Robot Therapist for Assisting in At-Home Rehabilitation of Shoulder Surgery Patients

Annually in the US, there are 53,000 shoulder replacements, 650,000 rotator cuff injury surgeries, and 4.5 million doctor visits for “extreme shoulder pain”. The mean age of shoulder surgery recipients is increasing (currently 50.9 years old), along with the number of surgeries performed. Of these surgery patients, 76% of traditional patients who do not fully adhere to their physical therapy regimen due to soreness, lack of supervision, and/or lack of incentive. Older patients are less likely to complete their post-surgery therapy regimen than younger patients. Patients are more likely to complete their exercise under supervision. However, patients prefer to exercise at home over traveling to a session. To resolve the abovementioned issues, we propose the use of an autonomous, interactive robot for patient rehabilitation as they recover from shoulder surgery. This robot introduces customized training and encouragement regimens, to increase physical therapy adherence and improve the patient\u27s recovery experience from the comfort of their own home. The initial range of motion and patient pain levels are recorded during calibration. OP2 guides the user through the physician prescribed rehabilitation program and movements. OP2 first demonstrates the exercises, then observes the user’s performance and gives performance feedback .The movements targeted for the purpose of this study include internal rotation, glenohumeral abduction, glenohumeral flexion, and passively performed pendulum movements. Research shows users respond positively to robots when they exhibit and respond to social cues. OP-2 provides active verbal feedback to the user based on performance. Role model systems: promotion (positive feedback) and prevention (negative feedback) vary in their effect on individual user motivation. Feedback style is altered throughout the rehabilitation program as OP-2 learns user preference and observes which styles generates the most patient improvement. Initial feasibility studies have proven promising. After completing the primary programming phase, the robot is currently able to: Perform calibration routines, demonstrate movements, retain user calibration information including pain values and initial range of motion of the joint, automatically generate a custom rehabilitation program based on initial calibration, track user movement, provide verbal feedback, and adapt feedback style based on patient performance