Embedded Controller for Artificial Limbs

Licentiate thesis, 2017

Promising developments are currently ongoing worldwide in the field of neuroprosthetics and artificial limb control. It is now possible to chronically connect a robotic limb to bone, nerves and muscles of a human being, and use the signals sourced from these connections to enable movements in the artificial limb. It is also possible to surgically redirect a nerve, deprived from its original target muscle due to amputation, to a new target in order to restore the original motor functionality. Intelligent signal processing algorithms can now utilize the bioelectric signals gathered from remaining muscles on the stump to decode the motor intention of the amputee, providing an intuitive control interface. Unfortunately for patients, clinical implementations still lag behind the advancements of research, and the conventional solutions for amputees remained basically unchanged since decades. More efforts are therefore needed from researchers to close the gap between scientific publications and hospital practices. The ultimate focus of this thesis is set on the intuitive control of a prosthetic upper limb. It was developed an embedded system capable of prosthetic control via the processing of bioelectric signals and pattern recognition algorithms. It includes a neurostimulator to provide direct neural feedback modulated by sensory information from artificial sensors. The system was designed towards clinical implementation and its functionality was proven by amputee subjects in daily life. It also constitutes a research platform to monitor prosthesis usage and training, machine learning based control algorithms, and neural stimulation paradigms.