Department of Clinical Neurophysiology, Rigshospitalet and University of Copenhagen, DK–1017 Copenhagen, Denmark.
IEEE Trans Biomed Eng. 2011 Oct;58(10):2707-18. doi: 10.1109/TBME.2010.2064773. Epub 2010 Aug 9.
We have developed an electromyography (EMG) decomposition system called EMGTools that can extract the constituent MUAPs and firing patterns (FPs) for quantitative analysis from the EMG signal recorded at slight effort for clinical evaluation. The aim was to implement a robust system able to handle the challenges and variations in clinically recorded signals. The system extracts MUAPs recorded by concentric needle electrodes and resolves superimposed MUAPs to produce FPs. Thus, critical fixed thresholds/parameters are avoided and replaced with adaptive solutions. The decomposition algorithm consists of three stages: segmentation, clustering, and resolution of compound segments. The results are validated using three different methods, comparing mean MUAP duration with previous methods, comparing dual channel recordings, and assessing the residual signal after decomposition. The advantages and limitations of the system are discussed.
我们开发了一种肌电图（EMG）分解系统，称为 EMGTools，它可以从临床评估中记录的轻微用力的 EMG 信号中提取组成 MUAP 和放电模式（FP）进行定量分析。目的是实现一个稳健的系统，能够处理临床记录信号中的挑战和变化。该系统提取同心针电极记录的 MUAP，并解析叠加的 MUAP 以产生 FP。因此，避免了关键的固定阈值/参数，并采用了自适应解决方案。分解算法由三个阶段组成：分段、聚类和复合段的解析。使用三种不同的方法验证结果，包括与以前的方法比较 MUAP 持续时间的平均值、比较双通道记录和评估分解后的剩余信号。讨论了系统的优点和局限性。
