Lower Limb Activity Recognition using sEMG Signals via Weighted Random Forest

Identifying the motion intention of the wearer and integrating it into the lower limb rehabilitation device is a rehabilitation interaction method that can effectively improve the rehabilitation effect. Existing surface electromyography (sEMG) based intention identification methods have the shortcomings of limited generalization ability in practical applications. In this paper, we proposed a weighted random forest (WRF) algorithm for the intention prediction. The performance of the ensemble model is improved by weighted ensemble of each base prediction decision tree. Detailed comparative studies between proposed method and conventional methods have been carried out through the lower limb activity identification data at different time periods. Compared with Linear Discriminant Analysis (LDA), Gradient Boosting (GB), AdaBoost, K-Nearest Neighbor (KNN), Random Forest (RF), the generalization ability of WRF is significantly superior considering the best accuracy and the best precision achieved by WRF. Compared to standard RF, the WRF improved classification accuracy performance by 1.19% and the classification precision accuracy performance by 1.1% on average. Conclusively, the proposed method can provide satisfactory generalization performance.

[1]  Saeid Rashidi,et al.  Finger movements classification based on fractional Fourier transform coefficients extracted from surface EMG signals , 2021, Biomed. Signal Process. Control..

[2]  Alireza Souri,et al.  Hyper-parameter tuned light gradient boosting machine using memetic firefly algorithm for hand gesture recognition , 2021, Appl. Soft Comput..

[3]  Xin Zhang,et al.  Novel Design and Adaptive Fuzzy Control of a Lower-Limb Elderly Rehabilitation , 2020 .

[4]  Shouqian Sun,et al.  A Low-Cost End-to-End sEMG-Based Gait Sub-Phase Recognition System , 2020, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[5]  Rafael R. Torrealba,et al.  Toward the Development of Knee Prostheses: Review of Current Active Devices , 2019, Applied Mechanics Reviews.

[6]  Ke Zhang,et al.  Surface electromyography–based hand movement recognition using the Gaussian mixture model, multilayer perceptron, and AdaBoost method , 2019, Int. J. Distributed Sens. Networks.

[7]  Min Sheng,et al.  A CNN-Based Method for Intent Recognition Using Inertial Measurement Units and Intelligent Lower Limb Prosthesis , 2019, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[8]  Krzysztof Gajowniczek,et al.  Reducing False Arrhythmia Alarms Using Different Methods of Probability and Class Assignment in Random Forest Learning Methods , 2019, Sensors.

[9]  Lev V. Utkin,et al.  A weighted random survival forest , 2019, Knowl. Based Syst..

[10]  Hao Wang,et al.  An Advanced Adaptive Control of Lower Limb Rehabilitation Robot , 2018, Front. Robot. AI.

[11]  Dingguo Zhang,et al.  sEMG Bias-Driven Functional Electrical Stimulation System for Upper-Limb Stroke Rehabilitation , 2018, IEEE Sensors Journal.

[12]  Ganesh R. Naik,et al.  An ICA-EBM-Based sEMG Classifier for Recognizing Lower Limb Movements in Individuals With and Without Knee Pathology , 2018, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[13]  José Luis Contreras-Vidal,et al.  Prediction of lower-limb joint kinematics from surface EMG during overground locomotion , 2017, 2017 IEEE International Conference on Systems, Man, and Cybernetics (SMC).

[14]  Nader Meskin,et al.  An EMG signal processing system for control of an ankle-foot orthosis , 2017, 2017 4th International Conference on Control, Decision and Information Technologies (CoDIT).

[15]  C. Brockett,et al.  Biomechanics of the ankle , 2016, Orthopaedics and trauma.

[16]  Xu Zhang,et al.  Random Forest-Based Recognition of Isolated Sign Language Subwords Using Data from Accelerometers and Surface Electromyographic Sensors , 2016, Sensors.

[17]  Siti Anom Ahmad,et al.  Classification of ankle joint movements based on surface electromyography signals for rehabilitation robot applications , 2015, Medical & Biological Engineering & Computing.

[18]  Yu Song,et al.  Estimation of Lower Limb Periodic Motions from sEMG Using Least Squares Support Vector Regression , 2015, Neural Processing Letters.

[19]  Marco Pirini,et al.  The ABC of EMG , 2014 .

[20]  Xuan Zhang,et al.  Human lower extremity joint moment prediction: A wavelet neural network approach , 2014, Expert Syst. Appl..

[21]  Dario Farina,et al.  The Extraction of Neural Information from the Surface EMG for the Control of Upper-Limb Prostheses: Emerging Avenues and Challenges , 2014, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[22]  Juan José Rodríguez Diez,et al.  A weighted voting framework for classifiers ensembles , 2014, Knowledge and Information Systems.

[23]  Angkoon Phinyomark,et al.  EMG feature evaluation for improving myoelectric pattern recognition robustness , 2013, Expert Syst. Appl..

[24]  George C. Runger,et al.  System Monitoring with Real-Time Contrasts , 2012 .

[25]  Chi-Woong Mun,et al.  Comparison of k-nearest neighbor, quadratic discriminant and linear discriminant analysis in classification of electromyogram signals based on the wrist-motion directions , 2011 .

[26]  Shuichi Obuchi,et al.  Effect of Ankle Weights on EMG Activity of the Semitendinosus and Knee Stability During Walking by Community-Dwelling Elderly , 2011 .

[27]  Manfred Morari,et al.  Automatic gait-pattern adaptation algorithms for rehabilitation with a 4-DOF robotic orthosis , 2004, IEEE Transactions on Robotics and Automation.

[28]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[29]  G. Lichtwark,et al.  Muscle fascicle and series elastic element length changes along the length of the human gastrocnemius during walking and running. , 2007, Journal of biomechanics.