Jose Shobha, George S Thomas, Subathra M S P, Handiru Vikram Shenoy, Jeevanandam Poornaselvan Kittu, Amato Umberto, Suviseshamuthu Easter Selvan
School of Engineering and TechnologyKarunya Institute of Technology and Sciences Coimbatore 641-114 India.
Center for Mobility and Rehabilitation Engineering ResearchKessler Foundation West Orange NJ 07052 USA.
IEEE Open J Eng Med Biol. 2020 Aug 17;1:235-242. doi: 10.1109/OJEMB.2020.3017130. eCollection 2020.
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. 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. 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 = [Formula: see text], sensitivity (normal) = [Formula: see text], sensitivity (myopathy) = [Formula: see text], sensitivity (neuropathy) = [Formula: see text], specificity (normal) = [Formula: see text], specificity (myopathy) = [Formula: see text], and specificity (neuropathy) = [Formula: see text]-surpassing the existing approaches. 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.
本文介绍了一种精确的自动诊断系统的设计与验证，该系统可将肌内肌电图（iEMG）信号分类为健康、肌病或神经病类别，以辅助神经肌肉疾病的诊断。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（FDs）。将LWT子带系数的FDs与从每个下采样信号导出的一维局部二值模式进行融合。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，将类别标签序列输入到博耶 - 摩尔多数投票（BMMV）算法中，该算法为每个iEMG信号分配一个类别。使用属于三个类别的250个iEMG信号对MLPNN - BMMV分类器进行了实验。与现有方法相比，验证了该分类器的性能。MLPNN - BMMV在使用10折交叉验证时产生了令人印象深刻的性能指标（%）——准确率 = [公式：见原文]，灵敏度（正常） = [公式：见原文]，灵敏度（肌病） = [公式：见原文]，灵敏度（神经病） = [公式：见原文]，特异性（正常） = [公式：见原文]，特异性（肌病） = [公式：见原文]以及特异性（神经病） = [公式：见原文]——超过了现有方法。未来的研究方向是使用多样化的iEMG数据集验证分类器性能，这将导致设计出一种用于神经肌肉疾病诊断的经济实惠的实时专家系统。
