tsta

Home / سی و دومین کنفرانس ملی و دهمین کنفرانس بین المللی مهندسی زیست پزشکی ایران

Unsupervised Gait Anomaly Detection Using Generative Adversarial Networks: A Feasibility Study

Authors :

Seyed Hooman Hosseini-Zahraei¹ Ali Chaibakhsh²

1- Intelligent Systems and Advanced Control Lab, Faculty of Mechanical Engineering, University of Guilan 2- Faculty of Mechanical Engineering, University of Guilan

Keywords :

Gait Analysis،Gait Analysis،Generative Adversarial Networks (GAN)،Unsupervised Learning،Wearable Sensors،Feasibility Study

Abstract :

Anstract: The automated classification of human gait, a critical indicator of neuromuscular health, is often hindered by the dependency of supervised machine learning on extensive labeled pathological datasets, which are scarce and difficult to obtain. This paper explores the feasibility of a paradigm towards unsupervised learning, proposing a framework based on a Wasserstein Generative Adversarial Network with Gradient Penalty (WGAN-GP). The WGAN-GP is trained exclusively on healthy gait patterns from a single young subject's shank-mounted Inertial Measurement Unit (IMU) to build a model of healthy movement. The framework utilizes a reconstruction-based anomaly detection strategy, where abnormalities are quantified by the magnitude of error when attempting to reconstruct a new gait cycle from the learned healthy model. Evaluated using real-world data from healthy subjects (both young and old) and individuals with Parkinson's disease, the model demonstrated strong performance, achieving an Area Under the Curve (AUC) of 0.96. Notably, the framework also showed sensitivity to non-pathological, age-related gait variations. This feasibility study provides compelling evidence for the efficacy of WGAN-GP-based unsupervised detection as a data-efficient and generalizable alternative, paving the way for future validation on larger clinical datasets to characterize mobility impairments across various disorders and age groups.