state-of-the-art models in both reconstruction fidelity and generation quality, accurately capturing critical pathology-specific gait features. A clinical user study confirms the realism and clinical relevance of our generated sequences. Moreover, incorporating GAITGen-generated data into downstream tasks improves parkinsonian gait severity estimation, highlighting its potential for advancing clinical gait analysis.

Problem

• Generates realistic gait sequences for clinical analysis
• Addresses limited datasets in parkinsonian gait evaluation
• Improves accuracy in pathology-specific gait severity estimation

Innovation

• Conditional Residual VQ-VAE for disentangled representations
• Mask and Residual Transformers for sequence generation
• Generates realistic gait sequences by severity levels