Abstract.
Recent work in macaques has shown that different areas of posterior
parietal cortex are specialized for planning hand and eye movements
(1; 2), and that it is possible to use recordings from these areas to
predict the direction of the planned movement (3). Preliminary studies
from our group have taken the first step toward identifying the human
homologue of the macaque parietal reach region (PRR), which is responsible
for planning hand movements (4). However, it is still unknown if neural
activity in human PRR exhibits the same spectral characteristics as
that in the macaque. To address this question we are working with human
participants who have chronically implanted electrodes placed on the
surface of cortex and within deep brain regions, often in partial cortex.
Recording taken from these participants while they execute delayed reaches
allow us to acquire high signal-to-noise intracranial EEG (iEEG) activity
from cortical areas during motor planning. Analysis of this neural activity
is aimed at determining which properties of the signal can be used to
decode and predict planned movement.
Additionally, in order for human PRR to serve as a substrate for neuroprosthetic
control signals it must be resistant to pathological reorganization
after cortico-spinal tract (CST) injury, an issue which is still a matter
of debate. To address this, we have begun using fMRI to examine differences
in motor planning activity in quadriplegic patients compared to normal
participants. This comparison will allow us to see to what degree the
activity in these areas degenerates after CST injury. The results of
these studies will provide an assessment of the feasibility of using
PRR recordings in patients with CST injury to control a prosthetic device.
Summary. A prosthetic system capable of decoding the desired
direction of movement from analysis of PRR activity would go a long
way toward improving the ability to communicate and the quality of life
of a locked-in patient. As defined by Plum and Posner (5), these patients
display a picture of quadriplegia, lower cranial nerve paralysis and
mutism with preserved consciousness and cortical function. Initially
such a prosthesis could be used to control a computer cursor, thereby
greatly improving the speed of communication, and enabling the locked-in
patient to utilize the internet as well as custom-designed software
applications. Eventually, a robotic arm (work currently being done by
our collaborators at E&AS and NASA/JPL) will be designed to implement
the decoded high-level commands of the PRR.
Prior to clinical trials of a PRR-based prosthesis the characteristics
of neural activity underlying motor planning in humans must be uncovered.
Recent work has shown that neural activity from macaque PRR is specific
to arm movements toward particular locations in space (6), while activity
in lateral intraparietal area can be used to predict the direction of
eye saccades in a working memory task (3). In the latter study, single
neuron firing rates (units) and local field potentials (LFPs) were recorded
as monkeys performed a delayed saccade to target task. Gamma band (25
- 90 Hz) power from both LFPs and single units exhibited maximal responses
for saccades in particular directions, with each recording location
exhibiting a different preferred direction. Most interesting was the
fact that not only were LFPs just as accurate in distinguishing between
saccades in the preferred and the anti-preferred direction as single
units, but they were more accurate in determining the timing of the
state transition associated with saccade execution. This opens up the
possibility of using human LFPs (as opposed to single unit activity)
to predict the direction and timing of PRR motor commands, which is
potentially quite important for the development of a cortical prosthesis
as LFPs are much easier to record than single units in awake behaving
animals.
Figure 1: a. Mobile recording apparatus. Apparatus includes detachable
touchscreen interface for presenting stimuli and monitoring patient
reaches, one rack mount server for controlling the behavioral stimulation,
one 64 channel amplifier, and one rack mount server for recording iEEG
signal. b. Raw trace recorded from left supplementary motor area in
one patient during the performance of the delayed reach task, illustrated
above the trace. The first red line indicates the time when the participant
started the trial by placing his hand on the fixation point in the center
of the screen; the green line indicates the onset of the target stimulus;
the second red line indicates the offset of the target and the beginning
of the delay period; the black line indicates the offset of the fixation
stimulus, which signals subject to make a movement to the former location
of the most recently presented target. c. MRI images from a a normal
participant during guided pointing. Coloring indicates regions exhibiting
significant increases in blood flow during actual pointing versus imagined
pointing. Red circles indicate putative human PRR.
To aid in the identification of human neural control signals we are
collecting intracranial recordings from participants at Huntington Memorial
Hospital while they perform delayed reaching tasks. Intracranial recordings
may be ethically obtained from patients with pharmacologically refractory
epilepsy. In these cases, invasive (Phase II) monitoring is frequently
used to precisely locate seizure foci and, by means of stimulation mapping,
locate functionally important regions of cortex. Patients undergoing
this protocol are observed over a period of 7 - 21 days (long enough
to obtain sufficient evidence to determine seizure foci) while recordings
are being taken from surgically implanted electrodes. Typically, patients
recover substantially within 2-4 days after the initial electrode implantation
and are generally willing, sometimes even eager, to participate in cognitive
experiments. iEEG recordings provide access to inferior and medial brain
regions, have a very high signal-to-noise ratio compared to scalp recorded
EEG, and are resistant to eye movement artifacts.
Figure 1a shows the mobile recording apparatus that we have built and
are currently using to record human iEEG. This device is mounted on
wheels and has a detachable touchscreen for easy transport and positioning
in the patients room. We have just finished recording from our first
participant using this apparatus and Figure 1b plots iEEG signal recorded
from a depth electrode in left supplementary motor area during the performance
of a delayed reach task (see Figure 1 legend for task details). Although
this first participant did not have any electrodes in parietal cortex,
we are currently analyzing the data using spectral methods to assess
the presence of task related activity in any of the 72 brain locations
that were sampled.
The second portion of this research project involves using fMRI to examine
activation in PRR when paralyzed patients plan arm movements and when
they think about executing arm movements. Because PRR is part of the
visuomotor system, rather than a part of the somatomotor system, it
is more isolated from direct connections with the spinal cord than primary
motor cortex and does not lose sensory feedback after paralysis. As
a result, it may be an optimal location for the placement of a neural
prosthetic. To test the hypothesis that human PRR does not undergo pathological
reorganization after CST injury we will compare fMRI activity from paralyzed
participants during a guided pointing task to that collected from a
normal control population. We have currently collected pilot fMRI data
from one normal participant during a real pointing task. The highlighted
regions in Figure 1c represent those brain areas in this participant
that exhibited significant increases in blood flow during actual pointing
versus fixation. The red circle indicates the putative human PRR, which
was significantly activated during this task along with pre-motor cortex
and motor cortex.
Results obtained in these experiments will assist in the assessment
of feasibility, target patient population, and eventual design of a
PRR-based prosthesis. The evaluation of iEEG during the performance
of a delayed reaching task will uncover the neural dynamics underlying
planned motor movements. This will assist in the development and refinement
of efficient decoding algorithms for utilizing a prosthetic device under
direct cortical control. Analyzing differences in fMRI activation between
paralyzed patients and normal controls will assess the feasibility of
using PRR as the neural command center for a cortical prosthesis by
assessing degeneration in this region due to CST injury. Given that
PRR codes for higher-level commands than primary motor cortex, it is
possible to speculate that this region may be less prone to reorganization
in the aftermath of CST injury than the primary motor region. In conclusion,
this research program will provide invaluable information for the development
of a practical neural prosthetic system in humans.
References
[1] Batista, A. P & Andersen, R. A. (2001) J. Neurosci. 85, 539–44.
[2] Snyder, L. H, Batista, A. P, & Andersen, R. A. (1997) Nature
386, 167–70.
[3] Pesaran, B, Pezaris, J. S, Sahani, M, Mitra, P. P, & Andersen,
R. A. (2002) Nat. Neurosci. 5, 805.
[4] Connolly, J. D, Andersen, R. A, & Goodale, M. A. (2003) Experimental
Brain Research In press.
[5] Plum, F & Posner, J. B. (1980) The Diagnosis of Stupor and Coma.
(FA Davis, Co).
[6] Batista, A. P, Buneo, C. A, Snyder, L. H, & Andersen, R. A.
(1999) Science 285, 257–60.
