1287_ch04.pdf

(1822 KB) Pobierz
Motor Cortex in Voluntary Movements
4
Neuronal
Representations of
Bimanual Movements
Eilon Vaadia and Simone Cardoso de Oliveira
CONTENTS
Unimanual Components
Types
Bimanual Coordination
Bimanual Coordination?
4.6Conclusion
References
0-8493-1287-6/05/$0.00+$1.50
© 2005 by CRC Press LLC
Copyright © 2005 CRC Press LLC
 
4.1 INTRODUCTION
Simultaneous movements of the two arms constitute a relatively simple example of
complex movements and may serve to test whether and how the brain generates
unique representations of complex movements from their constituent elements, as
suggested by Leyton and Sherrington: “[T]he motor cortex may be regarded as a
synthetic organ for compounding … movements … from fractional movements.”
1
This chapter describes studies in which we attempted to investigate how the brain
assembles coordinated complex movements from their constituents, using the rela-
tively simple example of bimanual coordination.*
To do so, we have taken a neurophysiological approach, investigating neuronal
activity in behaving monkeys. The first question we ask is how the neuronal repre-
sentations of unimanual movements are combined to form bimanual movements. To
answer it, we compare neuronal activity during bimanual movements to the activity
observed during performance of their unimanual constituents. This approach may
provide preliminary evidence as to whether complex movements are coded differ-
ently from simple movements. Second, we need to define an approach to deciphering
the neuronal code for complex movements; namely, how we can pinpoint which
parameters of neuronal activity contain relevant information about the movement to
be executed. Previous work has suggested that in the motor system, rates of neuronal
populations are especially informative about the directions of upcoming movements.
2
). Given that each arm is mainly controlled by the contralateral
hemisphere, it is also likely that the temporal relationships between the hemispheres
are relevant to bimanual movements.
This chapter summarizes results we have accumulated to answer the above
questions, at least partially. We present evidence that bimanual representations
indeed exist, both at the level of single neurons and at the level of neuronal popu-
lations (in local field potentials). We further show that population rates and dynamic
interactions between the hemispheres contain information about the kind of bimanual
movement to be executed.
4.2 BIMANUAL-RELATED ACTIVITY OF SINGLE
NEURONS IN MOTOR CORTICAL FIELDS
who trained a monkey to press buttons with the fingers of either
hand separately or with both hands together. They recorded cortical neurons in the
medial aspect of the frontal cortex, which was called at the time the supplementary
* The term “bimanual coordination” literally means “coordination of the two hands,” yet this term has
been used in the literature in studies that relate not only to the coordination of the left and right hands,
but also of the left and right fingers, or of the left and right arms. This is also how we use the term in this article.
Copyright © 2005 CRC Press LLC
However, a number of studies, mainly on the visual system, have suggested that
temporal correlations between neuronal activities may contain information that is
particularly related to the compositionality of the coded items (e.g., the coherence
of moving bars
3,4,5
One of the first efforts to resolve the first question electrophysiologically was made
by Tanji et al.,
6
found that a substantial fraction of neurons in this
area were active during bimanual finger tapping and not during movements of the
finger of the right or left hand separately. This finding suggests that there are some
neurons that seem to be specific to bimanual movements. Their work appeared after
a behavioral study by Brinkman
6
who reported bimanual deficits consecutive to
SMA lesion. These and other studies (including clinical reports; for review, see
Brust
8
) inspired further studies focusing on the SMA as a major candidate area for
the control of bimanual coordination. Neuronal activity in SMA that is specific to
bimanual movements has now been described by a number of groups using different
tasks, although this specificity has been defined differently by different groups.
Neuronal activity during performance of a “drawer pulling task” was tested by
Wiesendanger et al.,
where monkeys performed naturally coordinated movements
without specific training. This task involved whole arm movements, where the
monkey was required to open a drawer with one hand and retrieve a raisin from it
with the other. Bimanual specific SMA activity has also been described by Kermadi
et al.,
10
although a different study on the same task reported that only a small
percentage of neurons was exclusively activated during bimanual movements.
11
12
Our group (including the authors of this chapter and Opher Donchin, Orna
Steinberg, and Anna Gribova) took another approach in an attempt to capitalize on
knowledge from the extensive studies of neuronal activity during arm reaching in a
center-out task.
13
4.2.1 T
B
IMANUAL
T
ASK
Macaque monkeys were trained to operate two separate manipulanda, one with each
arm. The manipulanda were low weight, low friction, two-joint mechanical arms,
oriented in the horizontal plane. Movement of each manipulandum produced move-
ment of a corresponding cursor on a vertical 21” video screen. The movement of
each cursor was mapped to its corresponding manipulandum movement such that
each millimeter of manipulandum movement yielded one millimeter of movement
of the cursor on the video display. The angular origin, 0
°
, was to the monkey’s right,
was away from the monkey for the manipulandum movement and toward
the top of the screen for the display.
A trial began when the monkey aligned both cursors on 0.8 cm diameter origins,
as shown in Figure 4.1 (where both cursors, left and right, are at their respective
origins) and held them still for 500 msec. For each arm, one of eight peripheral
target circles (0.8 cm diameter) could appear at a distance of 3 cm from the origin.
This small movement amplitude was chosen to minimize postural adjustments while
performing the movements. Movements taking the cursor from the origin to the
target were primarily small elbow and shoulder movements. Figure 4.2 presents a
few examples of trial types. In unimanual trials, only one target appeared (the upper
°
* SMA was later divided into SMA-proper and pre-SMA. See Reference 7.
Copyright © 2005 CRC Press LLC
motor area (SMA*). Tanji et al.
9
In what follows, we summarize a number of studies in which we
used a bimanual center-out reaching task to explore neuronal representations of
bimanual movements in the cortex.
HE
and 90
The monkey sits in a primate chair holding two manipulanda and facing a video
screen. Two cursors indicating the location of the manipulanda are shown on the screen (
).
Each cursor appears in the corresponding origin. Possible target locations are shown as circles
surrounding each origin. (Modified with permission from Reference 15.)
+
The behavioral task illustrated by examples of types of trials that were used in
the various experiments. The empty circles are not visible to the monkey. The figure displays
examples of unimanual movements (upper two rows) to 90
°
(up) and 270
°
(down) and
bimanual movements (lower row).
Copyright © 2005 CRC Press LLC
FIGURE 4.1
FIGURE 4.2
356521638.001.png 356521638.002.png 356521638.003.png
two rows in Figure 4.2 ) and the monkey moved the appropriate arm to bring the
corresponding cursor into the target, but did not move the other arm. If two targets
appeared — signaling a bimanual trial — the monkey had to move both arms, such
that the two cursors were moved into the target circles on the screen.
These structured movements made it possible to study well-controlled bimanual
movements of various types. For example, parallel movements and opposite move-
ments (lower row, Figure 4.2) were composed of unimanual movements shown in
the upper rows of Figure 4.2. (Figure 4.2 shows only one direction per arm; in all
cases additional directions were studied.) Other combinations, where each arm was
required to move in a different direction or to cover a different distance, were also
tested, as for example the movements shown in Figure 4.2 (bottom-right plot) where
the arms move at 90° to each other.
4.2.2 M
ONKEY
B
EHAVIOR
Neuronal activity was sampled after the monkeys were over-trained to perform
bimanual trials with the two arms starting to move together and reaching the targets
together quite accurately. For example, the two monkeys used for the data presented
in this section initiated the bimanual movements with average interarm intervals
(IAIs) of 16 to 21 msec (SD = 56 to 74 msec) and reached the targets with an average
IAI of 5 to 15 msec (SD = 106 to 125 msec). These IAIs are quite short, much
shorter than would be required for successful performance of the task, meaning that
the monkeys tended, like humans, to synchronize their movements rather than
attempting to perform two separate movements. The movements used in the tasks
were small (a length of 3 cm for all movement types presented in this section and
up to 6 cm in some types of movements for the experiments described in Section 4.5 ) .
The hand trajectories made to a given direction were quite similar for different
movement types (but not identical; see Figure 4.6 ) . Further, video camera observa-
tions and electromyographic (EMG) recordings failed to detect consistent variations
in postural adjustments during the arm movements. EMGs of muscles on the forearm,
the upper arm, the shoulders, and the back were also recorded simultaneously with
neuronal activity (selected sessions). Various analyses were carried out in order to
detect changes in neuronal activity that could emerge from different patterns of
muscle activation. (For details about specific control measures comparing EMG
during unimanual and bimanual movements see Donchin et al.
14
)
4.2.3 N
EURONAL
R
ECORDINGS
IN
MI
AND
SMA
) The activity of 8 to 30 isolated
neurons and up to eight local field potential (LFP) channels was recorded each
session. The data discussed in this article were recorded from 3 monkeys and
included the activity of more than 438 neurons (232 in M1 and 206 in SMA). To
detect
et al
14
evoked activity,
we tested the firing rate in a 500-msec period from 100 msec
Copyright © 2005 CRC Press LLC
Single-unit activity and local field potentials were recorded from homologous sites
in the two hemispheres, from the primary motor cortex (M1) and from SMA proper.
(For details on recording sites see Donchin
.

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika:

Zgłoś jeśli naruszono regulamin