Stroke is a leading cause of long-term disability. Trunk compensation is the most common movement strategy to substitute limited arm reaching extent during goal-directed arm reaches for daily activities. There are a number of factors that impact motor control strategies, such as individual, environmental, and test condition factors.
Previous studies have focused on the individuals and environmental factors. However, the test condition factors have not been well studied regarding compensatory movement strategies in chronic stroke survivors. We designed this protocol to investigate the impact of motor test conditions or motor control strategies of goal-directed arm reaches in chronic stroke survivors.
Specifically, we aim to determine how post-stroke individuals utilize different movement strategies when they are performing goal-directed arm reaching movements with different tasks goals. We use this protocol to test our hypothesis that chronic stroke survivors will increase trunk compensation during goal-directed arm reaching movements as a functional task complexity and difficulty. In this section, we are going to explain the kinematic measurement setup.
We use 10 Vicon three-dimensional motion capture cameras with motion monitor software to record goal-directed arm reaching kinematics. Using the Vicon Nexus software version of 2.8.2, we set up the motion capture cameras, masking unwanted reflections, camera calibration, and volume origin setup. After that, we set up the kinematic data collection using the motion monitor software.
This procedure includes mapping markers, setting up the stylus, setting up world axes, and assigning virtual sensors to body segments. Then we put the marker triads on the participant. The trunk marker triad is positioned on the skin over the thoracic vertebrae between scapulas.
The upper arm triad is placed on the skin in the middle of each upper arms lateral aspect. The forearm triads are placed on the skin in the middle of the dorsal surface of each forearm. The hand triad is placed on the skin over the third metacarpal bone.
We also place a marker triad on the table. This marker triad is used to register the location of the home and target positions. We made a custom chopstick stylus to record the kinematics of the chopstick.
The chopstick stylus has a marker triad and this is also registered in the model. After we place all the marker triads on the participant, we set up subject sensors using digitization methods. The procedure registers the body segments to the model based on the marker triads location and the software calculates the positions of different joint centers.
By following onscreen prompts, we point to the following landmarks using the stylus. For upper trunk, one spot between C7 and T1 vertebrae. For lower trunk, one spot between T12 and L1 vertebrae.
For the shoulder joint, two spots equal distant from the middle of the head of the humerus. For the elbow joint, two spots on the medial and lateral elbow that are equidistant from the midline of the joint. For the wrist joint, two spots on the medial and lateral wrist that are equal distance from the middle of the joint.
For the hand, the tip of the third phalanx of each hand. For the home and target positions, a spot on the center of each position. For the chopstick stylus, a spot on the tip of the chopstick.
There are four different goal-directed arm reaching motor task conditions. Here, we are going to explain the details of those motor task conditions. There are two different target sizes indicating the task difficulty.
A large target would be an easier task condition. And the smaller target is a more difficult task condition. Also, there are two different test types, which indicate the task complexity.
A pointing task is a simpler motor task condition and picking up an object with a pair of chopsticks is a more complex motor task condition that requires high level hand dexterity. As a combination of those two tasks conditions we have four different motor task conditions. This figure shows the template for pointing to a large target.
There are home and target positions. Each square size is one by one square centimeter. The center to center distance between the two locations is 20 centimeters.
The task goal for pointing to a large target is to reach and tap the center of the target square with the chopsticks tip as quickly and accurately as possible. The participant holds a chopstick and locate the chopstick's tip on the center of the home position. When the participant hears a go sound, he or she reaches and taps the center of the target square as quickly and accurately as possible.
The participant has a three seconds to complete the task. There will be a stop signal after three seconds from the go signal. If the participant cannot complete the task within three seconds, it is considered a failed trial.
The participant repeats this task 10 times with ten second rests between each trial.Go.Stop. We use the same template for the picking up a large object task using a pair of chopsticks. A plastic cube, one centimeter on edge, is placed on the target location.
The task goal is to reach and pick up the plastic cube about an inch height with a pair of chopsticks as quickly as possible without dropping. The participant holds a pair of chopsticks and locates the tips on the center of the home position. When the participant hears a go sound, he or she reaches and picks up the cube as quickly as possible.
The participant needs to pick up the cube before the stop signal, which has given three seconds after the go signal. If the participant cannot pick up the cube within three seconds, it is considered a failed trial. He or she is asked to bring the tips of the chopsticks back to the home position.
Dropping or flying the plastic cube during the task is considered a failed trial.Go.Stop. The pointing to a small target is the same as pointing to a large target, but the square target size is 0.3 by 0.3 square centimeters. The participant holds the chopsticks and locates the chopstick's tip on the center of the home position.
When the participant hears a go sound he or she reaches and taps the center of the target square as quickly and accurately as possible.Go.Stop. The picking up a small object task is the same as picking up a large object task, but the target object is 0.3 centimeters on edge. A plastic cube 0.3 centimeters on edge will be placed on the target location.
The task goal is to reach and pick up the plastic cube about an inch height with a pair of chopsticks as quickly as possible without dropping. The participant holds a pair of chopsticks and locates the tips on the center of the home position. When the participant hears a go sound, he or she reaches and picks up the cube as quickly as possible Go.Stop.
In this section, we are going to explain the goal-directed arm reaching kinematic data analysis. We export the position data of the following landmarks from the motion monitor software. Tip of the chopstick stylus, home position on the table, target position on the table, each hand in the middle of the third phalanx, each center of elbow joints, each center of shoulder joints, C7 spine representing the trunk movement.
Each participant's upper extremity joint landmarks and the trunk position data are exported in the X, Y and Z axes as a text file for each task condition. The kinematic data are preprocessed using custom scripts and MATLAB software. The kinematic data pre-processing includes filtering using a third order Butterworth low-pass filter with a three Hertz cutoff.
Then we calculate the resultant of X, Y and Z directions of the performing hand position. After the pre-processing of the position data, we performed the kinematic data analysis using the resultant of the three-dimensional position of the performing hand to calculate the kinematic variables of the goal-directed arm reaches. We use custom scripts and MATLAB software for kinematic data analysis.
First of all, we calculate tangential velocity, acceleration and jerk of the performing hand, which are the first, second, and third derivatives of the position data, respectively. Then we use each trial's tangental velocity profile to determine the movement onset, offset and the peak velocity. The following kinematic variables were calculated from the kinematic data analysis, movement duration, peak velocity, absolute and relative time to peak velocity and log dimensionless jerk.
Here, we are going to describe these kinematic variables with an exemplar velocity profile. Movement onset and offsets are identified using the movement onset and offset thresholds, which is 0.01 meters per second. Movement onset is defined as the first frame of the reach where the tangential velocity is above 0.01 meters per second.
Movement offset is defined as the last frame of the reach where the tangental velocity is above 0.01 meters per second. The movement duration is defined as the time between the movement onset and offset. Peak velocity is the maximum velocity amplitude between the movement onset and offset.
Time to peak velocity is the time to reach the peak velocity from the movement onset. The movement onset, offset and peak velocity are automatically labeled using MATLAB software's custom scripts. After this automated labeling the labels are visualized and inspected by an investigator.
If the labels are incorrect, the investigator makes manual adjustments. Log dimensionless jerk is calculated from the jerk profile of the reach using this equation, which is the third derivative of the position. We calculated two measures of trunk compensation during goal-directed arm reaches.
First, trunk displacement was calculated. This measure is the distance difference of the trunk landmark, C7 between movement onset and offset. The other trunk compensation measure is the shoulder trajectory length.
This measure is the travel distance of the shoulder landmark between arm reaching movement onset and offset. The shoulder trajectory length is a new trunk compensation measure during goal-directed arm reaching. We employed this measure to capture the trunk compensation in all directions.
We use all three dimensions to calculate these trunk compensation measures. In this section, we are going to present our preliminary results. Our preliminary study has two chronic stroke survivors with mild upper extremity motor impairment on the right side and two non-disabled young adults.
All the participants had no or little previous experience of chopstick use. Chronic stroke survivors perform the motors task using their paretic right hand, which was dominant before stroke onset. Non-disabled young adults performed motor tasks with their right hand, which is their dominant hand.
Here, we compared two different trunk compensation measures to determine if the shoulder trajectory is a more sensitive measure than the trunk displacement to capture the trunk compensation. Figure one shows the violin plots of trunk compensation measures in two different participant populations. The green plots indicate the shoulder trajectory length and the red plots indicate the trunk displacement.
Each dot in the violin plot indicates each arm reaching movement. Figure one shows that non-disabled adults and chronic stroke survivors have no difference in trunk displacement. While the shoulder trajectory length of chronic stroke survivors is greater than that of non-disabled adults.
This result may indicate that the shoulder trajectory length is a more sensitive measure of trunk compensation during goal-directed arm reaches than the trunk displacement in post-stroke individuals. Figure two addresses differences in kinematic variables between participant populations and between different motor tasks. In this figure, red outlined box plots indicate the chronic stroke survivors and blue outlined box plots indicate the non-disabled young adults.
And the X-axis, we have four different tasks conditions pointing to a large target, pointing to a small target picking up a large object and picking up a small object. Chronic stroke survivors had different goal-directed arm reaching kinematic characteristics compared to non-disabled young adults across different task conditions. In summary, chronic stroke survivors had slower and jerkier arm reaching than the non-disabled young adults.
Also, chronic stroke survivors were more dependent on feedback-based adjustments of the reaching, which is indicated by the lower relative time to peak velocity. These results are consistent with previous findings. Lastly, figure two also demonstrates that the task complexity impacts the kinematics of goal-directed arm reaching movement.
Participants utilize slower and jerkier movements for more complex motor tasks than simple motor tasks for both participant populations. Besides they use more feedback-based control of the arm reaching for a more complex task. People also tend to use more trunk compensation for more complex motor tasks than simple motor tasks.
Our preliminary results support that this protocol can be used to investigate the impact of test conditions on goal-directed arm reaching movement strategies in chronic stroke survivors.
