Walker EHE, Perreault EJ. Laterality of motor control specialization is not expressed by long-latency stabilizing reflexes during posture. Engineering in Medicine and Biology Conference 2014, Chicago IL, August 2014. (podium)
UPPER EXTREMITY ROBOTICS EXOSKELETON: APPLICATION, STRUCTURE AND ACTUATIONijbesjournal
Robotic exoskeleton is getting important to human in many aspects such as power assist, muscle training, regain motor function and rehabilitation. The research and development towards these functions are expected to be combined and integrated with the human intelligent and machine power, eventually becoming another generation of robot which will enhance the machine intelligent and human power. This paper reviews the upper extremity exoskeleton with different functions, actuators and degree of freedom (DOF). Among the functions, rehabilitation and power assist have been highlighted while pneumatic actuator, pneumatic muscle, motor and hydraulic actuator are presented under the categories of actuator. In addition, the structure of exoskeleton is separated by its DOF in terms of shoulder, elbow, wrist and hand.
UPPER EXTREMITY ROBOTICS EXOSKELETON: APPLICATION, STRUCTURE AND ACTUATIONijbesjournal
Robotic exoskeleton is getting important to human in many aspects such as power assist, muscle training, regain motor function and rehabilitation. The research and development towards these functions are expected to be combined and integrated with the human intelligent and machine power, eventually becoming another generation of robot which will enhance the machine intelligent and human power. This paper reviews the upper extremity exoskeleton with different functions, actuators and degree of freedom (DOF). Among the functions, rehabilitation and power assist have been highlighted while pneumatic actuator, pneumatic muscle, motor and hydraulic actuator are presented under the categories of actuator. In addition, the structure of exoskeleton is separated by its DOF in terms of shoulder, elbow, wrist and hand.
1. Effect of Arm Dominance
on Long-Latency Stabilizing
Reflex Gain during Posture
Elise H. E. Walker
Eric J. Perreault
Northwestern University &
Rehabilitation Institute of Chicago
EMBC 2014, Chicago IL USA | Thursday, August 28 | Motor Learning and Neural Control II 1
2. Arm dominance as control specialization in
the cortical hemispheres
__DOMINANT__
• Predictive
• Trajectory
• Limb dynamics
• Reach initiation
NONDOMINANT
• Impedance
• Position
• Stabilizing
• Final position
EMBC 2014, Chicago IL USA | Thursday, August 28 | Motor Learning and Neural Control II 2
Bagesteiro & Sainburg, J Neurophys 2002
“Handedness: dominant arm advantages in control of limb dynamics”
Bagesteiro & Sainburg, J Neurophys 2003
“Nondominant arm advantages in load compensation during rapid elbow joint movements”
Mani, Mutha, Przybyla, Haaland, Good, Sainburg, Brain 2013
“Contralesional motor deficits after unilateral stroke reflect hemiphere-specific control mechanisms”
?
3. Modified from Shemmell et al. 2009 J Neurosci 29(42)
Time (ms)
PTB
0.1 mV
-50 0 50 100 150
BicepsEMG
Stiff DNI
Compliant DNI
Stabilizing stretch reflexes: task-appropriate
feedback responses during posture
Hypothesis: Nondominant arm will display
more effective stabilizing stretch reflex
EMBC 2014, Chicago IL USA | Thursday, August 28 | Motor Learning and Neural Control II 3
LLR
4. Postural experiment in right-handed subjects
• Dominant (R) & nondominant (L) arm
• Stable & unstable environments
• Elbow flexor & extensor perturbations
EMBC 2014, Chicago IL USA | Thursday, August 28 | Motor Learning and Neural Control II 4
5. Perturbations of elbow posture elicit reflexes
EMBC 2014, Chicago IL USA | Thursday, August 28 | Motor Learning and Neural Control II 5
6. EMBC 2014, Chicago IL USA | Thursday, August 28 | Motor Learning and Neural Control II 6
Perturbations of elbow posture elicit reflexes
Onset of
Perturbation
EMG trace
Background
Activity
BGA
Short
Latency
Early Long
Latency
Late Long
Latency
SLR LLR1 LLR2
10%
MVC
0 50 10025 75-100 milliseconds
No
change
↑ sensitivity for
unstable, nondominant
7. Biceps
Some individual subjects display differences
between arms
Brachioradialis
Feedforward mechanism
EMBC 2014, Chicago IL USA | Thursday, August 28 | Motor Learning and Neural Control II 7
-50 ms 0 50 10025 75
R (dominant)
L (nondominant)
25%
MVC
Feedback mechanism
8. Unmatched reflexes were slightly higher in
some left-arm muscles
EMBC 2014, Chicago IL USA | Thursday, August 28 | Motor Learning and Neural Control II 8
TricepsLateralTricepsLong
LLLL RRRR
*** ***
**
* *
..
LL LL RR RR
*** ***
****
** **
**
.
*
**
.
BGA SLR LLR1 LLR2
10%
MVC
BGA SLR LLR1 LLR2
10%
MVC
9. Matched reflexes show no difference in
sensitivity for the two arms
EMBC 2014, Chicago IL USA | Thursday, August 28 | Motor Learning and Neural Control II 9
* *
* *
TricepsLateral
L RL RL RL R
BGA SLR LLR1 LLR2
10%
MVC
L RL R
TricepsLong
L RL R
BGA SLR LLR1 LLR2
10%
MVC
Arm dominance does not
influence reflex sensitivity
10. Arm dominance affects feedforward strategy,
but does not affect feedback gains
• Stabilizing reflexes in the nondominant arm do
not display greater sensitivity.
• Differences in reflex amplitude stem from
feedforward strategies of muscle activation.
• Our current understanding of postural reflexes
in the upper limb is probably not affected by
arm dominance.
EMBC 2014, Chicago IL USA | Thursday, August 28 | Motor Learning and Neural Control II 10
11. ACKNOWLEDGEMENTS
Support & Funding:
Northwestern
NSF 0932263
SMPP
Guidance:
Eric Perreault
Robert Sainburg
NMCL:
Timothy Haswell
Daniel Ludvig
Claire Honeycutt
David Lipps
Hyunglae Lee
Rosalind Heckman
Emma Bailllargeon
John Spanias
Mariah Whitmore
Bing Wang
Andrea Beer
EMBC 2014, Chicago IL USA | Thursday, August 28 | Motor Learning and Neural Control II 11
12. EMBC 2014, Chicago IL USA | Thursday, August 28 | Motor Learning and Neural Control II 12
13. Arm strength and task difficulty not
significantly different between arms
EMBC 2014, Chicago IL USA | Thursday, August 28 | Motor Learning and Neural Control II A1
Editor's Notes
Thank you
I’m Elise, I’m a graduate student in Northwestern’s biomedical engineering department
I’m here to talk about the effect of arm dominance on long-latency stabilizing reflex gain during posture
0) I just wanted to start out by acknowledging that this study was inspired by some very interesting work on handedness that has been coming out of Robert Sainburg’s lab at Penn State. Let me quickly go over a few of the main ideas from their work.
When most people talk about handedness there is often an implied assumption that the dominant hand, or arm, is simply more skilled than the nondominant limb. However, the work from Sainburg’s group has been generating evidence that each arm is actually specialized for different types of motor control.
For instance, the dominant arm performs better in predictive, trajectory planning tasks, is better able to account for limb dynamics, and is particularly good at initiating reaches in the appropriate direction.
In contrast, the nondominant arm is particularly suited to impedance control, which is important in posture and stabilizing tasks, and it is superior at achieving consistency in the final position of a reach.
By studying subjects with unilateral stroke, Sainburg’s group has linked these differences in motor performance to the dominant and nondominant cortical hemispheres. It seems that just as there are specialized circuits for language on one side of the brain, there may be control centers on each side of the brain which are specialized for specific types of motor control.
If this difference is substantial, we may have a dominant-sided bias in our understanding of motor control, since most studies are performed on the right limb of right-handed subjects.
0) This inspired me to investigate the question of handedness in one of the paradigms we use in our lab. We study long-latency stretch reflexes, particularly for conditions when reflex activity is apparently modulated according to the task. One example is what I will be calling the stabilizing stretch reflex
When a person is trying to maintain their arm posture and is suddenly bumped in a certain direction, there is reflexive activity in the stretched muscles that will tend to return the limb toward the original posture.
If this perturbation occurs in an environment with less stiffness or inherent stability, the reflex activity is often increased.
Importantly, this task-appropriate modulation occurs during the long-latency reflex period which is 50-100ms in the human arm. A cortical pathway is believed to be at least one contributor to long-latency reflexes, which may be why this reflex activity seems to be intelligently modulated. This is also why we believed that these reflexes could manifest differently in each arm, if they involve lateralized circuitry in the cortex.
Since the nondominant arm seems to excel at positional, stabilizing control, we hypothesized that it would display more effective stabilizing reflexes that the dominant arm, specifically in this long-latency window.
0) To test our hypothesis, we used an experimental block design that was run on 18 right-handed subjects.
Subjects were required to perform a postural task in both the dominant and nondominant arm,
And in both a stabilizing and a destabilizing haptic environment. These haptic environments were simulated using an admittance controller that was configured as a 2nd order mechanical system with either positive or negative stiffness. This is a little hard to grasp until you have experienced it,
but think about maintaining your arm posture straight up, where gravity is destabilizing arm position, versus straight down, where gravity stabposture.
Finally, we used perturbations in both directions to stretch both elbow flexor and extensor muscles.
0) Each arm was tested in a separate session, where the subject’s limb was attached to a linear motor.
With the help of feedback, the subject was required to maintain a postural target
In either haptic environment.
Perturbations were applied during posture to stretch the elbow muscles and elicit reflex responses
0) Muscle activity was recorded using electromyography
And normalized to maximum voluntary contraction
Reflex amplitudes were quantified as the average rectified value in three reflex time windows.
The voluntary background activity was also quantified as a measure of feedforward muscle drive, and to ensure that reflex sensitivity was compared at equal levels of preload.
Just to clarify, we expected no change in reflex sensitivity for the short-latency window, since this is purely spinal. In the long-latency time frame, a portion of activity is from the cortex,
So it was in the long-latency windows that we expected to see higher sensitivity in the nondominant arm as well as the unstable environment.
0) EMG activity was recorded in biceps and brachioradialis, as well as the lateral and long heads of the triceps. I’m just going to be showing you the results from a few muscles today, but there are more details in my paper.
Here is an example of EMG averages from one subject’s brachioradialis, in each arm. We can see that even before the perturbation, this subject had higher muscle activity in their left BRD. This resulted in larger reflexes as well, but not necessarily a larger reflex gain, because we know that reflex amplitude increases approximately linearly with the level of voluntary background muscle activity.
In contrast, the biceps of this subject has background activity is approximately the same in each arm. However, the long latency activity is still much higher in the left arm. This indicates a change in what I am here calling the reflex gain, or reflex sensitivity. This illustrates two different mechanisms that influence overall reflex activity…
the feedforward muscle drive, where heightened BGA leads to larger reflexes
And the feedback gain, where the sensitivity of the reflex loop is altered.
Our hypothesis was that arm dominance would specifically affect the feedback sensitivity, but we looked at both of these mechanisms by analyzing the EMG data both before and after a background matching procedure.
0) So first I will show you the reflex results before background matching, which could include influences from both feedforward and feedback mechanisms.
First, we have results in the lateral head of the triceps.
We have average EMG amplitude for the left arm
And the right arm in the stable environment.
And also for each arm in the unstable environment.
EMG activity was significantly higher in the unstable environment for each reflex window, but there was no significant effect of arm in this muscle.
If we look at these same results in the triceps long head
We can see that there was a similar effect of environment
And there was also a small effect of arm, with slightly higher activity in the nondominant, left arm. This was statistically significant in both short and long latency time windows. However, this difference may not be attributable to differences in reflex sensitivity until we eliminate the effect of background activity.
0) After background matching we performed the same analysis to isolate changes in feedback gains.
In both triceps heads there was still a significant effect of environment
Which indicated higher reflex gain during unstable conditions, now only in the LATE long-latency time period. However, after accounting for background activity, there were no significant differences between the reflex gains in the dominant and nondominant arm.
This suggests that the observed differences in reflex size were only due to increased levels of voluntary background activation in the nondominant triceps long. Arm dominance does not seem to influence reflex sensitivity.
0) So in short, we have demonstrated that arm dominance may affect feedforward strategies, but not feedback sensitivity during posture.
We did not find evidence in support of our hypothesis that the nondominant arm would display greater sensitivity of stabilizing reflexes.
Rather, any differences seen in reflexes seemed to be a result of feedforward activation patterns. The dominant and nondominant arm may have slightly different feedforward strategies during posture – we saw slightly higher activity in both the nondominant biceps and triceps long, which, interestingly, are the two biarticular muscles in this task – perhaps providing greater multijoint stability.
Finally, although it was a bit disappointing not to see a difference in reflex sensitivity between arms, it probably indicates that our understanding of these postural reflexes is not adversely affected by dominant-sided biases in the literature.
In conclusion, I would like to acknowledge the valuable help and guidance of my advisor Eric Perreault, as well as the input of Robert Sainburg who was kind enough to offer some initial feedback on the project.
I would also like to thank the organizations that have supported this work,
As well as all the people in my lab who have provided plentiful help and feedback.
Thank you.