A multigrasp, bidirectional hand prosthesis delivers dynamic sensory feedback, allowing a user with a hand amputation to achieve fine grasping force control and realistic object sensing.
8 August 2018 in IEEE Journal of Biomedical and Health Informatics
Currently, most of the adopted myoelectric schemes for upper limb prostheses do not provide users with intuitive control. Higher accuracies have been reported using different classification algorithms but investigation on the reliability over time for these methods is very limited. In this study, we compared for the first time the longitudinal performance of selected state-of-the-art techniques for electromyography (EMG) based classification of hand motions. Experiments were conducted on ten able-bodied and six transradial amputees for seven continuous days. Linear discriminant analysis (LDA), artificial neural network (ANN), support vector machine (SVM), K-nearest neighbor (KNN), and decision trees (TREE) were compared. Comparative analysis showed that the ANN attained highest classification accuracy followed by LDA. Three-way repeated ANOVA test showed a significant difference (P < 0.001) between EMG types (surface, intramuscular, and combined), days (1–7), classifiers, and their interactions. Performance on the last day was significantly better (P < 0.05) than the first day for all classifiers and EMG types. Within-day, classification error (WCE) across all subject and days in ANN was: surface (9.12 ± 7.38%), intramuscular (11.86 ± 7.84%), and combined (6.11 ± 7.46%). The between-day analysis in a leave-one-day-out fashion showed that the ANN was the optimal classifier (surface (21.88 ± 4.14%), intramuscular (29.33 ± 2.58%), and combined (14.37 ± 3.10%). Results indicate that within day performances of classifiers may be similar but over time, it may lead to a substantially different outcome. Furthermore, training ANN on multiple days might allow capturing time-dependent variability in the EMG signals and thus minimizing the necessity for daily system recalibration.
18 April 2006 in IEEE Transactions on Biomedical Engineering
The mechanical behavior of an electrode during implantation into neural tissue can have a profound effect on the neural connections and signaling that takes place within the tissue. The objective of the present work was to investigate the in vivo implant mechanics of flexible, silicon-based ACREO microelectrode arrays recently developed by the VSAMUEL consortium (European Union, grant #IST-1999-10073). We have previously reported on both the electrical [1]-[3] and mechanical [4], [5] properties of the ACREO electrodes. In this paper, the tensile and compression forces were measured during a series of in vivo electrode insertions into the cerebral cortex of rats (7 acute experiments, 2-mm implant depth, 2-mm/s insertion velocity). We compared the ACREO silicon electrodes (4' opening angle, 1-8 shafts) to single-shaft tungsten electrodes (3' and 10' opening angles). The penetration force and dimpling increased with the cross-sectional area (statistical difference between the largest and the smallest electrode) and with the number of shafts (no statistical difference). We consistently observed tensile (drag) forces during the retraction phase, which indicates the brain tissue sticks to the electrode within a short time period. Treating the electrodes prior to insertion with silane (hydrophobic) or piranha (hydrophilic) significantly decreased the penetration force. In conclusion, our findings suggest that reusable electrodes for acute animal experiments must not only be strong enough to survive a maximal force that exceeded the penetration force, but must also be able to withstand high tension forces during retraction. Careful cleaning is not only important to avoid foreign body response, but can also reduce the stress applied to the electrode while penetrating the brain tissue
25 June 2013 in IEEE Transactions on Neural Systems and Rehabilitation Engineering
Neural prostheses are limited by the availability of peripheral neural electrodes to record the user's intention or provide sensory feedback through functional electrical stimulation. Our objective was to compare the ability of the novel “transverse intrafascicular multi-channel electrode” (TIME) and an earlier generation “thin-film longitudinal intrafascicular electrode” (tfLIFE) to selectively stimulate nerve fascicles and activate forelimb muscles in pigs. TIME was designed to access a larger subpopulation of fascicles than tfLIFE and should therefore be able to selectively activate a larger number of muscles. Electrodes were implanted in the median nerve, and sequential electric stimulation was applied to individual contacts. The compound muscle action potentials of seven muscles were recorded to quantify muscle recruitment. As expected, TIME was able to recruit more muscles with higher selectivity than tfLIFE (significant difference when comparing the performance of an entire electrode); a similar activation current was used (no significant difference). Histological analysis revealed that electrodes were located between fascicles, which influenced the selectivity and activation current level. In conclusion, TIME is a viable neural interface for selective activation of multiple fascicles in human-sized nerves that may assist to pave the way for future neuroprosthesis applications.
1 January 2012 in Journal of Rehabilitation Research and Development
Few studies are available in the literature on the sensations artificially created by dual-channel electrocutaneous stimulation. This study assessed the effect of a set of selected stimulation parameters on the sensations evoked by single- or dual-channel electrocutaneous stimulation. The investigated parameters included the stimulation site, the number of pulses, the number of stimulating channels (single- vs dual-channel), and the interleaved time between two channels. The modality, quality, location, and magnitude of the sensations were evaluated when the stimulations were applied on the forearm skin in 16 nondisabled subjects. Tactile perception was found to be induced more easily on the median and ulnar aspect than the dorsal and radial aspect of the forearm. Stimulation site significantly affected the magnitude of the sensation (p < 0.01). Dual-channel stimulation significantly increased the sensation magnitude (p < 0.05) only when the two electrodes were positioned closely. Moreover, a higher number of pulses evoked a movement perception more frequently and the interleaved time showed no significant effect on the magnitude of the sensation. The findings are expected to be useful for sensory substitution and augmentation applications. The results may also help improve users' acceptance of hand prostheses.
19 October 2011 in Annals of Biomedical Engineering