Integrated Robotic Evaluation and Training for Ataxia: Establishing A Pilot Framework

Jowy Tani, Ray Chen, Jia-Ying Sung, Ruzena Bajcsy, Sheng-Wen Su

Research output: Contribution to journalArticlepeer-review


Research Objectives
To collect the performance data of healthy individuals participating in a bimanual robot-assisted Constraint-Induced Movement Therapy (CIMT) shaping activity, and evaluate the performance data using computer vision techniques, in order to establish a reference interval for specific performance parameters of the robotic shaping activity.
The study is a cross-sectional study that collects data from healthy individuals. Data is collected during a 7-day period.
The study is performed in a robotics laboratory.
Eleven healthy individuals (aged 27.36±6.39 years, 3 females) participated in the study.
Shaping is an important component of CIMT. Shaping is "a desired motor or behavioral objective is approached in small steps, by successive approximations" (Taub, 1994). In accordance with the definition, we designed a shaping activity featuring interaction between subjects and a bimanual research robot running customized shaping activity software.
Two robotic shaping activity is carried out. In the first activity, the patients are instructed to move metal pieces from a bowl into a cup held by one of the robot’s gripper. The robot would move its gripper continuously following pre-set trajectories, and the patient is instructed to attempt to move as much as possible metal pieces into the moving cup during. In the second activity, the patients are instructed to continuously follow the movement of the gripper using his/her index finger of the dominant hand.
All patient motion is captured using a Kinect motion-sensing input device, and the accuracy, smoothness, and submovements indices are calculated using customized computer vision software.
Main Outcome Measure(s)
The parameters measured in the present study are shaping activity score (metal pieces placed per minute), and length ratio (ratio of path actually traveled by the subject and the desired one). Other accuracy, smoothness, and submovement indices are also obtained.
The 95% confidence interval for the primary parameters are: shaping activity score, 33.94 - 39.84; length ratio: 108.27 – 113.09%.
The present study has successfully established the basis of a reference interval for specific performance parameters of the robotic shaping activity for a specific age group. We would use the present study as a foundation for a future robotic coordination evaluation and training study in ataxia patients.
Original languageEnglish
Pages (from-to)E96-E96
Number of pages1
JournalArchives of Physical Medicine and Rehabilitation
Issue number10
Publication statusPublished - Oct 1 2018


Dive into the research topics of 'Integrated Robotic Evaluation and Training for Ataxia: Establishing A Pilot Framework'. Together they form a unique fingerprint.

Cite this