Related Experiment Video
Updated: May 14, 2026
08:15
Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision
Published on: March 28, 2025
Continuous estimation of finger joint angles using muscle activation inputs from surface EMG signals
Jimson Ngeo1, Tomoya Tamei, Tomohiro Shibata
1Graduate School of Information Science, Nara Institute of Science and Technology, 8916-5 Takayama, Ikoma City, Nara 630-0192, Japan. jimson-n@is.naist.jp
Summary
This study presents a novel method to predict finger joint angles using surface electromyography (sEMG) signals by accounting for electromechanical delay (EMD). This approach enhances control for virtual reality, prosthetics, and rehabilitation aids.
Area of Science:
- Biomedical Engineering
- Neuroscience
- Robotics
Background:
- Surface electromyography (sEMG) signals are crucial for inferring motor intentions.
- Accurate prediction of hand and finger movements is vital for human-computer interfaces.
- Electromechanical delay (EMD) presents a challenge in correlating sEMG with actual movement.
Purpose of the Study:
- To develop a method for estimating finger joint angles from sEMG signals.
- To incorporate and automatically optimize for electromechanical delay (EMD) in the prediction model.
- To validate the method for both periodic and non-periodic finger movements.
Main Methods:
- Utilized muscle activation derived from sEMG signals as input to a neural network.
- Developed a muscle activation model that parameterizes and optimizes for EMD.
- Applied the model to predict finger joint angles during various hand movements.
Main Results:
- Achieved high correlation (up to 0.92) between actual and predicted metacarpophalangeal (MCP) joint angles for periodic movements.
- Demonstrated strong prediction accuracy (up to 0.85) for more dynamic, non-periodic movements.
- Successfully estimated finger joint angles considering EMD in sEMG-based prediction.
Conclusions:
- The proposed method effectively predicts finger joint angles from sEMG signals, even with inherent EMD.
- This technique offers improved control for applications like virtual reality, robotic prosthetics, and rehabilitation.
- The accurate prediction of dynamic movements opens new possibilities for intuitive human-machine interaction.
