Article

Determining the optimal window length for pattern recognition-based myoelectric control: balancing the competing effects of classification error and controller delay.

Feinberg School of Medicine, Northwestern University, Chicago, IL 60611 USA.
IEEE transactions on neural systems and rehabilitation engineering: a publication of the IEEE Engineering in Medicine and Biology Society (Impact Factor: 2.42). 12/2010; 19(2):186-92. DOI: 10.1109/TNSRE.2010.2100828
Source: PubMed

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.

0 Bookmarks
 · 
77 Views
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Myoelectric control is filled with potential to significantly change human–robot interaction due to the ability to non-invasively measure human motion intent. However, current control schemes have struggled to achieve the robust performance that is necessary for use in commercial applications. As demands in myoelectric control trend toward simultaneous multifunctional control, multi-muscle coordinations, or synergies, play larger roles in the success of the control scheme. Detecting and refining patterns in muscle activations robust to the high variance and transient changes associated with surface electromyography is essential for efficient, user-friendly control. This article reviews the role of muscle synergies in myoelectric control schemes by dissecting each component of the scheme with respect to associated challenges for achieving robust simultaneous control of myoelectric interfaces. Electromyography recording details, signal feature extraction, pattern recognition and motor learning based control schemes are considered, and future directions are proposed as steps toward fulfilling the potential of myoelectric control in clinically and commercially viable applications.
    Journal of Neural Engineering 09/2014; 11(5):051001. · 3.42 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The potential for pattern recognition to improve powered prosthesis control has been discussed for many years. One remaining barrier to at-home use of these techniques is that practical methods of user prompting during system training are lacking. Most research and development of pattern recognition systems for prosthesis control has relied on on-screen cues to prompt the prosthesis wearer during signal collection; therefore most systems require connection to a computer or external device. We have developed a method called Prosthesis-Guided Training (PGT) to address this issue. In PGT, the prosthesis itself moves through a pre-programmed sequence of motions to prompt the wearer to elicit the appropriate muscle contractions. PGT requires no extra hardware and allows wearers to retrain, refresh, or recalibrate the controller in many locations and situations. Training via PGT is self-initiated and requires only about 1 minute of the wearer's time. Furthermore, PGT provides a practical mechanism for overcoming malfunctioning or changing inputs, addresses differences in routine donning, and results in acquisition of myoelectric signals representative of those elicited during functional use. Qualitative and quantitative data acquired to investigate the efficacy of PGT suggest that it is an intuitive, effective, and clinically viable method of training pattern recognition–controlled prostheses.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: A remaining barrier to the clinical accessibility of pattern recognition systems is the lack of practical methods to acquire the myoelectric signals required to train the system. Many current methods involve screen-guided training (SGT), where wearers connected to an external computer perform muscle contractions synchronized with a sequence of visual cues. The system complexity prevents easy retraining when signal conditions change. We have developed a method called prosthesis-guided training (PGT), where the prosthesis itself provides the cues by moving through a sequence of preprogrammed motions; screen prompting and external connections are eliminated. Five prosthesis wearers performed a repetitive clothespin placement task using pattern recognition control. Wearers demonstrated similar baseline functionality between systems trained with PGT (10 ± 4 clothespins) and SGT (12 ± 7 clothespins) (p = 0.56). To investigate the efficacy of PGT retraining, real-world issues (e.g. broken wires, external noise) were simulated to accelerate control degradation. Sessions ended when wearers indicated loss of functional control. On average, wearers maintained function through two malfunctioning inputs, placing 48 ± 17 clothespins in 31.6 ± 16.2 minutes when allowed to retrain using PGT. These results suggest that PGT acquires adequate training data and may enable longer-lasting functional use, potentially increasing prosthesis wear time and reducing device rejection.