Effect of upper-limb positions on motion pattern recognition using electromyography

Abstract

Previous studies of electromyographic (EMG) pattern recognition for neural prosthesis control mainly focused on the estimation of offline classification accuracy. Factors that may affect the performance in operating prosthesis in practice were rarely considered. In the preliminary study we investigated effects of the variation of limb positions on classification performance. Eight channels of myoelectric signals and a LDA classifier were used to identify seven classes of forearm movements in five transradial amputees. Our pilot results showed that the classification error of inter-position was about 4 times more than that of single position across the five transradial amputees (p<0.02), indicating the significant effect of limb position variation on classification performance. We further attempted to reduce the limb position effect by training the classifier with EMG from multiple EMG recording positions, and the classification performance was compared with that of single position training. It was found multiple positions training method could decrease the classification error as more positions involved, which reached its minimum value 11.25% when all five positions were included.

High-density force myography: A possible alternative for upper-limb prosthetic control

“…Scheme et al demonstrated that training the classifier in multiple positions reduces this degradation [14]. Chen et al performed a similar study using data from participants with transradial amputation and supported the notion that training the classifier in multiple positions reduces the position effect [17]. In another study, Scheme et al further showed that changing the limb position during both static and dynamic ADLs negatively affects myoelectric pattern recognition [15].…”

“…The results of applying HD-FMG for classification were compared with those of standard pattern recognition-based EMG control methods reported in the literature [13][14][15][16][17][18]. As shown in Figure 7, the mean overall error was 0.33 percent using pressure maps, while errors between 2.2 and 11.3 percent have been commonly reported using EMG signals [13][14][15][16][17][18].…”

“…To study the performance of the HD-FMG technique, we compared the classification results with those of standard pattern recognition-based EMG control methods that have been reported in the literature [13][14][15][16][17][18], and followed a similar experimental protocol as that used by Radmand et al [28]. Subjects were prompted to elicit contractions corresponding to eight classes of motion including wrist flexion/extension, wrist supination/pronation, pinch grip, power grip, hand open, and no movement.…”

“…HD-FMG uses standard pattern recognition algorithms to discriminate the patterns and determine the user's intent. In this study, we compared the performance of HD-FMG with that of previously reported pattern recognition-based EMG control approaches using an experimental protocol similar to those reported in the literature [13][14][15][16][17][18].…”

“…Low classification errors, in the range of 2.2 to 11.3 percent, have been reported for varying numbers (6 to 10) of wrist and hand movements using EMG pattern recognition techniques, such as linear discriminant analysis (LDA), artificial neural networks, and support vector machines (SVMs) [13][14][15][16][17][18].…”