Wearable wireless ultrasonic sensors for human gait analysis
Human motion capture has been used in rehabilitation clinics for evaluating, quantifying the severity of various neuromuscular and musculoskeletal diseases, and monitoring the progress of recovery. High accuracy motion tracking is necessary for gait analysis. The analyses of bilateral gait symmetry,...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis-Doctor of Philosophy |
| Language: | English |
| Published: |
Nanyang Technological University
2020
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/139650 |
| _version_ | 1811692020555579392 |
|---|---|
| author | Ashhar, Karalikkadan |
| author2 | Soh Cheong Boon |
| author_facet | Soh Cheong Boon Ashhar, Karalikkadan |
| author_sort | Ashhar, Karalikkadan |
| collection | NTU |
| description | Human motion capture has been used in rehabilitation clinics for evaluating, quantifying the severity of various neuromuscular and musculoskeletal diseases, and monitoring the progress of recovery. High accuracy motion tracking is necessary for gait analysis. The analyses of bilateral gait symmetry, coordination, and lower extremity joint angles during gait are important in rehabilitation after trauma, prosthesis, and other neurovascular diseases such as stroke. It can also act as a precursor to the freezing of gait in Parkinson’s disease. Current methods for gait analysis include optical motion tracking systems using multiple high-speed cameras, instrumented platforms, etc. which are complicated, costly and require controlled lighting conditions and dedicated laboratories. Systems using multiple cameras are sensitive to variations in the intensity of light, occlusion, and shadow. Recent advancements in the field of micro-electromechanical systems led to the use of inertial measurement units like accelerometers and gyroscopes for human motion tracking. However, the unavoidable integration drift in the calculated position values made the independent use of these systems unfit for human motion capture. The combination of various kinds of sensors to improve reliability makes the system more complex and costly. Narrowband ultrasonic sensors can be used for accurate ranging with considerably cheaper hardware. This research proposes a scalable, cost-effective, wearable ultrasonic gait analysis system, which is easy to use, portable, and does not require complex calibration procedures. The system uses four fixed anchors to track multiple mobile nodes. The Time of Arrival of ultrasonic pulses has been used for localization with the help of spherical positioning and Unscented Kalman filter. Initially, we compare the performance of a wireless ultrawideband (UWB) system and a wireless ultrasonic system for gait analysis. Ultrasonic sensors have been found to provide better performance compared to UWB for low-cost wireless implementations. Broadband ultrasonic sensors perform better with correlation receivers; however, they are costly and requires higher operating voltages and cannot be used for wearable applications, and this motivates us to design a scalable gait analysis system using narrowband ultrasonic sensors. The standard method for time of flight estimation is the correlation where the transmitted and the received signals are correlated to provide a peak delayed by the time of flight of the wave. In correlation receivers, the accuracy of ranging is inversely proportional to the width of the correlation peak and the width of the correlation peak is again inversely related to the bandwidth of the signal. An increase in the width of correlation peak due to the lower bandwidth of the signals prevents the identification and separation of close multi-path components at the receiver side. In the initial part of this thesis, a gait analysis system has been designed with one wearable mobile node where a novel method to compensate for the errors caused by close multi-path components has been proposed. Moving sensors emitting ultrasonic signals also cause Doppler shift in the received signal which affects the accuracy of ranging. We propose two methods for Doppler shift correction using chirp signals. Narrowband ultrasonic chirp signals have been used for ranging and localization of one mobile node along with a novel multi-path compensation and Doppler correction technique during gait analysis of healthy human subjects. Four passive anchor nodes and one active mobile node have been used. The mean values of the Root Mean Square Error (RMSE) with reference to the motion capture system obtained for tracking one mobile node using chirp signals were found to be 19.32, 9.79, and 18.40 mm which was again reduced to 16.04, 8.14, and 12.92 mm respectively for x, y, and z-axes after removing close multi-path components in the received signals using the proposed multi-path compensation method. The proposed Doppler correction methods were also compared to traditional methods using a pendulum experiment. To increase the number of mobile nodes, we test spatial division multiple access in the initial experiments in the second part of the thesis where sensors are arranged spatially in a way that multiple mobile nodes can transmit at the same time without interference. For tracking both the lower limbs simultaneously during gait analysis employing two mobile nodes using monotonous ultrasonic pulses and spatial division multiple access, net-RMSE was found to be 27.67, 32.01, and 35.63 mm for walking speed of 1, 2, and 3 kph. A pilot simulation study has been conducted to find the performance of tracking the hip and knee joint angles with a minimum number of mobile nodes attached to the body. To reduce the number of anchor nodes and make the system robust, we have designed an ultrasonic system in which orthogonal coded chirp signals are used for simultaneous multiple access and Doppler-correction. The ranging performance with the method is validated using a pendulum experiment with four ultrasonic transmitters. The method has been tested on healthy subjects for bilateral gait symmetry analysis and base of gait estimation using only two mobile tags and four anchor nodes. The mobile nodes were attached to the feet of the subject, and the anchor node plane was placed behind the subject walking on a treadmill. The results were compared with the Motion capture system and mean net-RMSE between the trajectories was found to be 30.30, 28.55, and 32.01 mm respectively for walking speed of 1, 2, and 3 kph. The spatial and temporal parameters extracted from the proposed method have been compared to the reference system. Experiments were conducted to evaluate the performance of the gait analysis system with proposed methods for signal design, Doppler correction, multiple access, and multi-path compensation compared to the optical motion capture reference system. The experimental results show that the proposed system can be used to assess the human gait symmetry and other spatial and temporal parameters in a convenient setup for a long duration with minimal errors. Moreover, a bilateral gait deviation measure has been proposed to quantify the deviations between the trajectories of the left and the right lower limbs. The deviation measures from both systems were determined for five healthy subjects and one subject with a simulated leg length discrepancy, and these values were compared.
In summary, this research proposes, evaluates and validates a convenient and scalable ultrasonic gait analysis system for rehabilitation clinics. |
| first_indexed | 2024-10-01T06:29:09Z |
| format | Thesis-Doctor of Philosophy |
| id | ntu-10356/139650 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T06:29:09Z |
| publishDate | 2020 |
| publisher | Nanyang Technological University |
| record_format | dspace |
| spelling | ntu-10356/1396502023-07-04T17:43:42Z Wearable wireless ultrasonic sensors for human gait analysis Ashhar, Karalikkadan Soh Cheong Boon School of Electrical and Electronic Engineering ECBSOH@ntu.edu.sg Engineering::Electrical and electronic engineering::Electronic systems::Signal processing Human motion capture has been used in rehabilitation clinics for evaluating, quantifying the severity of various neuromuscular and musculoskeletal diseases, and monitoring the progress of recovery. High accuracy motion tracking is necessary for gait analysis. The analyses of bilateral gait symmetry, coordination, and lower extremity joint angles during gait are important in rehabilitation after trauma, prosthesis, and other neurovascular diseases such as stroke. It can also act as a precursor to the freezing of gait in Parkinson’s disease. Current methods for gait analysis include optical motion tracking systems using multiple high-speed cameras, instrumented platforms, etc. which are complicated, costly and require controlled lighting conditions and dedicated laboratories. Systems using multiple cameras are sensitive to variations in the intensity of light, occlusion, and shadow. Recent advancements in the field of micro-electromechanical systems led to the use of inertial measurement units like accelerometers and gyroscopes for human motion tracking. However, the unavoidable integration drift in the calculated position values made the independent use of these systems unfit for human motion capture. The combination of various kinds of sensors to improve reliability makes the system more complex and costly. Narrowband ultrasonic sensors can be used for accurate ranging with considerably cheaper hardware. This research proposes a scalable, cost-effective, wearable ultrasonic gait analysis system, which is easy to use, portable, and does not require complex calibration procedures. The system uses four fixed anchors to track multiple mobile nodes. The Time of Arrival of ultrasonic pulses has been used for localization with the help of spherical positioning and Unscented Kalman filter. Initially, we compare the performance of a wireless ultrawideband (UWB) system and a wireless ultrasonic system for gait analysis. Ultrasonic sensors have been found to provide better performance compared to UWB for low-cost wireless implementations. Broadband ultrasonic sensors perform better with correlation receivers; however, they are costly and requires higher operating voltages and cannot be used for wearable applications, and this motivates us to design a scalable gait analysis system using narrowband ultrasonic sensors. The standard method for time of flight estimation is the correlation where the transmitted and the received signals are correlated to provide a peak delayed by the time of flight of the wave. In correlation receivers, the accuracy of ranging is inversely proportional to the width of the correlation peak and the width of the correlation peak is again inversely related to the bandwidth of the signal. An increase in the width of correlation peak due to the lower bandwidth of the signals prevents the identification and separation of close multi-path components at the receiver side. In the initial part of this thesis, a gait analysis system has been designed with one wearable mobile node where a novel method to compensate for the errors caused by close multi-path components has been proposed. Moving sensors emitting ultrasonic signals also cause Doppler shift in the received signal which affects the accuracy of ranging. We propose two methods for Doppler shift correction using chirp signals. Narrowband ultrasonic chirp signals have been used for ranging and localization of one mobile node along with a novel multi-path compensation and Doppler correction technique during gait analysis of healthy human subjects. Four passive anchor nodes and one active mobile node have been used. The mean values of the Root Mean Square Error (RMSE) with reference to the motion capture system obtained for tracking one mobile node using chirp signals were found to be 19.32, 9.79, and 18.40 mm which was again reduced to 16.04, 8.14, and 12.92 mm respectively for x, y, and z-axes after removing close multi-path components in the received signals using the proposed multi-path compensation method. The proposed Doppler correction methods were also compared to traditional methods using a pendulum experiment. To increase the number of mobile nodes, we test spatial division multiple access in the initial experiments in the second part of the thesis where sensors are arranged spatially in a way that multiple mobile nodes can transmit at the same time without interference. For tracking both the lower limbs simultaneously during gait analysis employing two mobile nodes using monotonous ultrasonic pulses and spatial division multiple access, net-RMSE was found to be 27.67, 32.01, and 35.63 mm for walking speed of 1, 2, and 3 kph. A pilot simulation study has been conducted to find the performance of tracking the hip and knee joint angles with a minimum number of mobile nodes attached to the body. To reduce the number of anchor nodes and make the system robust, we have designed an ultrasonic system in which orthogonal coded chirp signals are used for simultaneous multiple access and Doppler-correction. The ranging performance with the method is validated using a pendulum experiment with four ultrasonic transmitters. The method has been tested on healthy subjects for bilateral gait symmetry analysis and base of gait estimation using only two mobile tags and four anchor nodes. The mobile nodes were attached to the feet of the subject, and the anchor node plane was placed behind the subject walking on a treadmill. The results were compared with the Motion capture system and mean net-RMSE between the trajectories was found to be 30.30, 28.55, and 32.01 mm respectively for walking speed of 1, 2, and 3 kph. The spatial and temporal parameters extracted from the proposed method have been compared to the reference system. Experiments were conducted to evaluate the performance of the gait analysis system with proposed methods for signal design, Doppler correction, multiple access, and multi-path compensation compared to the optical motion capture reference system. The experimental results show that the proposed system can be used to assess the human gait symmetry and other spatial and temporal parameters in a convenient setup for a long duration with minimal errors. Moreover, a bilateral gait deviation measure has been proposed to quantify the deviations between the trajectories of the left and the right lower limbs. The deviation measures from both systems were determined for five healthy subjects and one subject with a simulated leg length discrepancy, and these values were compared. In summary, this research proposes, evaluates and validates a convenient and scalable ultrasonic gait analysis system for rehabilitation clinics. Doctor of Philosophy 2020-05-20T12:26:57Z 2020-05-20T12:26:57Z 2020 Thesis-Doctor of Philosophy Ashhar, K. (2020). Wearable wireless ultrasonic sensors for human gait analysis. Doctoral thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/139650 10.32657/10356/139650 en This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0). application/pdf Nanyang Technological University |
| spellingShingle | Engineering::Electrical and electronic engineering::Electronic systems::Signal processing Ashhar, Karalikkadan Wearable wireless ultrasonic sensors for human gait analysis |
| title | Wearable wireless ultrasonic sensors for human gait analysis |
| title_full | Wearable wireless ultrasonic sensors for human gait analysis |
| title_fullStr | Wearable wireless ultrasonic sensors for human gait analysis |
| title_full_unstemmed | Wearable wireless ultrasonic sensors for human gait analysis |
| title_short | Wearable wireless ultrasonic sensors for human gait analysis |
| title_sort | wearable wireless ultrasonic sensors for human gait analysis |
| topic | Engineering::Electrical and electronic engineering::Electronic systems::Signal processing |
| url | https://hdl.handle.net/10356/139650 |
| work_keys_str_mv | AT ashharkaralikkadan wearablewirelessultrasonicsensorsforhumangaitanalysis |