This thesis considers Neuro-prosthetic systems using Functional Electrical Stimulation (FES) to restore or support motor functions in neurologically impaired patients. An often encountered difficulty is the precise control of limb motion by adjusting the stimulation intensity. The musculoskeletal systems involving muscle mechanics, limb motion dynamics, and muscle recruitment due to FES and possibly volitional activity are highly complex and non-linear. Hence, the stimulation effect is hard to model and predict. An intensive effort during a calibration phase would be required to adapt models describing such systems to the full extent, which is unfeasible in clinical environments or at home. Further, many parameters vary under the rapid progression of muscle fatigue with FES. Hence, most practically relevant devices focus on simple open-loop strategies that trigger pre-defined stimulation patterns on different events caused by the user. This thesis presents new methods to improve the applicability and movement precision of closed-loop FES systems and applies them to practically relevant applications to support or to restore functional arm reaching movements in entirely or partially paralyzed patients. The compensation of muscle fatigue and the linearization of the highly non-linear process of muscle recruitment is addressed in this thesis. Electromyography (EMG)-measurements can give insight into the muscular recruitment process. Therefore, a feedback of the stimulation-evoked EMG (eEMG) enforcing the desired recruitment level is presented. Because of this enforcement, a more predictable behavior is obtained such that muscles are much easier to model and control using this method as an underlying feedback loop. The positive effects on the predictability are shown in a study involving five healthy subjects. Further, the applicability of this approach was successfully tested in two stroke patients. In partially paralyzed patients it is often beneficial to take account the residual volitional activity to, e.g., realize a user-control of the neuroprosthesis. Devices exist that control FES based on a measurement the volitional EMG (vEMG). The low signal to noise ratio, however, typically leads to many restrictions in control performance. Hence, a different approach based on angular measurements is presented: An arm weight relief is realized by a FES-activation proportional to the elevation angle. This FES-activation reduces the required volitional effort significantly as demonstrated in five healthy subjects. In two acute stroke patients, arm function could be partially and fully restored, respectively, as long as the controller was active. For patients who have a complete paralysis of the upper extremity, the functional restoration of movements is difficult due to the numerous degrees of freedom (DoF). Hence, a system using a combination of a passive, light-weight exoskeleton with FES is proposed, wherein each DoF can be locked by brakes to reduce muscle fatigue during holding-postures and to guide movements. FES is feedback controlled, and the movement precision was evaluated in five healthy subjects showing a sufficient performance for performing simple reaching tasks. Further considered aspects include the improvement of motor precision by artificially introducing co-activations in antagonistic muscle pairs and the potential reduction of muscle fatigue by varying the stimulation frequency as desired for the contrary objectives "motor precision" and "muscle fatigue." For the former aspect, a non-linear system inversion based approach is developed to realize desired joint torques and co-activation levels and, further, applied to joint angle control yielding good results in a healthy subject. For the latter, a time-discretization method for linear time-invariant systems is presented that also considers variable sampling rates. Further, a system linearization approach is suggested allowing to apply standard discrete-time controllers for variable sampling rate systems.


    Access

    Download

    Check availability in my library


    Export, share and cite



    Title :

    Feedback-controlled functional electrical stimulation to restore upper extremity functions.
    Regelung der muskulären Rekrutierung und automatische Anpassung für die Vereinfachung des klinischen Einsatzes von neuartigen Neuroprothesen


    Subtitle :

    improving the clinical feasibility of elaborated neuroprosthetic devices by automatic adaptation and feedback of muscle-recruiment


    Additional title:

    Geregelte funktionelle Elektrostimulation zur Wiederherstellung motorischer Fähigkeiten der oberen Extremität


    Contributors:

    Publication date :

    2019



    Type of media :

    Miscellaneous


    Type of material :

    Electronic Resource


    Language :

    English



    Classification :

    DDC:    629



    Automatische Regelung von Kesselanlagen

    Gerasimov, S. G. | TIBKAT | 1952


    Automatische Regelung des Reifenluftdrucks

    von Hohenthal,M.Y. / Steyr-Daimler-Puch,AT | Automotive engineering | 2005


    Automatische Luftdruck Regelung ALR

    Skoff,G. / Steyr-Daimler-Puch,AT | Automotive engineering | 2005


    Regelung des Drehfreiheitsgrades eines neuartigen Gesamtfahrzeugpruefstands

    Fietzek,R. / Meier,T. / Rinderknecht,S. et al. | Automotive engineering | 2011


    Regelung des Drehfreiheitsgrades eines neuartigen Gesamtfahrzeugprüfstands

    Fietzek, Rafael / Meier, Torben / Rinderknecht, Stephan | Tema Archive | 2011