The COMPWALK-ACL: A Dataset of Multi-pace IMU Gait Kinematics in Adolescents, Adults, and ACL Injured Patients

TL;DR

This paper introduces a dataset of gait kinematics collected using wearable IMUs from adolescents, adults, and ACL-injured patients walking at different speeds.

Contribution

The novel contribution is a diverse, multi-pace IMU gait dataset for studying normative and injured movement patterns.

Findings

The dataset includes spatiotemporal and kinematic data from 92 participants across three groups.

It supports research on gait differences due to age, ACL injury, and walking speed.

Follow-up data from ACL patients three months post-surgery is included.

Abstract

Gait analysis provides objective, quantitative parameters essential for assessing mobility, identifying movement impairments, and monitoring the progress of rehabilitation. While traditional lab-based systems offer high accuracy, wearable Inertial Measurement Units (IMUs) enable portable, cost-effective gait assessments outside the laboratory environment. However, the reliability and applicability of IMU-derived data across diverse populations and walking conditions require robust datasets. This paper presents a lower limb kinematic dataset acquired with the Xsens Awinda IMU system. Data were collected from 92 unique participants: healthy adults (n = 25), healthy adolescents (n = 27), and individuals with ACL injuries assessed before surgery (n = 40), with 27 completing a follow-up three months post-reconstruction. Participants walked overground at self-selected slow, normal, and fast speeds. The dataset contains spatiotemporal parameters, as well as lower limb joint kinematics. It enables research on normative gait across age groups, the effects of ACL injury and early recovery on movement patterns, and the development of IMU-based gait analysis methods under different walking speeds and clinical conditions.

Figures