Neuroprosthetics and Solutions for Restoring Sensorimotor Functions
- Saturday, 01 May 2010
A new class of prosthetic devices looks, feels, and functions like natural limbs.
Advances in body armor and life-saving technology have increased survival rates of severely injured military personnel. Unfortunately, the survivors of improvised explosive devices are often left with amputations and/or spinal cord injuries. The increase in amputations and paralysis among military personnel requires significant advances in prosthetics and functional electrical stimulation (FES) systems so that the soldiers can return to the field or to productive civilian lives.
A new class of prosthetic devices has been developed that mimics more closely the full range of sensory and motor capabilities of natural limbs. By providing a communication link between the prosthesis and the user’s nervous system, the prosthetic limb is integrated as a natural component of the user’s sensorimotor apparatus.
The first project deals directly with the issue of providing somatosensory input to soldiers with amputation or paralysis. The overall goal is to restore natural sensations of limb posture and movement through multichannel microstimulation of the normal afferent pathways involved in proprioception. Two objectives must be met to achieve this overall goal. First, an appropriate location in the somatosensory nervous system must be identified for implanting microelectrode arrays to stimulate afferent neurons. The ideal site for microstimulation is one in which the neurons for each body location (somatotopy) and afferent modality (e.g. muscle spindles, tendon organs) are colocated, allowing a single electrode to activate multiple neurons of a similar class. The second objective is to quantify the amount of information that can be transmitted by multichannel afferent microstimulation.
The second project deals with improving the chronic stability of the neural interface and testing novel polymer surface modification methods for improving the long-term reliability of the implanted microelectrodes. Although neural control is the ultimate goal, there is useful control information available in the muscles of the residual limb. The implanted neural interface must remain stable throughout the lifespan of the user, but immune and inflammatory reactions at the implant site are known to degrade the performance of implanted microelectrodes. Since tissue reactions vary in different parts of the nervous system, the first objective is to compare the responses in the DRG, dorsal root nerve, and spinal cord. The second objective is to test whether surface coating, with agents that encourage specific neuronal survival and growth and reduce inflammation, will be effective in improving the biocompatibility.
Two types of coatings are being developed for the electrodes: one for surface immobilization of neural adhesion molecules, and the other for the controlled release of anti-inflammatory drugs. A surface immobilization method was developed that can coat the surface of parylene C with neural adhesion molecule L1, which has shown to be potent in promoting neuronal health and growth while inhibiting glial cells.
The third project uses virtual reality to place patients in an environment with a simulated neuroprosthesis. In this environment, one can discover the degree of remaining electromyographic (EMG) signal content and begin to train patients to control their neuroprosthetic. The virtual environment will allow amputees to: 1) test simulated neuroprosthetics and control algorithms, and 2) practice using the neuroprosthetic in a virtual training environment.
The critical hardware and software components for the virtual reality training system are:
- A three-dimensional real-time motion tracking system with four cameras and a base station running a real-time Linux operating system. All the cameras and the base station communicate with each other through a dedicated local network. All cameras are wall-mounted in the human subject testing room.
- A 16-channel EMG recording system equipped with dry EMG surface electrodes, which reduce setup time and skin irritation caused by conductive gels.
- A 64-channel brain-computer interface system that is FDA-approved for recording physiological signals from humans, including EMG, EOG, EEG, and ECoG.
- A 4-channel oscilloscope for data visualization and system testing.
- An MEG and fMRI-compatible 2-axis joystick with USB interface. It has no metal or electronic components, and it uses fiber optics to transmit the movement signal from the MEG recording room to an outside base station.
- BCI2000 software, an open-source software package based on C++. It is a general-purpose software package for neuroprosthetics and brain-computer interface research. It has three modules: signal source, signal processing, and application.
- Virtual Reality Toolbox that provides an interface for users to create and manipulate virtual objects through either a MATLAB or Simulink interface in real time.