The project targets stroke survivors to investigate the effect of augmented feedback (using robotic force cues and visual feedback) on their upper limb reaching patterns and trunk compensatory movements.

Purpose:

For stroke survivors, the use of compensatory movements can lead to a reduction of range of motion, pain, and a pattern of "learned non-use". A common compensatory movement present during upper limb reaching is trunk displacement. Although this motion has been identified as an important one to be reduced, few strategies for addressing this problem have been considered. The existing strategies require physical restraint of the person to the back of a chair, making them undesirable for use in unsupervised therapy. As a result, there is a current need for alternate methods that promote the use of correct movement patterns both in the clinic and in the home. In this sense, technology can act as an enabler to create new ways of reducing trunk compensation. Still, there is a gap in the literature as trunk compensation has only been investigated as a secondary theme in robotic and computer-aided rehabilitation. Consequently, in this project the investigators will look into the reduction of trunk compensation using robotic devices and commercially available technology, to enable a focus on the quality of the movements in unsupervised therapy. The potential results from this project could later be applied and generalized to other modes of compensation in stroke and other neurological disabled populations.

Objective:

Our objective is to demonstrate that feedback cues could be used to reduce trunk compensatory movements in unsupervised therapy.

Research Question:

Will the use of force cues attain similar results when compared to visual feedback to reduce trunk compensation?

Equipment:

• 1 Kinect (Microsoft, Inc.) markerless motion capture system.
• 2 Kinova Jaco assistive robotic arms. These devices are used to assist disabled people on daily tasks. The arms are designed for safe interaction (low forces) with the user.
• 1 Desktop computer and a monitor to deliver the visual feedback.

Method:

The investigators will design the force feedback cues using two Kinova Jaco robotic devices to deliver them. The investigators will then investigate if the force cues could achieve similar or better results when compared to visual feedback (using a computer monitor to deliver the feedback) to reduce compensatory movements.

Even though robotic devices are capable of exerting force cues, these devices can be difficult to implement in a home setting due to their complexity and price. Consequently, in this project the investigators will study the use of commercially available technology to deliver feedback about the users' compensatory movements. The investigators will compare the use of force and visual feedback to study their effect on reducing compensatory levels. The investigators have chosen to use visual feedback as it can be delivered by systems readily available in the home (e.g. television sets and personal computers) without the need to purchase more complex robotic devices. If the results from this project support the use of commercially available technology, this will provide further evidence of the potential use of this technology for substituting or complementing robotic devices once the users continue their therapy programs at home.

The force feedback cues will be provided as resistance to move the robots' handles. These cues will be applied when the user moves outside a certain error band, based on a "normal" reaching pattern. In addition, the magnitude of the cue will be proportional to the magnitude of trunk compensation. In this project two conditions will be compared: visual feedback and force feedback.

The visual condition will use a monitor to display two cursors (empty circles) that will represent the participant's hands, and the circles will fill with red ink as the user starts to compensate. As the magnitude of compensation increases, the amount of red ink will gradually increase to indicate the level at which the user is compensating.

Summary of Procedures:

(Total Time: 2-2.5 hours):

1. Participants will be recruited.

2. Introduce the study and equipment. Participants will be asked to inform the investigator if they feel uncomfortable or fatigued at any point during the experiment, and will be given as many breaks as needed.

3. Ask participants to fill out consent forms, or if their unable to provide consent due to their health condition, their caregivers/guardians will provide consent and the subject will provide assent.

4. Participant will fill out a background questionnaire and a registered physical therapist will conduct a clinical assessment based on recognized impairment scales (Fugl-Meyer Upper Extremity Assessment and Reaching Performance Scale) to use the scores as a baseline for the comparisons that will be performed in the statistical analyses at the end of this phase. The Reaching performance Scale requires the use of video recording of the assessment for scoring.

   If the participant wants to know their clinical assessment results, at the end of the session, the therapist will provide a photocopy of the results, and will give the participant and explanation of these results and answer any questions that the participant may have about these scales. In the case of the Reaching Performance Scale as the scoring is done after the study, the participant could receive their scores at a later date via a telephone call.

5. All the study sessions will be conducted at the University of British Columbia (UBC) Point Grey Campus. During the test, the participants will be asked to interact with a computer through the use of the following input technologies: 2 Jaco Kinova robotic arms and a Microsoft Kinect. Using these technologies, the participants will perform bimanual symmetric movements with their arms/hands to control a simple cursor/target videogame. The Kinect will measure the participant's movement as data points for every joint, no video will be recorded. The robotic devices will be used to record the hands' movements and will increase their resistance to be moved based on the level of trunk compensation of the participant. The monitor will be used to provide visual feedback about the participant's trunk compensation and to display the target game.

6. Ask participants to sit in a chair and adjust footrest to have their feet fully supported, their knees at a 90 degree angle, and their back against the chair.

7. Ask participants to hold on to the handles of the two robotic devices.

8. In case participants are not able to hold the handle due to hand weakness, an adjustable fabric and elastic strap will be fitted around their palm to hold the hand on top of the handle.

9. The maximum force that the user can produce to push the robots will be measured by reading the robot's sensors.

10. Ask the participant to perform a series of unimanual reaches to calibrate the system based on the participant's arm's length.

11. Ask participant to perform 5 practice bimanual reaches to become familiar with the system and the motion mapping.

12. Ask participant to perform 15 baseline (no feedback) bimanual reaches to measure their trunk compensation. The investigators will use the value of average trunk compensation to set the error bands for the visual and force feedback.

13. Ask participant to perform 5 practice bimanual reaches to become familiar with either the visual or force feedback (depending on the randomization of subjects).

14. Ask the participant to perform 60 trials of bimanual reaches to 1 target at knee height with arms fully extended. Participants will receive feedback about their compensation through the computer's monitor or as increased resistance to move of the robots.

15. The participant will be able to rest between targets if requested. In addition there will be 1 minute rests after every 15 targets.

16. The participant will perform 15 reaches without any feedback (Post measurement)

17. The participants will have 5 minutes of break before starting the second feedback condition.

18. Repeat steps 14, 15 and 16, but with the other type of feedback (visual or force).

19. A note taker will record the occurrence of obstacles encountered by the participants during the study.

20. The motion tracking data and assessment videos will be saved on a computer file, backed up on a UBC-based file server and on optical media.

21. At the end of the session the participant will answer a usability questionnaire.

22. The data from the note taker, motion logs, assessment videos and questionnaires will be used to conduct a quantitative and qualitative analysis to gain further insight into how augmented feedback can reduce compensatory trunk movements, and the ease of use and functionality of the system. All the data will be identified using participants' numbers.

23. The video recordings will be erased/destroyed 5 years after publication of results.

Study Design:

The investigators will follow a within-subjects crossover design with the independent variable being the feedback type and the levels will be: force feedback and visual feedback. The primary dependent variable will be the measure of trunk compensation. The investigators will follow a counterbalanced strategy to reduce the carryover effects from performing the two conditions in a certain order.

Data collection will include motion log files, scores from the game, discussions with participants and exit surveys.

Inclusion Criteria:

• At least 19 years old
• Hemiplegia as a result of a non-traumatic cerebral stroke (ischaemic or hemorrhagic)
• Stroke occurred at least 3 months prior to study
• Ability to understand/follow directions and answer questions in English
• Ability to maintain a sitting position in a standard office chair without arm rests, independently or with minimal supervision, for 1.5 hours.
• Have the ability to perform the following movement several times with their weak arm (while seated): move their hand to their hip (on the same side as the weak arm), it's acceptable if participants use their trunk to help themselves, and from that point of flexion moving the hand forward (without touching their thigh) to touch their knee (on the same side as the weak arm). Participants should be able to do this movement without any help from their strong hand.

Exclusion Criteria:

• Upper limb orthopaedic surgery in the past 3 months
• Shoulder subluxation or significant shoulder pain
• Trunk pain
• Other orthopaedic or neurological conditions affecting the arm or trunk
• Uncorrected visual impairment