An official website of the United States government
Abstract Objective . Neural decoding is an important tool in neural engineering and neural data analysis. Of various machine learning algorithms adopted for neural decoding, the recently introduced deep learning is promising to excel. Therefore, we sought to apply deep learning to decode movement trajectories from the activity of motor cortical neurons. Approach . In this paper, we assessed the performance of deep learning methods in three different decoding schemes, concurrent, time-delay, and spatiotemporal. In the concurrent decoding scheme where the input to the network is the neural activity coincidental to the movement, deep learning networks including artificial neural network (ANN) and long-short term memory (LSTM) were applied to decode movement and compared with traditional machine learning algorithms. Both ANN and LSTM were further evaluated in the time-delay decoding scheme in which temporal delays are allowed between neural signals and movements. Lastly, in the spatiotemporal decoding scheme, we trained convolutional neural network (CNN) to extract movement information from images representing the spatial arrangement of neurons, their activity, and connectomes (i.e. the relative strengths of connectivity between neurons) and combined CNN and ANN to develop a hybrid spatiotemporal network. To reveal the input features of the CNN in the hybrid network that deep learning discovered for movement decoding, we performed a sensitivity analysis and identified specific regions in the spatial domain. Main results . Deep learning networks (ANN and LSTM) outperformed traditional machine learning algorithms in the concurrent decoding scheme. The results of ANN and LSTM in the time-delay decoding scheme showed that including neural data from time points preceding movement enabled decoders to perform more robustly when the temporal relationship between the neural activity and movement dynamically changes over time. In the spatiotemporal decoding scheme, the hybrid spatiotemporal network containing the concurrent ANN decoder outperformed single-network concurrent decoders. Significance . Taken together, our study demonstrates that deep learning could become a robust and effective method for the neural decoding of behavior.
more »
« less
Deep Learning-Based Approaches for Decoding Motor Intent From Peripheral Nerve Signals
Previous literature shows that deep learning is an effective tool to decode the motor intent from neural signals obtained from different parts of the nervous system. However, deep neural networks are often computationally complex and not feasible to work in real-time. Here we investigate different approaches' advantages and disadvantages to enhance the deep learning-based motor decoding paradigm's efficiency and inform its future implementation in real-time. Our data are recorded from the amputee's residual peripheral nerves. While the primary analysis is offline, the nerve data is cut using a sliding window to create a “pseudo-online” dataset that resembles the conditions in a real-time paradigm. First, a comprehensive collection of feature extraction techniques is applied to reduce the input data dimensionality, which later helps substantially lower the motor decoder's complexity, making it feasible for translation to a real-time paradigm. Next, we investigate two different strategies for deploying deep learning models: a one-step (1S) approach when big input data are available and a two-step (2S) when input data are limited. This research predicts five individual finger movements and four combinations of the fingers. The 1S approach using a recurrent neural network (RNN) to concurrently predict all fingers' trajectories generally gives better prediction results than all the machine learning algorithms that do the same task. This result reaffirms that deep learning is more advantageous than classic machine learning methods for handling a large dataset. However, when training on a smaller input data set in the 2S approach, which includes a classification stage to identify active fingers before predicting their trajectories, machine learning techniques offer a simpler implementation while ensuring comparably good decoding outcomes to the deep learning ones. In the classification step, either machine learning or deep learning models achieve the accuracy and F1 score of 0.99. Thanks to the classification step, in the regression step, both types of models result in a comparable mean squared error (MSE) and variance accounted for (VAF) scores as those of the 1S approach. Our study outlines the trade-offs to inform the future implementation of real-time, low-latency, and high accuracy deep learning-based motor decoder for clinical applications.
more »
« less
- Award ID(s):
- 1845709
- PAR ID:
- 10315041
- Date Published:
- Journal Name:
- Frontiers in Neuroscience
- Volume:
- 15
- ISSN:
- 1662-453X
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Objective: Robust neural decoding of intended motor output is crucial to enable intuitive control of assistive devices, such as robotic hands, to perform daily tasks. Few existing neural decoders can predict kinetic and kinematic variables simultaneously. The current study developed a continuous neural decoding approach that can concurrently predict fingertip forces and joint angles of multiple fingers. Methods: We obtained motoneuron firing activities by decomposing high-density electromyogram (HD EMG) signals of the extrinsic finger muscles. The identified motoneurons were first grouped and then refined specific to each finger (index or middle) and task (finger force and dynamic movement) combination. The refined motoneuron groups (separate matrix) were then applied directly to new EMG data in real-time involving both finger force and dynamic movement tasks produced by both fingers. EMG-amplitude-based prediction was also performed as a comparison. Results: We found that the newly developed decoding approach outperformed the EMG-amplitude method for both finger force and joint angle estimations with a lower prediction error (Force: 3.47±0.43 vs 6.64±0.69% MVC, Joint Angle: 5.40±0.50° vs 12.8±0.65°) and a higher correlation (Force: 0.75±0.02 vs 0.66±0.05, Joint Angle: 0.94±0.01 vs 0.5±0.05) between the estimated and recorded motor output. The performance was also consistent for both fingers. Conclusion: The developed neural decoding algorithm allowed us to accurately and concurrently predict finger forces and joint angles of multiple fingers in real-time. Significance: Our approach can enable intuitive interactions with assistive robotic hands, and allow the performance of dexterous hand skills involving both force control tasks and dynamic movement control tasks.more » « less
-
null (Ed.)Continuous and accurate decoding of intended motions is critical for human-machine interactions. Here, we developed a novel approach for real-time continuous prediction of forces in individual fingers using parallel convolutional neural networks (CNNs). We extracted populational motor unit discharge frequency using CNNs organized in a parallel structure. The CNNs parameters were trained based on two features from high-density electromyogram (HD-EMG), namely temporal energy heatmaps and frequency spectrum maps. The populational motor unit discharge frequency was then used to continuously predict finger forces based on a linear regression model. The force prediction performance was compared with a motor unit decomposition method and the conventional EMG amplitude-based method. Our results showed that the correlation coefficient between the predicted and the recorded forces of the parallel CNN approach was on average 0.91, compared with an offline decomposition method of 0.89, an online decomposition method of 0.82, and the EMG amplitude method of 0.81. Additionally, the CNN based approach showed generalizable performance, with CNN trained on one finger applying to a different finger. The outcomes suggest that our CNN based algorithm can offer an accurate and efficient force decoding method for human-machine interactions.more » « less
-
null (Ed.)Objective: A reliable neural-machine interface offers the possibility of controlling advanced robotic hands with high dexterity. The objective of this study was to develop a decoding method to estimate flexion and extension forces of individual fingers concurrently. Methods: First, motor units (MUs) firing information were identified through surface electromyogram (EMG) decomposition, and the MUs were further categorized into different pools for the flexion and extension of individual fingers via a refinement procedure. MU firing rate at the populational level was calculated, and the individual finger forces were then estimated via a bivariate linear regression model (neural-drive method). Conventional EMG amplitude-based method was used as a comparison. Results: Our results showed that the neural-drive method had a significantly better performance (lower estimation error and higher correlation) compared with the conventional method. Conclusion: Our approach provides a reliable neural decoding method for dexterous finger movements. Significance: Further exploration of our method can potentially provide a robust neural-machine interface for intuitive control of robotic hands.more » « less
-
Abstract Sensitivity analysis is a popular feature selection approach employed to identify the important features in a dataset. In sensitivity analysis, each input feature is perturbed one-at-a-time and the response of the machine learning model is examined to determine the feature's rank. Note that the existing perturbation techniques may lead to inaccurate feature ranking due to their sensitivity to perturbation parameters. This study proposes a novel approach that involves the perturbation of input features using a complex-step. The implementation of complex-step perturbation in the framework of deep neural networks as a feature selection method is provided in this paper, and its efficacy in determining important features for real-world datasets is demonstrated. Furthermore, the filter-based feature selection methods are employed, and the results obtained from the proposed method are compared. While the results obtained for the classification task indicated that the proposed method outperformed other feature ranking methods, in the case of the regression task, it was found to perform more or less similar to that of other feature ranking methods.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.3389/fnins.2021.667907
