|
|
|
|||||||||||
|
Wrist
versus Elbow/Shoulder :
How do MI neurons represent movement about these interconnected yet
discrete joints? Growing evidence suggests that these forelimb
components are controlled as an integrated unit; the classical view of
discrete hand / forearm / shoulder areas has proven less appropriate.
Providing amputees with intricate control of prosthetic arms and
hands will require a much better understanding of how the brain parcels
out representation of integrated movement plans within and across cortical
areas. Kinematics
and Dynamics
: What do action potential patterns in MI motor cortical neuron
really encode? There are many correlates with hand position (and other
kinematics) as well as forces involved in movement (dynamics).
Understanding in detail how the population of neurons works together to
represent and execute a movement will help us build better decoding
algorithms for the control of prosthetic devices. Chip
integration
: The BrainGate clinical trial has reported cursor control by an
individual with tetraplegia with an intracortical microelectrode array
implant more than three years after implant. At Brown, our
team is working closely with Dr. Nurmikko's engineering group to
develop the next generation of brain-implantable microelectronics,
advancing state-of-the-art miniaturization of the electronics
required to capture and decode the neural signature of movement
intentions. The end goal of this research, currently in in-vivo testing,
is a fully-implantable wireless recording device that makes high-fidelity
neural signals (30kS/s) available externally for neural prosthesis
applications. Neural
Decoding
: The BrainGate team works closely with the Brown Computer
Science department, particularly Phil Kim, to understand how
various theoretical computational and statistical frameworks might
best be applied to decoding neural activity. We currently favor Bayesian
methods because they provide a formal probabilistic framework in which to
evaluate the power of our decoding strategies. Together we strive
to communicate their expertise into our neuroprosthetic
application. Arm Models : To help translate our laboratory's work into neuroprosthetic products, we are beginning to pursue models of the arm and wrist musculature. Many outstanding laboratories and clinics have worked on this problem and we hope to benefit from their knowledge as we bring this technology into our realm. With the capability we will be able to test the appropriateness of our decoding algorithms and tune our neural decoding for improved limb control. |
||||||||||||