Abstract:Conventional measurement of spasticity in stroke patients, e.g., modified Ashworth scale (MAS), has been challenged about its reliability issues. Surface electromyography (sEMG) has been used to identify neuromuscular abnormalities since it directly measures electrical activity in the muscle, however its performance is affected by the placement of electrodes and crosstalk, and it cannot detect the activities of deep muscles. This study proposes a novel spasticity measurement method by quantifying the difference between the impaired and unaffected sides in the elbow maximal voluntary contraction (MVC) task. Five inertial measurement units (IMUs) and a force sensor were used to capture the movement dynamics for the MVC test, by which a neuromusculoskeletal model is established to estimate the muscle activation using the inverse dynamics and optimization techniques. Normalized keeping time of peak activation is a quantitative feature that identifies the disparity between the impaired and unaffected side in the MVC test is defined as a measurement of spasticity. Six stroke patients and eight healthy subjects were recruited to evaluate the muscle activation estimation model. The outcomes of our measurement for patients were compared with the spasticity rated by an experienced physical therapist (PT) using MAS. The estimated muscle activation shows promising accuracy compared to the sEMG profiles (patients: mean ; healthy: mean ). The outcomes of our approach are highly correlated with MAS (Pearson's , ). These findings indicate that our approach can provide a quantitative - easure of spasticity and can be used as a complementary measurement along with the existing clinical methods. This approach will also enhance the efficiency of upper limb robot-aided rehabilitation in stroke patients.

Estimation of Muscle Activation During Ankle Rehabilitation.

A Subject-Specific EMG-Driven Musculoskeletal Model for Applications in Lower-Limb Rehabilitation Robotics

SEMG-based multifeatures and predictive model for knee-joint-angle estimation

Research on a Calculation Model of Ankle-Joint-Torque-Based sEMG

Ankle Joint Torque Prediction Based On Surface Electromyographic And Angular Velocity Signals

Prediction of Actively Exerted Torque From Ankle Joint Complex Based on Muscle Synergy

Evaluation of Different Robot-assisted Ankle Muscle Strength Training Modes based on Angle/Torque and sEMG Characteristics

A Close-Loop EMG Model for Continuous Joint Movements Estimation of a Rehabilitation Robot

Assessment of ankle muscle activation by muscle synergies in healthy and post-stroke gait.

Muscle force estimation from lower limb EMG signals using novel optimised machine learning techniques

Simultaneous and Continuous Estimation of Joint Angles Based on Surface Electromyography State-Space Model

A State Equation Model for Joint Moment Estimation Based on SEMG

Quantitative Elbow Spasticity Measurement Based on Muscle Activation Estimation Using Maximal Voluntary Contraction

A Temporally Smoothed MLP Regression Scheme for Continuous Knee/Ankle Angles Estimation by Using Multi-Channel Semg

Semg Based Movement Quantitative Estimation of Joins Using SVM Method

Estimation and Correlation Analysis of Lower Limb Joint Angles Based on Surface Electromyography

Evaluation of Non-Invasive Ankle Joint Effort Prediction Methods for Use in Neurorehabilitation Using Electromyography and Ultrasound Imaging

Single-Channel Semg-Based Estimation of Knee Joint Angle Using a Decomposition Algorithm with a State-Space Model

A Novel Estimation Approach of Semg-Based Joint Movements Via RBF Neural Network

A Subject-Specific Neuromusculoskeletal Model for Muscle Force Estimation Used in Rehabilitation

The Classification of Surface Electromyographic for Ankle Eversion and Inversion Based on Cerebellar Model Neural Networks