The Influence of Gait Phase on Predicting Lower-Limb Joint Angles

Machine learning has seen a rapid increase in applications that harness wearable signals to improve human mobility. Previous work has used machine learning predictors as a means of continuously estimating locomotor intent. Although previous studies perform gait phase classification and joint-level angular prediction, there are currently no studies that compare joint-level prediction performance at various phases of gait. As such, the purpose of this offline study was to analyze how machine learning and deep learning models perform at predicting future joint angles during various phases of gait. EMG, IMU, and joint kinematics collected during level-ground walking from thirty participants and data was separated into six distinct gait phases. Random forest, long short-term memory (LSTM), and bidirectional LSTM was used to predict lower-limb joint angles during the phases of gait. Results indicate that bidirectional LSTM is the most robust performer across the gait cycle, with a mean prediction RMSE of 1.42-5.71 degrees. Our study shows how deep learning methods, such as bidirectional LSTM, can accurately estimate joint angles throughout the gait cycle. Furthermore, we propose future work of deploying models which accurately predict future joint angles throughout the gait cycle for users to sufficiently operate a wearable exoskeleton during locomotion.

[1]  Mykel J. Kochenderfer,et al.  Personalizing exoskeleton assistance while walking in the real world , 2022, Nature.

[2]  Dean D. Molinaro,et al.  Subject-Independent Continuous Locomotion Mode Classification for Robotic Hip Exoskeleton Applications , 2022, IEEE Transactions on Biomedical Engineering.

[3]  Dean D. Molinaro,et al.  Predicting biological joint moment during multiple ambulation tasks. , 2022, Journal of biomechanics.

[4]  Qiaoling Meng,et al.  Flexible lower limb exoskeleton systems: A review. , 2022, NeuroRehabilitation.

[5]  Dean D. Molinaro,et al.  Subject-Independent, Biological Hip Moment Estimation During Multimodal Overground Ambulation Using Deep Learning , 2022, IEEE Transactions on Medical Robotics and Bionics.

[6]  P. Kent,et al.  Predicting Knee Joint Kinematics from Wearable Sensor Data in People with Knee Osteoarthritis and Clinical Considerations for Future Machine Learning Models , 2022, Sensors.

[7]  Juanjuan Zhang,et al.  Improving Walking Economy With an Ankle Exoskeleton Prior to Human-in-the-Loop Optimization , 2022, Frontiers in Neurorobotics.

[8]  Sean Gallagher,et al.  EMG and Joint Angle-Based Machine Learning to Predict Future Joint Angles at the Knee , 2021, Sensors.

[9]  Kurt L. Mudie,et al.  The need for a paradigm shift in the development of military exoskeletons , 2021, European journal of sport science.

[10]  N. Fey,et al.  Continuous Classification of Locomotion in Response to Task Complexity and Anticipatory State , 2021, Frontiers in Bioengineering and Biotechnology.

[11]  Sanam Moghaddamnia,et al.  Efficiency of deep neural networks for joint angle modeling in digital gait assessment , 2021, EURASIP Journal on Advances in Signal Processing.

[12]  Gwendolyn M. Bryan,et al.  A hip–knee–ankle exoskeleton emulator for studying gait assistance , 2020, Int. J. Robotics Res..

[13]  Wen Wu Using Reinforcement Learning to Estimate Human Joint Moments From Electromyography or Joint Kinematics: An Alternative Solution to Musculoskeletal-Based Biomechanics , 2021 .

[14]  Goldie Nejat,et al.  End-to-End Deep Reinforcement Learning for Exoskeleton Control , 2020, 2020 IEEE International Conference on Systems, Man, and Cybernetics (SMC).

[15]  R. Nataraj,et al.  Disproportionate positive feedback facilitates sense of agency and performance for a reaching movement task with a virtual hand , 2020, PloS one.

[16]  Wolfgang Potthast,et al.  Estimation of Gait Mechanics Based on Simulated and Measured IMU Data Using an Artificial Neural Network , 2020, Frontiers in Bioengineering and Biotechnology.

[17]  Pratik Kunapuli,et al.  Real-Time Neural Network-Based Gait Phase Estimation Using a Robotic Hip Exoskeleton , 2020, IEEE Transactions on Medical Robotics and Bionics.

[18]  Xiaorong Guan,et al.  Estimation of Knee Movement from Surface EMG Using Random Forest with Principal Component Analysis , 2019, Electronics.

[19]  W. Potthast,et al.  Prediction of lower limb joint angles and moments during gait using artificial neural networks , 2019, Medical & Biological Engineering & Computing.

[20]  Longhan Xie,et al.  A Continuous Estimation Model of Upper Limb Joint Angles by Using Surface Electromyography and Deep Learning Method , 2019, IEEE Access.

[21]  Reed D. Gurchiek,et al.  Estimating Biomechanical Time-Series with Wearable Sensors: A Systematic Review of Machine Learning Techniques , 2019, Sensors.

[22]  Emily N. Ussery,et al.  Walking as an Opportunity for Cardiovascular Disease Prevention , 2019, Preventing chronic disease.

[23]  Gian Maria Gasparri,et al.  Proportional Joint-Moment Control for Instantaneously Adaptive Ankle Exoskeleton Assistance , 2019, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[24]  Bing Chen,et al.  User-Adaptive Assistance of Assistive Knee Braces for Gait Rehabilitation , 2018, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[25]  Jun Morimoto,et al.  EMG-Based Model Predictive Control for Physical Human–Robot Interaction: Application for Assist-As-Needed Control , 2018, IEEE Robotics and Automation Letters.

[26]  Elliott J. Rouse,et al.  Design and Validation of a Torque-Controllable Knee Exoskeleton for Sit-to-Stand Assistance , 2017, IEEE/ASME Transactions on Mechatronics.

[27]  Edward D. Lemaire,et al.  Gait phase detection from thigh kinematics using machine learning techniques , 2017, 2017 IEEE International Symposium on Medical Measurements and Applications (MeMeA).

[28]  Dirk De Clercq,et al.  Reducing the metabolic cost of walking with an ankle exoskeleton: interaction between actuation timing and power , 2017, Journal of NeuroEngineering and Rehabilitation.

[29]  Daniel P. Ferris,et al.  State of the Art and Future Directions for Lower Limb Robotic Exoskeletons , 2017, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[30]  François Michaud,et al.  Exoskeletons' design and usefulness evidence according to a systematic review of lower limb exoskeletons used for functional mobility by people with spinal cord injury , 2016, Disability and rehabilitation. Assistive technology.

[31]  Robert Riener,et al.  A survey of sensor fusion methods in wearable robotics , 2015, Robotics Auton. Syst..

[32]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[33]  Robert Riener,et al.  Control strategies for active lower extremity prosthetics and orthotics: a review , 2015, Journal of NeuroEngineering and Rehabilitation.

[34]  Hugh M. Herr,et al.  Volitional control of ankle plantar flexion in a powered transtibial prosthesis during stair-ambulation , 2014, 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[35]  Quoc V. Le,et al.  Sequence to Sequence Learning with Neural Networks , 2014, NIPS.

[36]  Wendy L. Boehm,et al.  Ankle torque control that shifts the center of pressure from heel to toe contributes non-zero sagittal plane angular momentum during human walking. , 2014, Journal of biomechanics.

[37]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[38]  Brendan J. Frey,et al.  Adaptive dropout for training deep neural networks , 2013, NIPS.

[39]  Fan Zhang,et al.  Source Selection for Real-Time User Intent Recognition Toward Volitional Control of Artificial Legs , 2013, IEEE Journal of Biomedical and Health Informatics.

[40]  Michael H Schwartz,et al.  Predicting the outcome of intramuscular psoas lengthening in children with cerebral palsy using preoperative gait data and the random forest algorithm. , 2013, Gait & posture.

[41]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[42]  Hugh M. Herr,et al.  A method to determine the optimal features for control of a powered lower-limb prostheses , 2011, 2011 Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[43]  Richard W. Bohannon,et al.  Normal walking speed: a descriptive meta-analysis. , 2011, Physiotherapy.

[44]  Fan Zhang,et al.  Continuous Locomotion-Mode Identification for Prosthetic Legs Based on Neuromuscular–Mechanical Fusion , 2011, IEEE Transactions on Biomedical Engineering.

[45]  He Huang,et al.  A Strategy for Identifying Locomotion Modes Using Surface Electromyography , 2009, IEEE Transactions on Biomedical Engineering.

[46]  P. Cavanagh,et al.  Electromechanical delay in human skeletal muscle under concentric and eccentric contractions , 1979, European Journal of Applied Physiology and Occupational Physiology.

[47]  Jürgen Schmidhuber,et al.  Applying LSTM to Time Series Predictable through Time-Window Approaches , 2000, ICANN.

[48]  T P Andriacchi,et al.  A point cluster method for in vivo motion analysis: applied to a study of knee kinematics. , 1998, Journal of biomechanical engineering.

[49]  Howard Jay Chizeck,et al.  Fuzzy model identification for classification of gait events in paraplegics , 1997, IEEE Trans. Fuzzy Syst..

[50]  Michael Y. Hu,et al.  Effect of data standardization on neural network training , 1996 .

[51]  V M Zatsiorsky,et al.  Tendon action of two-joint muscles: transfer of mechanical energy between joints during jumping, landing, and running. , 1994, Journal of biomechanics.

[52]  H.B.K. Boom,et al.  Automatic stance-swing phase detection from accelerometer data for peroneal nerve stimulation , 1990, IEEE Transactions on Biomedical Engineering.

[53]  E S Grood,et al.  A joint coordinate system for the clinical description of three-dimensional motions: application to the knee. , 1983, Journal of biomechanical engineering.