Goal: This article presents the design and validation of an accurate automatic diagnostic system to classify intramuscular EMG (iEMG) signals into healthy, myopathy, or neuropathy categories to aid the diagnosis of neuromuscular diseases. Methods: First, an iEMG signal is decimated to produce a set of “disjoint” downsampled signals, which are decomposed by the lifting wavelet transform (LWT). The Higuchi's fractal dimensions (FDs) of LWT coefficients in the subbands are computed. The FDs of LWT subband coefficients are fused with one-dimensional local binary pattern derived from each downsampled signal. Next, a multilayer perceptron neural network (MLPNN) determines the class labels of downsampled signals. Finally, the sequence of class labels is fed to the Boyer-Moore majority vote (BMMV) algorithm, which assigns a class to every iEMG signal. Results: The MLPNN-BMMV classifier was experimented with 250 iEMG signals belonging to three categories. The performance of the classifier was validated in comparison with state-of-the-art approaches. The MLPNN-BMMV has resulted in impressive performance measures (%) using a 10-fold cross-validation—accuracy = $99.87\pm 0.25$ , sensitivity (normal) = $99.97\pm 0.13$ , sensitivity (myopathy) = $99.68\pm 0.95$ , sensitivity (neuropathy) = $99.76\pm 0.66$ , specificity (normal) = $99.72\pm 0.61$ , specificity (myopathy) = $99.98\pm 0.10$ , and specificity (neuropathy) = $99.96\pm 0.14$ —surpassing the existing approaches. Conclusions: A future research direction is to validate the classifier performance with diverse iEMG datasets, which would lead to the design of an affordable real-time expert system for neuromuscular disorder diagnosis.

中文翻译：

---

肌内 EMG 信号的稳健分类以帮助诊断神经肌肉疾病

目标：本文介绍了精确自动诊断系统的设计和验证，用于将肌内 EMG (iEMG) 信号分类为健康、肌病或神经病类别，以帮助诊断神经肌肉疾病。方法：首先，对 iEMG 信号进行抽取以产生一组“不相交”的下采样信号，这些信号通过提升小波变换 (LWT) 进行分解。计算子带中 LWT 系数的 Higuchi 分形维数 (FD)。LWT 子带系数的 FD 与从每个下采样信号导出的一维局部二进制模式融合。接下来，多层感知器神经网络 (MLPNN) 确定下采样信号的类别标签。最后，类标签序列被输入到 Boyer-Moore 多数投票 (BMMV) 算法，该算法为每个 iEMG 信号分配一个类。结果：MLPNN-BMMV 分类器对属于三个类别的 250 个 iEMG 信号进行了实验。与最先进的方法相比，分类器的性能得到了验证。MLPNN-BMMV 使用 10 倍交叉验证——准确度 =$99.87\pm 0.25$ ，灵敏度（正常）=$99.97\pm 0.13$ ，敏感性（肌病）=$99.68\pm 0.95$ ，敏感性（神经病）=$99.76\pm 0.66$ ，特异性（正常）=$99.72\pm 0.61$ ，特异性（肌病）=$99.98\pm 0.10$ ，和特异性（神经病）=$99.96\pm 0.14$ ——超越现有的方法。结论：未来的研究方向是使用不同的 iEMG 数据集验证分类器性能，这将导致设计用于神经肌肉疾病诊断的负担得起的实时专家系统。