The inability to replace human muscle in surgical practice is a significant challenge. An artificial muscle controlled by the nervous system is considered a potential solution for this. We defined it as neuromuscular prosthesis. Muscle loss and dysfunction related to musculoskeletal oncological impairments, neuromuscular diseases, trauma or spinal cord injuries can be treated through artificial muscle implantation. At present, the use of dielectric elastomer actuators working as capacitors appears a promising option. Acrylic or silicone elastomers with carbon nanotubes functioning as the electrode achieve mechanical performances similar to human muscle in vitro. However, mechanical, electrical, and biological issues have prevented clinical application to date. In this study, materials and mechatronic solutions are presented which can tackle current clinical problems associated with implanting an artificial muscle controlled by the nervous system. Progress depends on the improvement of the actuation properties of the elastomer, seamless or wireless integration between the nervous system and the artificial muscle, and on reducing the foreign body response. It is believed that by combining the mechanical, electrical, and biological solutions proposed here, an artificial neuromuscular prosthesis may be a reality in surgical practice in the near future.

Introduction and Aims: Following injuries leading to loss of hand function, rehabilitative exercises are often critical for recovery. This project utilises artificial muscles to create a glove that can provide specific rehabilitation options to the hand, as well as provide grasp and release functions to those unable to move their hand.

Method: Artificial muscles are composed of electrically activated polymers or shape memory alloys (SMA) that shorten under current, in a way similar to muscles. In this project, three glove prototypes were developed and tested initially with SMA Nitinol wires. Each wire was placed such that its activation caused a movement of an individual joint in a hand model. Wire activation was controlled by a pocket PC, thereby enabling portability. Artificial proprioception was provided for the joints, using force/position transducers. A software interface in the controller was used to program desired passive movement.

Results: Each joint could be independently moved to a desired angle. The force and position transducers were tested against calibrated inputs and were found to be reliable. These were used as feedback for the control of motion and force. The control software allowed the therapist to program desired force, speed and range of motion for each of the 15 finger/thumb joints in the hand model. As such, continuous passive joint movements can be administered by the glove for hours at a time through a portable glove and controller device. Another experiment verified the ability of this unique rehabilitation glove to provide a light hand grasp with controllable holding force. This function can be triggered using an external switch such as a shoulder stick. Other functions implemented include the provision of dynamic splinting to the finger joints using the system software and the ability to log the performance of the hand through time. The latter function would be valuable for monitoring the effectiveness of the therapy, as well as for continuously adjusting the exercises to increase performance.

Conclusion: The individual tailoring of therapeutic programs for the hand joints will create a safer and more effective therapy than current devices available on the market. As well as being portable, the provision of hand grasp and release with the glove provides an additional functional benefit for those with paralysed hands.