sEMG-Based Motion Recognition of Upper Limb Rehabilitation Using the Improved Yolo-v4 Algorithm
Abstract
:1. Introduction
2. Methods
2.1. System Overview
2.2. sEMG Acquisition
2.3. sEMG Signal Preprocessing
2.4. sEMG Channels Reduction
2.5. Feature Extraction Network Model
2.6. The Prediction Network
2.7. The Evaluation Criteria
3. Results
3.1. The Performance of Feature Extraction Network
3.2. The Performance of Prediction Network
4. Conclusions and Discussion
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
References
- Xu, H.; Xiong, A. Advances and Disturbances in sEMG-Based Intentions and Movements Recognition: A Review. IEEE Sens. J. 2021, 21, 13019–13028. [Google Scholar] [CrossRef]
- Yang, X.; Yan, J.; Fang, Y.; Zhou, D.; Liu, H. Simultaneous Prediction of Wrist/Hand Motion via Wearable Ultrasound Sensing. IEEE Trans. Neural Syst. Rehabil. Eng. 2020, 28, 970–977. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Chen, T.; Casas, R.; Lum, P.S. An Elbow Exoskeleton for Upper Limb Rehabilitation with Series Elastic Actuator and Cable-Driven Differential. IEEE Trans. Robot. 2019, 35, 1464–1474. [Google Scholar] [CrossRef] [PubMed]
- Nicholson-Smith, C.; Mehrabi, V.; Atashzar, S.F.; Patel, R.V. Multi-Functional Lower- and Upper-Limb Stroke Rehabilitation Robot. IEEE Trans. Med. Robot. Bionics 2020, 2, 549–552. [Google Scholar] [CrossRef]
- Xiong, D.; Zhang, D.; Zhao, X.; Zhao, Y. Deep Learning for EMG-based Human-Machine Interaction A Review. IEEE/CAA J. Autom. Sin. 2021, 8, 512–533. [Google Scholar] [CrossRef]
- Song, Z.; Guo, S.; Xiao, N.; Gao, B.; Shi, L. Implementation of human-machine synchronization control for active rehabilitation using an inertia sensor. Sensors 2012, 12, 16046–16059. [Google Scholar] [CrossRef]
- Song, Z.; Guo, S.; Pang, M.; Zhang, S.; Xiao, N.; Gao, B.; Shi, L. Implementation of resistance training using an upper-limb exoskeleton rehabilitation device for elbow joint. J. Med. Biol. Eng. 2014, 34, 188–196. [Google Scholar] [CrossRef]
- Bi, L.; Feleke, A.; Guan, C. A review on EMG-based motor intention prediction of continuous human upper limb motion for human-robot collaboration. Biomed. Signal Process. Control. 2019, 51, 113–127. [Google Scholar] [CrossRef]
- Zhang, S.; Guo, S.; Gao, B.; Huang, Q.; Pang, M.; Hirata, H.; Ishihara, H. Muscle strength assessment system using sEMG-based force prediction method for wrist joint. J. Med. Biol. Eng. 2016, 36, 121–131. [Google Scholar] [CrossRef]
- Yang, Z.; Guo, S.; Liu, Y.; Hirata, H.; Tamiya, T. An intention-based online bilateral training system for upper limb motor rehabilitation. Microsyst. Technol. 2021, 27, 211–222. [Google Scholar] [CrossRef]
- Toledo-Perez, D.C.; Rodriguez-Resendiz, J.; Gomez-Loenzo, R.A. A Study of Computing Zero Crossing Methods and an Improved Proposal for EMG Signals. IEEE Access 2020, 8, 8483–8490. [Google Scholar] [CrossRef]
- Toledo-Pérez, D.C.; Martínez-Prado, M.A.; Gómez-Loenzo, R.A.; Paredes-García, W.J.; Rodríguez-Reséndiz, J. A Study of Movement Classification of the Lower Limb Based on up to 4-EMG Channels. Electronics 2019, 8, 259. [Google Scholar] [CrossRef] [Green Version]
- Toledo-Pérez, D.C.; Rodríguez-Reséndiz, J.; Gómez-Loenzo, R.A.; Jauregui-Correa, J.C. Support Vector Machine-Based EMG Signal Classification Techniques: A Review. Appl. Sci. 2019, 9, 4402. [Google Scholar] [CrossRef] [Green Version]
- Crocher, V.; Sahbani, A.; Robertson, J. Constraining upper limb synergies of hemiparetic patients using a robotic exoskeleton in the perspective of neuro-rehabilitation. IEEE Trans. Neural Syst. Rehabil. Eng. 2012, 20, 247–257. [Google Scholar] [CrossRef] [Green Version]
- Zhang, S.; Guo, S.; Fu, Y.; Boulardou, L.; Huang, Q.; Hirata, H.; Ishihara, H. Integrating Compliant Actuator and Torque Limiter Mechanism for Safe Home-Based Upper-Limb Rehabilitation Device Design. J. Med. Biol. Eng. 2017, 37, 357–364. [Google Scholar] [CrossRef]
- Trifonov, A.; Filist, S.; Degtyarev, S.; Serebrovsky, V.; Shatalova, O. Human–Machine Interface of Rehabilitation Exoskeletons with Redundant Electromyographic Channels. In Proceedings of the 15th International Conference on Electromechanics and Robotics “Zavalishin’s Readings”, Ufa, Russia, 15–18 April 2020. [Google Scholar] [CrossRef]
- Ding, B.; Tian, F.; Zhao, L. Digital Evaluation Algorithm for Upper Limb Motor Function Rehabilitation Based on Micro Sensor. J. Med. Imaging Health Inform. 2021, 11, 391–401. [Google Scholar] [CrossRef]
- Zhang, S.; Guo, S.; Pang, M.; Gao, B.; Guo, P. Mechanical Design and Control Method for SEA and VSA-based Exoskeleton Devices for Elbow Joint Rehabilitation. Neurosci. Biomed. Eng. 2015, 2, 142–147. [Google Scholar] [CrossRef] [Green Version]
- Maqsood, K.; Luo, J.; Yang, C.; Ren, Q.; Li, Y. Iterative learning-based path control for robot assisted upper-limb rehabilitation. Neural Comput. Appl. 2021. [Google Scholar] [CrossRef]
- Miao, Q.; Peng, Y.; Liu, L.; McDaid, A.; Zhang, M. Subject-specific compliance control of an upper-limb bilateral robotic system. Robot. Auton. Syst. 2020, 126, 103478. [Google Scholar] [CrossRef]
- Liu, Y.; Guo, S.; Yang, Z.; Hirata, H.; Tamiya, T. A Home-based Bilateral Rehabilitation System with sEMG-based Real-time Variable Stiffness. IEEE J. Biomed. Health Inform. 2020, 25, 1529–1541. [Google Scholar] [CrossRef]
- Yang, Z.; Guo, S.; Hirata, H.; Kawanishi, M. A Mirror Bilateral Neuro-Rehabilitation Robot System with the sEMG-Based Real-Time Patient Active Participant Assessment. Life 2021, 11, 1290. [Google Scholar] [CrossRef] [PubMed]
- Zhang, Y.; Chen, Y.; Yu, H.; Yang, X.; Lu, W. Learning Eective Spatial-Temporal Features for sEMG Armband. IEEE Internet Things J. 2020, 7, 6979–6992. [Google Scholar] [CrossRef]
- Zhou, Y.; Chen, C.; Cheng, M.; Alshahrani, Y.; Franovic, S.; Lau, E.; Xu, G.; Ni, G.; Cavanaugh, J.M.; Muh, S.; et al. Comparison of machine learning methods in sEMG signal processing for shoulder motion recognition. Biomed. Signal Process. Control. 2021, 68, 102577. [Google Scholar] [CrossRef]
- Hussain, S.; Jamwal, P.K.; Van Vliet, P.; Ghayesh, M.H. State-of-the-Art Robotic Devices for Wrist Rehabilitation: Design and Control Aspects. IEEE Trans. Hum.-Mach. Syst. 2020, 50, 361–372. [Google Scholar] [CrossRef]
- Huang, Y.; He, Z.; Liu, Y.; Yang, R.; Zhang, X.; Cheng, G.; Yi, J.; Ferreira Joao, P.; Liu, T. Real-time intended knee joint motion prediction by deep-recurrent neural networks. IEEE Sens. J. 2019, 19, 11503–11509. [Google Scholar] [CrossRef]
- Xiao, F.; Chen, Y.; Zhu, Y. GADF/GASF-HOG: Feature extraction methods for hand movement classification from surface electromyography. J. Neural Eng. 2020, 17, 046016. [Google Scholar] [CrossRef]
- Ding, Z.; Yang, C.; Wang, Z.; Yin, X.; Jiang, F. Online Adaptive Prediction of Human Motion Intention Based on sEMG. Sensors 2021, 21, 2882. [Google Scholar] [CrossRef]
- Desai, S.; Alhadad, R.; Mahmood, A.; Chilamkurti, N.; Rho, S. Multi-State Energy Classifier to Evaluate the Performance of the NILM Algorithm. Sensors 2019, 19, 5236. [Google Scholar] [CrossRef] [Green Version]
- Spanias, J.A.; Perreault, E.J.; Hargrove, L.J. Detection of and compensation for EMG disturbances for powered lower limb prosthesis control. IEEE Trans. Neural Syst. Rehabil. Eng. 2015, 24, 226–234. [Google Scholar] [CrossRef]
- Qi, S.; Wu, X.; Chen, W.H.; Liu, J.; Zhang, J.; Wang, J. sEMG-based recognition of composite motion with convolutional neural network. Sens. Actuators A Phys. 2020, 311, 112046. [Google Scholar] [CrossRef]
- Han, K.; Wang, Y.; Tian, Q.; Guo, J.; Xu, C.; Xu, C. GhostNet: More Features from Cheap Operations. In Proceedings of the in 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 13–19 June 2020. [Google Scholar] [CrossRef]
- Liu, S.; Qi, L.; Qin, H.; Shi, J.; Jia, J. Path Aggregation Network for Instance Segmentation. In Proceedings of the in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Salt Lake City, UT, USA, 18–23 June 2018. [Google Scholar] [CrossRef] [Green Version]
- Guo, C.; Lv, X.; Zhang, Y.; Zhang, M. Improved YOLOv4-tiny network for real-time electronic component detection. Sci. Rep. 2021, 11, 22744. [Google Scholar] [CrossRef]
Experimental Scheme | Biceps Brachii | Triceps Brachii | Brachialis | Experimental Result | |
---|---|---|---|---|---|
Accuracy (%) | Time Consuming (s) | ||||
Scheme 1 | ▲ | 61.9 | 103.29 | ||
Scheme 2 | ▲ | ▲ | 80.6 | 132.05 | |
Scheme 3 | ▲ | ▲ | ▲ | 82.3 | 193.27 |
Experimental Scheme | Extensor Carpi Ulnaris | Flexor Carpi Radialis | Extensor Carpi Radialis | Flexor Carpi Ulnaris | Experimental Result | |
---|---|---|---|---|---|---|
Accuracy (%) | Time Consuming (s) | |||||
Scheme 1 | ▲ | 69.7 | 97.93 | |||
Scheme 2 | ▲ | ▲ | 70.4 | 128.05 | ||
Scheme 3 | ▲ | 74.3 | 102.05 | |||
Scheme 4 | ▲ | ▲ | 75.1 | 130.57 | ||
Scheme 5 | ▲ | ▲ | 81.1 | 127.17 | ||
Scheme 6 | ▲ | ▲ | ▲ | ▲ | 81.9 | 193.27 |
Experimental Scheme | Extensor Carpi Ulnaris | Flexor Carpi Radialis | Extensor Carpi Radialis | Flexor Carpi Ulnaris | Experimental Result | |
---|---|---|---|---|---|---|
Accuracy (%) | Time Consuming (s) | |||||
Scheme 1 | ▲ | 71.7 | 97.93 | |||
Scheme 2 | ▲ | ▲ | 72.5 | 128.05 | ||
Scheme 3 | ▲ | 64.9 | 101.93 | |||
Scheme 4 | ▲ | ▲ | 65.7 | 132.57 | ||
Scheme 5 | ▲ | ▲ | 80.3 | 131.53 | ||
Scheme 6 | ▲ | ▲ | ▲ | ▲ | 80.9 | 197.01 |
Input Size | Network Block | Expansion Factor (t) | Output Channels (c) | Number of Repetitions (n) | Step (s) |
---|---|---|---|---|---|
448 × 448 × 3 | Conv2D | - | 32 | 1 | 2 |
224 × 224 × 32 | BottleNeck | 1 | 16 | 1 | 1 |
224 × 224 × 16 | BottleNeck | 6 | 24 | 2 | 2 |
112 × 112 × 24 | BottleNeck | 6 | 32 | 3 | 2 |
56 × 56 × 32 | Ghost | - | 64 | 4 | 2 |
28 × 28 × 64 | BottleNeck | 6 | 96 | 3 | 2 |
14 × 14 × 96 | BottleNeck | 6 | 128 | 3 | 1 |
14 × 14 × 128 | BottleNeck | 6 | 160 | 3 | 2 |
7 × 7 × 160 | BottleNeck | 6 | 320 | 1 | 1 |
7 × 7 × 320 | Conv2D 1 × 1 | - | 1280 | 1 | 1 |
7 × 7 × 1280 | Avgpool 7 × 7 | - | - | 1 | - |
1 × 1 × 1280 | Conv2D | - | K = 6 | - | - |
Feature Extraction Layer | Input Size | Parameter Quantity | Model Size |
---|---|---|---|
3 × 3 × 3 × 32 | 448, 448, 3 | 864 | - |
dw 3 × 3 × 32 | 224, 224, 32 | 288 | - |
pw 1 × 1 × 32 × 16 | 224, 224, 32 | 512 | - |
pw 1 × 1 × 16 × (16 × 6) | 224, 244, 16 | 1536 | 2 |
dw 3 × 3 × (16 × 6) | 224, 224, (16 × 6) | 864 | 2 |
pw 1 × 1 × (16 × 6) × 24 | 112, 112, (16 × 6) | 2304 | 2 |
pw 1 × 1 × 24 × (24 × 6) | 112, 112, 24 | 3456 | 3 |
dw 3 × 3 × (24 × 6) | 56, 56, (24 × 6) | 1296 | 3 |
pw 1 × 1 × (24 × 6) × 32 | 56, 56, (24 × 6) | 4608 | 3 |
pc 1 × 1 × 32 × (64/2) | 56, 56, 32 | 1024 | 4 |
co 3 × 3 × (64/2) | 56, 56, 32 | 288 | 4 |
pw 1 × 1 × 64 × (64 × 6) | 28, 28, 64 | 24,576 | 3 |
dw 3 × 3 × (64 × 6) | 14, 14, (64 × 6) | 3456 | 3 |
pw 1 × 1 × (64 × 6) × 96 | 14, 14, (64 × 6) | 36,864 | 3 |
pw 1 × 1 × 96 × (96 × 6) | 14, 14, 96 | 55,296 | 3 |
dw 3 × 3 × (96 × 6) | 7, 7, (96 × 6) | 5184 | 3 |
pw 1 × 1 × (96 × 6) × 160 | 7, 7, (96 × 6) | 92,160 | 3 |
pw 1 × 1 × 160 × (160 × 6) | 7, 7, 160 | 153,600 | 1 |
dw 3 × 3 × (160 × 6) | 7, 7, (160 × 6) | 8640 | 1 |
pw 1 × 1 × (160 × 6) × 320 | 7, 7, (160 × 6) | 307,200 | 1 |
1 × 1 × 320 × 1280 | 1, 1, 320 | 409,600 | - |
1 × 1 × 1280 × 6 | 1, 1, 1280 | 7716 | - |
All | - | 1,583,464 | 1.52 M- |
Feature Extract Network Model | Accuracy(%) | Time Consuming (s) |
---|---|---|
MobileNet-V2 | 81.73 | 131.533 |
GhostNet | 67.56 | 97.702 |
MGNet | 81.91 | 109.392 |
Processing Speed | Response Time | mAP | |
---|---|---|---|
Verification data effect | 1.7 s/900 images | 17.97 ms for each image | 80.7% |
Methods | Best Accuracy | mAP | Angle Predict Error (°) | Predict | ||
---|---|---|---|---|---|---|
Train | Validation | Train | Validation | |||
SVM-RBF | 88.7% | 85.8% | 84.8% | 83.6% | - | Only six motion classifications |
Our methods | 85.7% | 82.4% | 82.3% | 80.7% | Six montion classifications and joint movement angle |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Bu, D.; Guo, S.; Li, H. sEMG-Based Motion Recognition of Upper Limb Rehabilitation Using the Improved Yolo-v4 Algorithm. Life 2022, 12, 64. https://doi.org/10.3390/life12010064
Bu D, Guo S, Li H. sEMG-Based Motion Recognition of Upper Limb Rehabilitation Using the Improved Yolo-v4 Algorithm. Life. 2022; 12(1):64. https://doi.org/10.3390/life12010064
Chicago/Turabian StyleBu, Dongdong, Shuxiang Guo, and He Li. 2022. "sEMG-Based Motion Recognition of Upper Limb Rehabilitation Using the Improved Yolo-v4 Algorithm" Life 12, no. 1: 64. https://doi.org/10.3390/life12010064