Related Experiment Video
Updated: Dec 6, 2025
08:15
Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision
Published on: March 28, 2025
1.0K
Real-time finger force prediction via parallel convolutional neural networks: a preliminary study.
Summary
This study introduces a new method using convolutional neural networks (CNNs) to predict finger forces from high-density electromyogram (HD-EMG) signals in real-time. The CNN approach achieved high accuracy, outperforming traditional methods for human-machine interaction.
Area of Science:
- Biomedical Engineering
- Neuroscience
- Machine Learning
Background:
- Accurate decoding of intended motions is crucial for effective human-machine interaction (HMI).
- Existing methods for predicting force from electromyogram (EMG) signals have limitations in accuracy and real-time application.
- Motor unit discharge frequency is a key neural signal for force control.
Purpose of the Study:
- To develop a novel, real-time continuous prediction method for individual finger forces.
- To utilize parallel convolutional neural networks (CNNs) for extracting motor unit discharge frequency from high-density EMG (HD-EMG) without spike sorting.
- To compare the performance of the CNN-based approach against traditional motor unit decomposition and EMG amplitude methods.
Main Methods:
- Developed a parallel CNN architecture to extract populational motor unit discharge frequency from HD-EMG signals.
- Trained CNN parameters using temporal energy heatmaps and frequency spectrum maps derived from HD-EMG.
- Used predicted motor unit discharge frequency to estimate finger forces via a linear regression model.
- Validated the approach by comparing predicted forces against recorded forces, assessing correlation coefficients.
Main Results:
- The CNN-based approach achieved a high average correlation coefficient of 0.91 between predicted and recorded forces.
- This performance surpassed offline decomposition (0.89), online decomposition (0.82), and conventional EMG amplitude methods (0.81).
- Demonstrated generalizable performance, with CNNs trained on one finger showing applicability to others.
Conclusions:
- The developed CNN-based algorithm provides an accurate and efficient method for real-time continuous force decoding.
- This approach holds significant potential for advancing human-machine interaction technologies.
- The method offers a robust alternative to traditional EMG signal processing techniques for force prediction.
Related Concept Videos
01:20Parallel Processing
482
The brain processes sensory information rapidly due to parallel processing, which involves sending data across multiple neural pathways at the same time. This method allows the brain to manage various sensory qualities, such as shapes, colors, movements, and locations, all concurrently. For instance, when observing a forest landscape, the brain simultaneously processes the movement of leaves, the shapes of trees, the depth between them, and the various shades of green. This enables a quick and...
482
01:22Force Classification
2.1K
Forces play a crucial role in the study of physics and engineering. They are essential in describing the motion, behavior, and equilibrium of objects in the physical world. Forces can be classified based on their origin, type, and direction of action.
Contact and non-contact forces are two of the most widely used categories of forces. As the name suggests, contact forces require physical contact between two objects to act upon each other. Examples of contact forces include frictional,...
Contact and non-contact forces are two of the most widely used categories of forces. As the name suggests, contact forces require physical contact between two objects to act upon each other. Examples of contact forces include frictional,...
2.1K