GaitTracker: 3D Skeletal Tracking for Gait Analysis Based on Inertial Measurement Units

Gait analysis is essential for understanding human locomotion patterns and diagnosing gait disorders. While traditional methods typically rely on complex equipment and laboratory settings, recent advances in technology have enabled gait analysis outside of controlled laboratory settings. This study explores the impact of sensor placement on the accuracy of gait parameter estimation using inertial measurement units (IMUs). We conducted experiments involving 31 healthy participants walking under controlled conditions while wearing multiple IMU sensors at different locations on their bodies, including three locations to simulate commercially available wearable devices - earphones, watches, and smartphones. In addition, participants wore a robotic hip exoskeleton to test an IMU sensor and joint angle sensors in it. The Combination of sensors in the exoskeleton was found to be the best when averaged for four gait parameters: speed, stride length, step length, and gait phase; and for each gait parameter, the best sensor location was all different. These findings have important implications for the design of wearable devices for gait analysis and rehabilitation applications, as well as for clinical practice and sports medicine.

To identify the gait parameters used to assess gait disorders in children with bilateral spastic cerebral palsy (CP) and evaluate their responsiveness to treatments. A systematic search within PubMed, Web of Science, and Scopus (in English, 2000-2016) for randomized controlled trials of children with bilateral spastic CP who were assessed by instrumented gait analysis (IGA) was performed. Data related to participants and study characteristics, risk of bias, and outcome measures were collected. A list of gait parameters responsive to clinical interventions was obtained. Twenty-one articles met the inclusion criteria. Eighty-nine gait parameters were identified, 56 of which showed responsiveness to treatments. Spatiotemporal and kinematic parameters were widely used compared to kinetic and surface electromyography data. The majority of responsive gait parameters were joint angles at the sagittal plane (flexion-extension). The IGA yields responsive outcome measures for the gait assessment of children with bilateral spastic CP. Spatiotemporal and kinematic (at sagittal plane) parameters are the gait parameters used most frequently. Further research is needed to establish the relevant gait parameters for each clinical problem. Fifty-six responsive gait parameters for children with bilateral spastic cerebral palsy were identified. Most responsive gait parameters belong to joint angles time-series at sagittal plane. Spatiotemporal and kinematic parameters are widely used compared to kinetic and surface electromyography parameters.

Gait rehabilitation is a common method of postoperative recovery after the user sustains an injury or disability. However, traditional gait rehabilitations are usually performed under the supervision of rehabilitation specialists, meaning the patients can not receive adequate care continuously. In this paper, we propose IMU-Kinect, a novel system to remotely and continuously monitor the gait rehabilitation via the wearable kit. This system consists of a wearable hardware platform and a user-friendly software application. The hardware platform is composed of four Inertial Measurement Units (IMU), which are attached on the shanks and thighs of the human body. The software application is able to estimate the rotation and displacement of these sensors, then reconstruct the gait movements and calculate the gait parameters according to the geometric model of human lower limbs. Based on IMU-Kinect system, the users of gait rehabilitation just need to walk normally by wearing the IMU-Kinect kit, and then the rehabilitation specialists can analyze the status of postoperative recovery by remotely viewing the animations about users' gait movements and charts of the general gait parameters. Extend experiments in real environment show that our system can efficiently track the gait movements with 9% rotation and displacement error.

Changes in gait are associated with an increased risk of falling and may indicate the presence of movement disorders related to neurological diseases or age-related weakness. Continuous monitoring based on inertial measurement unit (IMU) sensor data can effectively estimate gait parameters that reflect changes in gait dynamics. Monitoring using a waist-level IMU sensor is particularly useful for assessing such data, as it can be conveniently worn as a sensor-integrated belt or observed through a smartphone application. Our work investigates the efficacy of estimating gait events and gait parameters based on data collected from a waist-worn IMU sensor. The results are compared to measurements obtained using a GAITRite® system as reference. We evaluate two machine learning (ML)-based methods. Both ML methods are structured as sequence to sequence (Seq2Seq). The efficacy of both approaches in accurately determining gait events and parameters is assessed using a dataset comprising 17,643 recorded steps from 69 subjects, who performed a total of 3588 walks, each covering approximately 4 m. Results indicate that the Convolutional Neural Network (CNN)-based algorithm outperforms the long short-term memory (LSTM) method, achieving a detection accuracy of 98.94% for heel strikes (HS) and 98.65% for toe-offs (TO), with a mean error (ME) of 0.09 ± 4.69 cm in estimating step lengths.

Gait rehabilitation commonly relies on bodyweight unloading mechanisms, such as overhead mechanical support and underwater buoyancy. Lightweight and wireless inertial measurement unit (IMU) sensors provide a cost-effective tool for quantifying body segment motions without the need for video recordings or ground reaction force measures. Identifying the instant when the foot contacts and leaves the ground from IMU data can be challenging, often requiring scrupulous parameter selection and researcher supervision. We aimed to assess the use of machine learning methods for gait event detection based on features from foot segment rotational velocity using foot-worn IMU sensors during bodyweight-supported treadmill walking on land and underwater. Twelve healthy subjects completed on-land treadmill walking with overhead mechanical bodyweight support, and three subjects completed underwater treadmill walking. We placed IMU sensors on the foot and recorded motion capture and ground reaction force data on land and recorded IMU sensor data from wireless foot pressure insoles underwater. To detect gait events based on IMU data features, we used random forest machine learning classification. We achieved high gait event detection accuracy (95–96%) during on-land bodyweight-supported treadmill walking across a range of gait speeds and bodyweight support levels. Due to biomechanical changes during underwater treadmill walking compared to on land, accurate underwater gait event detection required specific underwater training data. Using single-axis IMU data and machine learning classification, we were able to effectively identify gait events during bodyweight-supported treadmill walking on land and underwater. Robust and automated gait event detection methods can enable advances in gait rehabilitation.

For healthcare and clinical use, ambulatory gait monitoring systems using inertial sensors have been developed to estimate the user gait parameters, such as walking speed, stride time, and stride length. However, to adapt the systems effectively to daily-life activities, they need to be able to classify the gait activities of daily-life to obtain the parameters for each activity. In this study, we propose a simple classification algorithm based on a single inertial sensor for ease of use, which classifies three major gait activities: leveled walk, ramp walk, and stair walk. The classification can be performed with gait parameter estimation simultaneously. The developed system that includes classification and parameter estimation algorithms was evaluated with eight healthy subjects within a gait lab and on an outdoor daily-life walking course. The results showed that the estimated gait parameters were comparable to existing studies (range of walking speed root mean square error: 0.059-0.129 m/s), and the classification accuracy was sufficiently high for all three gait activities: 98.5% for the indoor gait lab experiment and 95.5% for the outdoor complex daily-life walking course experiment. The proposed system is simple and effective for daily-life gait analysis, including gait activity classification and gait parameter estimation for each activity.

The use of foot mounted inertial and other auxiliary sensors for kinematic gait analysis has been extensively investigated during the last years. Although, these sensors still yield less accurate results than those obtained employing optical motion capture systems, the miniaturization and their low cost have allowed the estimation of kinematic spatiotemporal parameters in laboratory conditions and real life scenarios. The aim of this work was to present a comprehensive approach of this scientific area through a systematic literature research, breaking down the state-of-the-art methods into three main parts: (1) zero velocity interval detection techniques; (2) assumptions and sensors' utilization; (3) foot pose and trajectory estimation methods. Published articles from 1995 until December of 2018 were searched in the PubMed, IEEE Xplore and Google Scholar databases. The research was focused on two categories: (a) zero velocity interval detection methods; and (b) foot pose and trajectory estimation methods. The employed assumptions and the potential use of the sensors have been identified from the retrieved articles. Technical characteristics, categorized methodologies, application conditions, advantages and disadvantages have been provided, while, for the first time, assumptions and sensors' utilization have been identified, categorized and are presented in this review. Considerable progress has been achieved in gait parameters estimation on constrained laboratory environments taking into account assumptions such as a person walking on a flat floor. On the contrary, methods that rely on less constraining assumptions, and are thus applicable in daily life, led to less accurate results. Rule based methods have been mainly used for the detection of the zero velocity intervals, while more complex techniques have been proposed, which may lead to more accurate gait parameters. The review process has shown that presently the best-performing methods for gait parameter estimation make use of inertial sensors combined with auxiliary sensors such as ultrasonic sensors, proximity sensors and cameras. However, the experimental evaluation protocol was much more thorough, when single inertial sensors were used. Finally, it has been highlighted that the accuracy of setups using auxiliary sensors may further be improved by collecting measurements during the whole foot movement and not only partially as is currently the practice. This review has identified the need for research and development of methods and setups that allow for the robust estimation of kinematic gait parameters in unconstrained environments and under various gait profiles.

Background Gait and balance disorders in patients with idiopathic Parkinson's disease (PD) lead to major mobility limitations. To counteract this, physical therapy such as gait, balance, or resistance training is applied. Integrative training methods, which combine these elements, could be particularly effective. Objective The objective of this study is to evaluate and compare the effects of two integrative interventions on gait and balance of patients with PD. Methods Twenty-six patients with PD received either resistance training in combination with gait training (gait resistance training, GRT) or resistance training in combination with balance training (stability resistance training, SRT) for six weeks. Gait and balance outcome parameters were assessed before, immediately after, and six weeks after the interventions. The primary outcome parameters were the functional reach test to evaluate balance and stride length to evaluate gait. Secondary outcomes included further gait analysis parameters, knee extension strength, the timed up and go test, and the six-minute walk test. Results The functional reach test results were significantly better after the intervention in both groups. Stride length increased significantly only in the GRT group. Several further gait parameters and the six-minute walk test improved in the GRT group, and the increase in gait speed was significantly higher than in the SRT group. The SRT group performed better after the intervention regarding the timed up and go test and knee extension strength, the latter being significantly more improved than in the SRT group. At six-week follow-up, the improvement in functional reach was maintained in the SRT group. Conclusions Integrative therapies, combining gait or balance training with resistance training, have specific positive effects in PD rehabilitation. More pronounced effects on gait parameters are achieved by GRT, while SRT has more impact on balance. Thus, the combination of both training methods might be particularly efficient in improving the mobility of PD patients.

Introduction: The Inertial Measurement Unit (IMU) is electronic device that enabled us to record joint angles, gait event and spatiotemporal parameter using accelerometers. IMU contain sensors known as inertial sensor which measures its movement by using the inertia principle. This study aimed to determine validity and reliability of spatiotemporal gait parameter using the IMU sensor. Methods: This study are prospective cross-sectional recruited thirteen convenience subjects (four men, nine women; 50.0 +/-15.0 years) diagnosed with chronic Non-Specific Lower Back Pain (LBP) from Physiotherapy Department, upon an Orthopedic Specialist’s referral, at Hospital Sultanah Aminah Johor Bahru. Spatiotemporal parameters interested: left and right velocity, cadence, stride/step time and stride/step length recorded by Vicon system and IMU sensors synchronously. Results: Higher validity was indicated at the Trial 2 detected by the IMU sensors comparing Vicon system, with significant correlation p ≤ 0.05 except stride time left shank (r = 0.539, p = 0.06) , left foot (r = 0.495, p = 0.11) and step length left shank (r = 0.532, p = 0.06). The result of study also indicated that the reliability of the IMU sensors based on ICCs ≥ 0.75 and 95% CI 0.180 – 0.993, p ≤ 0.01 in Non-specific LBP patients for spatiotemporal gait parameters comparing Trial 1 and Trial 2. Conclusion: The IMU system performs to be valid and reliable for determine spatiotemporal gait parameters in Non-specific LBP patients. IMU provides a possible solution to measure spatiotemporal gait in a clinical setting without requiring specific working area and professional technician.

Background:Spinal mobility is an important assessment outcome in axial spondyloarthritis (axSpA). Until now, conventional metrology (Schober test, lateral flexion, BASMI, …) has been used to assess spinal mobility, however, new technologies have been developed that provide better accuracy, reliability and responsiveness. Motion capture has been validated and Inertial Measurement Unit (IMU) sensors, appears to be a promising alternative. To use this IMU sensors in axSpA patients, wireless systems must be developed and validated allowing to doctors and patients to use them in hospitals and at home.Objectives:To develop an easy to use mobile app and IMU sensors system for analyse mobility for axSpA patients.Methods:A mobile app has been developed (iUCOTrack) that communicates with two IMU sensors (Shimmer 3©, Fig-a). These sensors are attached in different locations: at forehead and T12 for cervical mobility (Fig-c) and T12 and Sacrum for thoracolumbar mobility (Fig-b). The app provides mobility results for the different tests (Fig-d) and store results in the cloud. Validation tests of these sensors, using Matlab©, were done previously [1]. Our study test the validity of this app against a motion capture system, the UCOTrack®, and its metrology index, the UCOASMI [2], and conventional metrology as reference standards. Patiens with axSpA were recruited consecutively from the COSPAR cohort. Conventional metrology, PRO questionnaires and mobility (Cervical and thoracolumbar - flexion, lateral bending, rotation) using the iUCOTrack app and the UCOTrack were registered. Intraclasss Correlation Coefficients (ICC 3,1) between systems and correlations (spearman) with other axSpA outcome measures were performed for testing validity.Results:15 axSpA patients (47% female, age 52±12 years, disease duration 21±16 years) were included. Table shows ROM (SD) in degrees obtained for cervical and thoracolumbar spine measured by motion capture (UCOTrack) and the app (iUCOTrack). In the last column appears the UCOASMI (SD) calculated using angles obtained by each system. All ICC were good (ICC>0.8), and correlations were significant (p<0.05, r>0.8) specially the UCOASMI. Cervical rotation using a goniometer was 106.2±36°, with a significant correlation with both systems (p<0.05; r>0.8). Schober correlation with lumbar flexion was poor (NS;r>0.5) but a good correlation appeared with lateral flexion (p<0.01;r>0.9). Mean BASMI was 4.0± 1.8 with an excellent correlation with UCOASMI measured by Mocap (p<0.01;r=0.93) and by IMU (p<0.001;r=0.98).CervicalThoracolumbarFlexRotLatFlexRotLatUCOASMIUCOTrack79.5(24.7)109.8(29.6)62.5(25.1)100.7(21.6)61.8(25.3)54.7(22.9)6.07(1.66)iUCOTrack83.0(33.6)112.6(44.3)73.9(29.7)114.4(28.1)51.4(16.1)59.4(15.4)6.15(1.65)ICC0.8640.9030.8120.9360.7980.9010.970Corr0.89*0.96**0.82*0.97**0.88*0.97***0.97**Conclusion:New metrology tools are needed to improve features of convencional metrology. Motion Capture has proved to be valid but has feasibility problems. IMU sensor based systems provide similar results to motion capture but it can be faster and cheaper. A system based on mobile app connected to wireless IMU sensors could be a solution to improve metrology in axSpA. Further studies and developments are needed to introduce these technologies in research and clinical daily practice.

A markerless gait analysis system can measure useful gait metrics to determine effective clinical treatment. Although this gait analysis system does not require a large space, several markers, or time constraints, it inaccurately measure lower limb joint kinematics during gait. In particular, it has a substantial ankle joint angle error. In this study, we investigated the markerless gait analysis method capability using single RGB camera-based pose estimation by OpenPose (OP) and an inertial measurement unit (IMU) sensor on the foot segment to measure ankle joint kinematics under various gait conditions. Sixteen healthy young adult males participated in the study. We compared temporo-spatial parameters and lower limb joint angles during four gait conditions with varying gait speeds and foot progression angles. These were measured by optoelectronic motion capture, markerless gait analysis method using OP, and proposed method using OP and IMU. We found that the proposed method using OP and an IMU significantly decreased the mean absolute errors of peak ankle joint angles compared with OP in the four gait conditions. The proposed method has the potential to measure temporo-spatial gait parameters and lower limb joint angles, including ankle angles, in various gait conditions as a clinical settings gait assessment tool.

This study aimed to validate the application of an inertial measurement unit (IMU) sensor for calculating the spatiotemporal parameters of gait in healthy young and older adults. Six healthy young adults and four older adults participated in this study; each wore an IMU sensor on the fifth lumbar vertebra (L5) while the sensor was synchronized with a three-dimensional motion capture system. Nineteen gait parameters, including duration, variability, and asymmetry, were calculated from initial contact and final contact event times. Results showed that the gait parameters calculated from the wearable device system were reliable and had a small absolute mean difference from the parameters calculated using a motion capture system, and the highest absolute mean difference was only 0.01 s in the stance and swing phases among the young subjects.

Inertial measurement unit (IMU)-based gait analysis systems have become popular in clinical environments because of their low cost and quantitative measurement capability. When a shank is selected as the IMU mounting position, an inverted pendulum model (IPM) can accurately estimate its spatial gait parameters. However, the stride-by-stride estimation of gait parameters using one IMU on each shank and the IPMs has not been validated. This study validated a spatial gait parameter estimation method using a shank-based IMU system. Spatial parameters were estimated via the double integration of the linear acceleration transformed by the IMU orientation information. To reduce the integral drift error, an IPM, applied with a linear error model, was introduced at the mid-stance to estimate the update velocity. the gait data of 16 healthy participants that walked normally and slowly were used. The results were validated by comparison with those extracted from an optical motion-capture system; the results showed strong correlation (r>0.9) and good agreement with the gait metrics (stride length, stride velocity, and shank vertical displacement). In addition, the biases of the stride length and stride velocity extracted using the motion capture system were smaller in the IPM than those in the previous method using the zero-velocity-update. The error variabilities of the gait metrics were smaller in the IPM than those in the previous method. These results indicated that the reconstructed shank trajectory achieved a greater accuracy and precision than that of previous methods. This was attributed to the IPM, which demonstrates that shank-based IMU systems with IPMs can accurately reflect many spatial gait parameters including stride velocity.

Inertial Measurement Unit (IMU) sensors are used in many applications that include aviation, vehicle systems, unmanned aircraft, indoor navigation, health, and robotic systems. An IMU consists of accelerometers and gyroscope sensors combined in a single module. However, the accelerometer or gyroscope alone cannot produce reliable data, and so the outputs are combined to determine accurate data for measurements such as direction, velocity, angular velocity and position. The data collected from IMU sensors may differ due to measurement errors, calibration issues, and errors due to ambient noise. Small errors in IMU sensors can cause large deviations in applications. There is no clear distinction between the performance and area of use of commercially available sensors. Therefore, when selecting a sensor, the requirements for performance should be determined for the area of use and choosen accordingly. This study investigates the performance of three IMU sensors that have no specific application area and are in common use. An experimental setup was designed and implemented to test the accuracy of the acceleration and gyroscopic information obtained from the IMU sensors. The test apparatus consists of IMU sensor, encoder, stepper motor and Raspberry Pi. The stepper motor and encoder are connected to a shaft, and the IMU sensor is mounted on a rotating moving mechanism. The apparatus is controlled by a Raspberry Pi. The Python programming language has been used for the control software. The apparatus provides rotation of a desired angle and velocity. Acceleration and gyroscopic data received from the IMU sensor are drawn in real time. All sensors were first calibrated and then data were taken. The performance of the sensors was compared using the angular values around the x-axis. The test setup was rotated at a certain angle in the x-axis using a stepper motor. The gyroscopic data on the x-axis for each IMU sensor were then read and processed through a Kalman filter. The accuracy of the IMU sensors was determined with reference to the encoder data.

Gait analysis is an important medical diagnostic process which can be used in many applications like mobility assessment, healthcare, sports training, therapy, rehabilitation and many more. This analysis is usually carried out in gait laboratories, which maj ority of population is not aware especially in developing countries like India. Hence the development of shoe-based system that can track the walking pattern of a person by obtaining gait parameters is necessary. Depending on these gait parameters it can be concluded whether a person has a normal walking pattern or there are any abnormalities while walking. The system comprises of three Force Sensing Resistors (FSR), one Inertial Measurement Unit (IMU) sensor and one Ultrasound sensor on each leg which is connected to Arduino controller. FSRs are placed at pressure points located under foot. The three FSRs are placed at heel point, fore-foot point and thumb point under each foot. IMU and ultrasound sensor is placed above the shoe. These sensor values are sent using Bluetooth to the system wirelessly for further analysis. The parameters Stance (ST), Swing (SW), Heel-Strike (HS), Heel-Off (HO) and Stride Length (SL) are obtained based on the sensor values. Analysis of these parameters is performed to any abnormalities in walking pattern.