A Comprehensive Review of Vision-Based Sensor Systems for Human Gait Analysis

Sensors, Journal Year: 2025, Volume and Issue: 25(2), P. 498 - 498

Published: Jan. 16, 2025

Analysis of the human gait represents a fundamental area investigation within broader domains biomechanics, clinical research, and numerous other interdisciplinary fields. The progression visual sensor technology machine learning algorithms has enabled substantial developments in creation analysis systems. This paper presents comprehensive review advancements recent findings field vision-based systems over past five years, with special emphasis on role vision sensors, algorithms, technological innovations. relevant papers were subjected to using PRISMA method, 72 articles that met criteria for this research project identified. A detailing most commonly used systems, parameters, optimal camera placement, parameter extraction methods is presented analysis. indicate non-invasive depth cameras are gaining increasing popularity field. Furthermore, such as convolutional neural networks (CNNs) long short-term memory (LSTM) networks, being employed frequency. seeks establish foundations future innovations will facilitate development more effective, versatile, user-friendly tools, potential significantly enhance mobility, health, overall quality life. work was supported by [GOBIERNO DE ESPANA/PID2023-150967OB-I00].

Language: Английский

---

Validation of a 3D Markerless Motion Capture Tool Using Multiple Pose and Depth Estimations for Quantitative Gait Analysis

Sensors, Journal Year: 2024, Volume and Issue: 24(22), P. 7105 - 7105

Published: Nov. 5, 2024

Gait analysis is essential for evaluating walking patterns and identifying functional limitations. Traditional marker-based motion capture tools are costly, time-consuming, require skilled operators. This study evaluated a 3D Marker-less Motion Capture (3D MMC) system using pose depth estimations with the gold-standard (MOCAP) measuring hip knee joint angles during gait at three speeds (0.7, 1.0, 1.3 m/s). Fifteen healthy participants performed tasks which were captured by both systems. The MMC demonstrated good accuracy (LCC > 0.96) excellent inter-session reliability (RMSE < 3°). However, moderate-to-high constant biases was observed specific events, due to differences in sample rates kinematic methods. Limitations include use of only limited key points estimation model. shows potential as reliable tool analysis, offering enhanced usability clinical research applications.

Language: Английский

---

Reliability and Validity Examination of a New Gait Motion Analysis System

Sensors, Journal Year: 2025, Volume and Issue: 25(4), P. 1076 - 1076

Published: Feb. 11, 2025

Recent advancements have made two-dimensional (2D) clinical gait analysis systems more accessible and portable than traditional three-dimensional (3D) systems. This study evaluates the reliability validity of measurements using monocular composite camera setups with VisionPose, comparing them to Vicon 3D motion capture system as a reference. Key parameters-including hip knee joint angles, time distance factors-were assessed under normal, maximum speed, tandem conditions during level walking. The results show that intraclass correlation coefficient (ICC(1,k)) for 2D model exceeded 0.969 0.963 parameters. Time-distance parameters demonstrated excellent relative agreement across walking styles, while range showed overall strong agreement. However, accuracy was lower Cronbach's alpha time-distance ranged from 0.932 0.999 (monocular) 0.823 0.998 (composite). In contrast, motion, varied widely, ranging 0.826 0.985 0.314 0.974 coefficients spatiotemporal were greater 0.933 0.837 angle parameters, underscores potential models in applications highlights areas improvement enhance their application scope.

Language: Английский

---

Reliability and validity of current computer vision based motion capture systems in gait analysis: A systematic review

Gait & Posture, Journal Year: 2025, Volume and Issue: 120, P. 150 - 160

Published: April 17, 2025

Language: Английский

---

Estimation of Vertical Ground Reaction Force during Single-leg Landing Using Two-dimensional Video Images and Pose Estimation Artificial Intelligence

Physical Therapy Research, Journal Year: 2024, Volume and Issue: 27(1), P. 35 - 41

Published: Feb. 25, 2024

Assessment of the vertical ground reaction force (VGRF) during landing tasks is crucial for physical therapy in sports. The purpose this study was to determine whether VGRF a single-leg can be estimated from two-dimensional (2D) video image and pose estimation artificial intelligence (AI).

Language: Английский

---

Validity of the estimated angular information obtained using an inertial motion capture system during standing trunk forward and backward bending

BMC Sports Science Medicine and Rehabilitation, Journal Year: 2024, Volume and Issue: 16(1)

Published: July 17, 2024

Abstract Background Bending the trunk forward and backward while standing are common daily activities can have various patterns. However, any dysfunction in these movements considerably affect living activities. Consequently, a comprehensive evaluation of spinal motion during precise identification movement abnormalities important to facilitate an effective rehabilitation. In recent years, with development measurement technology, patterns using inertial capture system (motion sensor) has become easy. accuracy estimated angular information obtained via sensor measurements be affected by velocity. This study aimed compare validity assessing bending at different speeds three-dimensional analysis system. Methods The current included 12 healthy older men. A were used for measurement. participants performed comfortable maximum speeds, five sensors attached their spine. Statistical was paired t-test, intraclass correlation coefficient, mean absolute error, multiple coefficient. Results showed that each not velocity had high validity. Conclusions Therefore, this applied clinically objective

Language: Английский

---

Gait signature changes with walking speed are similar among able-bodied young adults despite persistent individual-specific differences

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Aug. 26, 2024

Understanding individuals' distinct movement patterns is crucial for health, rehabilitation, and sports. Recently, we developed a machine learning-based framework to show that "gait signatures" describing the neuromechanical dynamics governing able-bodied post-stroke gait kinematics remain individual-specific across speeds. However, only evaluated signatures within limited speed range number of participants, using sagittal plane (i.e., 2D) joint angles. Here characterized changes in wide speeds, from very slow (0.3 m/s) exceptionally fast (above walk-to-run transition speed) 17 young adults. We further assessed whether 3D kinematic and/or kinetic (ground reaction forces, moments, powers) data would improve discrimination signatures. Our study showed remained walking speeds: Notably, achieved exceptional accuracy (99.8%, confidence interval (CI) 99.1–100%) classifying individuals, surpassing both 2D kinetics. Moreover, participants exhibited consistent, predictable linear their entire range. These were associated with participants' preferred balance ability, cadence, step length. findings support as tool characterize individual differences predict speed-induced dynamics.

Language: Английский

---

Validity Verification of Human Pose-Tracking Algorithms for Gait Analysis Capability

Sensors, Journal Year: 2024, Volume and Issue: 24(8), P. 2516 - 2516

Published: April 14, 2024

Two-dimensional (2D) clinical gait analysis systems are more affordable and portable than contemporary three-dimensional (3D) models. Using the Vicon 3D motion capture system as standard, we evaluated internal statistics of Imasen open-source OpenPose measurement systems, both designed for 2D input, to validate their output based on similarity results legitimacy inner statistical processes. We measured time factors, distance joint angles hip knee joints in sagittal plane while varying speeds gaits during level walking three in-person experiments under normal, maximum-speed, tandem scenarios. The intraclass correlation coefficients models were greater 0.769 all parameters compared with those Vicon, except some angles. relative agreement was excellent time–distance parameter moderate-to-excellent each contraction range, high Cronbach’s alpha 0.899–0.993 but low 0.298–0.971. Correlation 0.571 lower angle parameters, particularly Our study elucidates areas which improve widespread application.

Language: Английский

---

Gait signature changes with walking speed are similar among able-bodied young adults despite persistent individual-specific differences

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: May 3, 2024

Abstract Understanding individuals’ distinct movement patterns is crucial for health, rehabilitation, and sports. Recently, we developed a machine learning-based framework to show that “gait signatures” describing the neuromechanical dynamics governing able-bodied post-stroke gait kinematics remain individual-specific across speeds. However, only evaluated signatures within limited speed range number of participants, using sagittal plane (i.e., 2D) joint angles. Here characterized changes in wide speeds, from very slow (0.3 m/s) exceptionally fast (above walk-to-run transition speed) 17 young adults. We further assessed whether 3D kinematic and/or kinetic (ground reaction forces, moments, powers) data would improve discrimination signatures. Our study showed remained walking speeds: Notably, achieved exceptional accuracy (99.8%, confidence interval (CI): 99.1-100%) classifying individuals, surpassing both 2D kinetics. Moreover, participants exhibited consistent, predictable linear their entire range. These were associated with participants’ preferred balance ability, cadence, step length. findings support as tool characterize individual differences predict speed-induced dynamics.

Language: Английский

---

A Step Forward Understanding Directional Limitations in Markerless Smartphone-Based Gait Analysis: A Pilot Study

Sensors, Journal Year: 2024, Volume and Issue: 24(10), P. 3091 - 3091

Published: May 13, 2024

The progress in markerless technologies is providing clinicians with tools to shorten the time of assessment rapidly, but raises questions about potential trade-off accuracy compared traditional marker-based systems. This study evaluated OpenCap system against a system-Vicon. Our focus was on its performance capturing walking both toward and away from two iPhone cameras same setting, which allowed Timed Up Go (TUG) test. that standard by comparing spatial-temporal kinematic parameters 10 participants. focused identifying discrepancies results using correlation analysis. Case examples further explored our results. demonstrated good faced challenges accurately parameters, especially direction facing cameras. Notably, directions observed significant differences pelvic obliquity, hip abduction, ankle flexion. findings suggest areas for improvement technologies, highlighting their clinical settings.

Language: Английский

---