Determining the optimal window length for pattern recognition-based myoelectric control: balancing the competing effects of classification error and controller delay.
ABSTRACT Pattern recognition-based control of myoelectric prostheses has shown great promise in research environments, but has not been optimized for use in a clinical setting. To explore the relationship between classification error, controller delay, and real-time controllability, 13 able-bodied subjects were trained to operate a virtual upper-limb prosthesis using pattern recognition of electromyogram (EMG) signals. Classification error and controller delay were varied by training different classifiers with a variety of analysis window lengths ranging from 50 to 550 ms and either two or four EMG input channels. Offline analysis showed that classification error decreased with longer window lengths (p < 0.01 ). Real-time controllability was evaluated with the target achievement control (TAC) test, which prompted users to maneuver the virtual prosthesis into various target postures. The results indicated that user performance improved with lower classification error (p < 0.01 ) and was reduced with longer controller delay (p < 0.01 ), as determined by the window length. Therefore, both of these effects should be considered when choosing a window length; it may be beneficial to increase the window length if this results in a reduced classification error, despite the corresponding increase in controller delay. For the system employed in this study, the optimal window length was found to be between 150 and 250 ms, which is within acceptable controller delays for conventional multistate amplitude controllers.
- SourceAvailable from: Henning Müller[Show abstract] [Hide abstract]
ABSTRACT: There has been increasing interest in applying learning algorithms to improve the dexterity of myoelectric prostheses. In this work, we present a large-scale benchmark evaluation on the second iteration of the publicly released NinaPro database, which contains surface electromyography data for 6 DOF force activations as well as for 40 discrete hand movements. The evaluation involves a modern kernel method and compares performance of three feature representations and three kernel functions. Both the force regression and movement classification problems can be learned successfully when using a non-linear kernel function, while the exp-2 kernel outperforms the more popular Radial Basis Function kernel in all cases. Furthermore, combining surface electromyography and accelerometry in a multimodal classifier results in significant increases in accuracy as compared to when either modality is used individually. Since window-based classification accuracy should not be considered in isolation to estimate prosthetic controllability, we also provide results in terms of classification mistakes and prediction delay. To this extent, we propose the Movement Error Rate as an alternative to the standard window-based accuracy. This error rate is insensitive to prediction delays and it allows therefore to quantify mistakes and delays as independent performance characteristics. This type of analysis confirms that the inclusion of accelerometry is superior, as it results in fewer mistakes while at the same time reducing prediction delay.IEEE transactions on neural systems and rehabilitation engineering: a publication of the IEEE Engineering in Medicine and Biology Society 01/2014; · 2.42 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Pattern recognition (PR) based strategies for the control of myoelectric upper limb prostheses are generally evaluated through offline classification accuracy, which is an admittedly useful metric, but insufficient to discuss functional performance in real time. Existing functional tests are extensive to set up and most fail to provide a challenging, objective framework to assess the strategy performance in real time.Journal of NeuroEngineering and Rehabilitation 05/2014; 11(1):91. · 2.57 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Fine-wire intramuscular electrodes were used to obtain EMG signals from six extrinsic hand muscles associated with the thumb, index, and middle fingers. Subjects' EMG activity was used to control a virtual three-DOF hand as they conformed the hand to a sequence of hand postures testing two controllers: direct EMG control and pattern recognition control. Subjects tested two conditions using each controller: starting the hand from a pre-defined neutral posture before each new posture and starting the hand from the previous posture in the sequence. Subjects demonstrated their ability to simultaneously, yet individually, move all three DOFs during the direct EMG control trials, however results showed subjects did not often utilize this feature. Performance metrics such as failure rate and completion time showed no significant difference between the two controllers.IEEE transactions on bio-medical engineering 07/2014; · 2.15 Impact Factor