IEEE Trans Biomed Eng. 2019 Nov;66(11):3192-3203. doi: 10.1109/TBME.2019.2901882. Epub 2019 Feb 26.
The performance of traditional approaches to decoding movement intent from electromyograms (EMGs) and other biological signals commonly degrade over time. Furthermore, conventional algorithms for training neural network based decoders may not perform well outside the domain of the state transitions observed during training. The work presented in this paper mitigates both these problems, resulting in an approach that has the potential to substantially improve the quality of life of the people with limb loss.
This paper presents and evaluates the performance of four decoding methods for volitional movement intent from intramuscular EMG signals.
The decoders are trained using the dataset aggregation (DAgger) algorithm, in which the training dataset is augmented during each training iteration based on the decoded estimates from previous iterations. Four competing decoding methods, namely polynomial Kalman filters (KFs), multilayer perceptron (MLP) networks, convolutional neural networks (CNN), and long short-term memory (LSTM) networks, were developed. The performances of the four decoding methods were evaluated using EMG datasets recorded from two human volunteers with transradial amputation. Short-term analyses, in which the training and cross-validation data came from the same dataset, and long-term analyses, in which the training and testing were done in different datasets, were performed.
Short-term analyses of the decoders demonstrated that CNN and MLP decoders performed significantly better than KF and LSTM decoders, showing an improvement of up to 60% in the normalized mean-square decoding error in cross-validation tests. Long-term analyses indicated that the CNN, MLP, and LSTM decoders performed significantly better than a KF-based decoder at most analyzed cases of temporal separations (0-150 days) between the acquisition of the training and testing datasets.
The short-term and long-term performances of MLP- and CNN-based decoders trained with DAgger demonstrated their potential to provide more accurate and naturalistic control of prosthetic hands than alternate approaches.
传统的从肌电图（EMG）和其他生物信号中解码运动意图的方法的性能通常会随时间退化。此外，基于神经网络的解码器的传统训练算法在训练期间观察到的状态转换之外的域中可能表现不佳。本文提出并评估了从肌内 EMG 信号中解码意愿运动的四种解码方法的性能。该解码器使用数据集聚合（DAgger）算法进行训练，其中在每个训练迭代过程中，根据前一个迭代的解码估计值来增强训练数据集。开发了四种竞争解码方法，即多项式卡尔曼滤波器（KF）、多层感知器（MLP）网络、卷积神经网络（CNN）和长短时记忆（LSTM）网络。使用从两名桡骨截肢的人类志愿者记录的 EMG 数据集评估了这四种解码方法的性能。进行了短期分析，其中训练和交叉验证数据来自同一数据集，以及长期分析，其中训练和测试在不同的数据集上进行。短期分析表明，CNN 和 MLP 解码器的性能明显优于 KF 和 LSTM 解码器，在交叉验证测试中，归一化均方解码误差的改善高达 60%。长期分析表明，在训练和测试数据集采集之间的时间分离（0-150 天）的大多数分析情况下，CNN、MLP 和 LSTM 解码器的性能明显优于基于 KF 的解码器。结论：基于 DAgger 训练的 MLP 和 CNN 解码器的短期和长期性能表明，它们有可能提供比替代方法更准确和自然的假肢手控制。
