Improved Surface Electromyogram-Based Hand–Wrist Force Estimation Using Deep Neural Networks and Cross-Joint Transfer Learning

Deep neural networks (DNNs) and transfer learning (TL) have been used to improve surface electromyogram (sEMG)-based force estimation. However, prior studies focused mostly on applying TL within one joint, which limits dataset size and diversity. Herein, we investigated cross-joint TL between two up...

Full description

Saved in:
Bibliographic Details
Main Authors: Haopeng Wang, He Wang, Chenyun Dai, Xinming Huang, Edward A. Clancy
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/24/22/7301
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850227786320445440
author Haopeng Wang
He Wang
Chenyun Dai
Xinming Huang
Edward A. Clancy
author_facet Haopeng Wang
He Wang
Chenyun Dai
Xinming Huang
Edward A. Clancy
author_sort Haopeng Wang
collection DOAJ
description Deep neural networks (DNNs) and transfer learning (TL) have been used to improve surface electromyogram (sEMG)-based force estimation. However, prior studies focused mostly on applying TL within one joint, which limits dataset size and diversity. Herein, we investigated cross-joint TL between two upper-limb joints with four DNN architectures using sliding windows. We used two feedforward and two recurrent DNN models with feature engineering and feature learning, respectively. We found that the dependencies between sEMG and force are short-term (<400 ms) and that sliding windows are sufficient to capture them, suggesting that more complicated recurrent structures may not be necessary. Also, using DNN architectures reduced the required sliding window length. A model pre-trained on elbow data was fine-tuned on hand–wrist data, improving force estimation accuracy and reducing the required training data amount. A convolutional neural network with a 391 ms sliding window fine-tuned using 20 s of training data had an error of 6.03 ± 0.49% maximum voluntary torque, which is statistically lower than both our multilayer perceptron model with TL and a linear regression model using 40 s of training data. The success of TL between two distinct joints could help enrich the data available for future deep learning-related studies.
format Article
id doaj-art-8553c584bac749189a624ee5e38982a3
institution OA Journals
issn 1424-8220
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj-art-8553c584bac749189a624ee5e38982a32025-08-20T02:04:44ZengMDPI AGSensors1424-82202024-11-012422730110.3390/s24227301Improved Surface Electromyogram-Based Hand–Wrist Force Estimation Using Deep Neural Networks and Cross-Joint Transfer LearningHaopeng Wang0He Wang1Chenyun Dai2Xinming Huang3Edward A. Clancy4Department of Electrical and Computer Engineering, Worcester Polytechnic Institute, Worcester, MA 01609, USADepartment of Electrical and Computer Engineering, Worcester Polytechnic Institute, Worcester, MA 01609, USASchool of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai 200241, ChinaDepartment of Electrical and Computer Engineering, Worcester Polytechnic Institute, Worcester, MA 01609, USADepartment of Electrical and Computer Engineering, Worcester Polytechnic Institute, Worcester, MA 01609, USADeep neural networks (DNNs) and transfer learning (TL) have been used to improve surface electromyogram (sEMG)-based force estimation. However, prior studies focused mostly on applying TL within one joint, which limits dataset size and diversity. Herein, we investigated cross-joint TL between two upper-limb joints with four DNN architectures using sliding windows. We used two feedforward and two recurrent DNN models with feature engineering and feature learning, respectively. We found that the dependencies between sEMG and force are short-term (<400 ms) and that sliding windows are sufficient to capture them, suggesting that more complicated recurrent structures may not be necessary. Also, using DNN architectures reduced the required sliding window length. A model pre-trained on elbow data was fine-tuned on hand–wrist data, improving force estimation accuracy and reducing the required training data amount. A convolutional neural network with a 391 ms sliding window fine-tuned using 20 s of training data had an error of 6.03 ± 0.49% maximum voluntary torque, which is statistically lower than both our multilayer perceptron model with TL and a linear regression model using 40 s of training data. The success of TL between two distinct joints could help enrich the data available for future deep learning-related studies.https://www.mdpi.com/1424-8220/24/22/7301electromyogramtransfer learningdeep neural networksforce estimationCNNLSTM
spellingShingle Haopeng Wang
He Wang
Chenyun Dai
Xinming Huang
Edward A. Clancy
Improved Surface Electromyogram-Based Hand–Wrist Force Estimation Using Deep Neural Networks and Cross-Joint Transfer Learning
Sensors
electromyogram
transfer learning
deep neural networks
force estimation
CNN
LSTM
title Improved Surface Electromyogram-Based Hand–Wrist Force Estimation Using Deep Neural Networks and Cross-Joint Transfer Learning
title_full Improved Surface Electromyogram-Based Hand–Wrist Force Estimation Using Deep Neural Networks and Cross-Joint Transfer Learning
title_fullStr Improved Surface Electromyogram-Based Hand–Wrist Force Estimation Using Deep Neural Networks and Cross-Joint Transfer Learning
title_full_unstemmed Improved Surface Electromyogram-Based Hand–Wrist Force Estimation Using Deep Neural Networks and Cross-Joint Transfer Learning
title_short Improved Surface Electromyogram-Based Hand–Wrist Force Estimation Using Deep Neural Networks and Cross-Joint Transfer Learning
title_sort improved surface electromyogram based hand wrist force estimation using deep neural networks and cross joint transfer learning
topic electromyogram
transfer learning
deep neural networks
force estimation
CNN
LSTM
url https://www.mdpi.com/1424-8220/24/22/7301
work_keys_str_mv AT haopengwang improvedsurfaceelectromyogrambasedhandwristforceestimationusingdeepneuralnetworksandcrossjointtransferlearning
AT hewang improvedsurfaceelectromyogrambasedhandwristforceestimationusingdeepneuralnetworksandcrossjointtransferlearning
AT chenyundai improvedsurfaceelectromyogrambasedhandwristforceestimationusingdeepneuralnetworksandcrossjointtransferlearning
AT xinminghuang improvedsurfaceelectromyogrambasedhandwristforceestimationusingdeepneuralnetworksandcrossjointtransferlearning
AT edwardaclancy improvedsurfaceelectromyogrambasedhandwristforceestimationusingdeepneuralnetworksandcrossjointtransferlearning