London, United Kingdom 2013 Poster Session Blue Cerebrovasc Dis 2013; 35 (suppl 3)1-854 785 831 Rehabilitation and reorganisation after stroke An EMG-controlled upper limb neuroprosthesis to improve motor recovery after stroke E. Ambrosini1, S. Ferrante2, G. Ferrigno3, F. Molteni4, A. Pedrocchi5 Politecnico di Milano, Dipartimento di Elettronica, Informazione e Bioingegneria, NearLab, Milano, ITALY1, Politecnico di Milano, Dipartimento di Elettronica, Informazione e Bioingegne-ria, NearLab, Milano, ITALY2, Politecnico di Milano, Dipartimento di Elettronica, Informazione e Bioingegneria, NearLab, Milano, ITALY3, Valduce Hospital, Villa Beretta, Rehabilitation Center, Costa Masnaga, Lecco, ITALY4, Politecnico di Milano, Dipartimento di Elettronica, Informazione e Bioingegneria, NearLab, Milano, ITALY5 Background Neuro Muscular Electrical Stimulation (NMES) has been used in literature to restore motor func-tions after stroke (Sheffler and Chae, 2007). When NMES is associated with a voluntary attempt to move the limb the improvement is further enhanced (Rushton, 2003; de Kroon et al., 2005). The work aims at developing an EMG-controlled neuroprosthesis for the upper limb and testing its feasi-bility on post-stroke patients. Methods The experimental setup (Fig. 1) consisted of a stimulator (RehaStim, Hasomed), an EMG amplifier (Porti 32, TMS International), a passive exoskeleton for weight relief (ArmeoSpring, Hocoma), and a PC. The EMG electrodes were placed on the biceps between the stimulation electrodes. The EMG signal was acquired at 2048 Hz. The current amplitude was set on each subject, the pulse width (PW) varied based on the volitional EMG, and the stimulation frequency was fixed at 25 Hz. To extract the volitional EMG from hybrid muscle activation, an adaptive filter (Sennels et al, 1997) was applied on the measured EMG (EMGm). For each stimulation period, the mean value of the rectified volitional EMG (eEMGv) was used to define the PW value. A double-threshold controller was designed (Fig. 1): when eEMGv overcame the activation threshold the PW value was increased to the maximum value; when eEMGv became lower than the deactivation threshold, the PW value was decreased to zero. During training, subjects were asked to perform elbow flex-extension movements with NMES and gravity support, while tracking a target angle displayed on a screen. Results and Conclusion Fig. 2 reports an example of a patient performing the tracking trial: panel (a) depicts the tracking performance, while panel (b) shows the ability of the subject to activate and de-activate the stimula-tion when desired. Our study offers a method to combine NMES with the attempted voluntary movement and paves the way for providing a better understanding of the benefits of this combination on stroke patients.
Karger_ESC London_2013
To see the actual publication please follow the link above