Estimating the Relationship Between Multivariable Standing Ankle Impedance and Lower Extremity Muscle Activation

Knop, Lauren; Ribeiro, Guilherme Aramizo; Ficanha, Evandro M.; Rastgaar, Mo

doi:10.1109/biorob.2018.8488121

Cited by 3 publications

(7 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This small change suggests that either the parameter was not modulated enough to be related to the muscle signal, or the parameter was more susceptible to noise in the model. When comparing to the previous study [8], the results of the 1-vs-All method improved with the addition of new EMG features and new model types. The previous work, which used only an ANN model to relate EMG and impedance, resulted in an average NMSE of 0.67 for KDP, 0.18 for KIE, 0.19 for BDP, and -0.07 for BIE.…”

Section: Regression Model Trainingmentioning

confidence: 76%

“…The same experimental procedure was used in our previous studies [8,9], which estimated standing ankle impedance with the use of the instrumented platform. This platform ( Fig.…”

Section: Methodsmentioning

confidence: 99%

“…Each model was trained using the training data set, which consisted of EMG features (ranging from 4 up to 24) as inputs and 4 impedance parameters as the target. Three methods of training were tested, as previously described in [8]. The first method included generating an individual model for each of the subjects and testing the performance using only the data from the subject-specific model.…”

Section: Training and Testingmentioning

confidence: 99%

“…As expected, ankle stiffness generally increased with higher muscle activation; however, not all subjects exhibited a linear relationship. To better explain this relationship, an Artificial Neural Network (ANN) with a single hidden layer was implemented [8,9]. The resulting subject-dependent models were able to predict ankle impedance in DP and IE directions with greater than 95% accuracy using the corresponding lower extremity muscle co-contraction levels.…”

Section: Introductionmentioning

confidence: 99%

“…The purpose of this study was to examine ways to improve the performance of a generalized model that can estimate ankle impedance based on a subject's EMG signals that were unseen to the model during training. Previous work by the authors studied the relationship between standing ankle impedance and lower extremity muscle co-contraction level [8,9]. These studies determined the performance of (1) an individual ANN model optimized for each separate subject, (2) an aggregated ANN model, which used the data of all the subject population (12 subjects) to train and test a single ANN model, and (3) a "1-vs-All" method, where a model was trained with the data of − 1 subjects and the performance was tested with the remaining subject to verify the generalization of the model.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Towards a Generalized Model of Multivariable Ankle Impedance During Standing Based on the Lower Extremity Muscle EMG

Knop

Ribeiro

Rastgaar

2019

2019 Design of Medical Devices Conference

Self Cite

View full text Add to dashboard Cite

The ankle mechanical impedance of healthy subjects was estimated during the standing pose while they co-contracted their lower-leg muscles. Subsequently, the impedance parameters were modeled as a function of the level of co-contraction using machine learning regression methods. From the experimental results, the average ankle stiffness coefficients in dorsi-plantar flexion (DP) showed more dependence to the muscle contraction than stiffness in inversion-eversion (IE): 4.6 Nm/rad per %MVC (percent of the maximum voluntary contraction) and 1.1 Nm/rad per %MVC, respectively. To accurately estimate the ankle impedance parameters as a function of the electromyography (EMG) signals, multiple EMG feature selection methods, regression models, and types of models were evaluated. Using a 1-vs-All model validation approach, the best regression model to fit the stiffness and damping in DP was the Least Square method with Regularization, and the best IE stiffness was the Gaussian Process Regression. No model was able to estimate the IE damping well, possibly because this parameter is not modulated with a changing co-contraction of the lower-leg muscles.

show abstract

Section: Regression Model Trainingmentioning

confidence: 76%

“…The same experimental procedure was used in our previous studies [8,9], which estimated standing ankle impedance with the use of the instrumented platform. This platform ( Fig.…”

Section: Methodsmentioning

confidence: 99%

Section: Training and Testingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations