Background: Our team developed ROBiGAME, a serious game implemented on an end-effector upper limb (UL) robot that allows simultaneous rehabilitation of motor and cognitive functions. This serious game uses an auto-adaptive mechanism that dynamically adjusts task-difficulty to patients’ degree of impairments. Game characteristics (target position, level of assistance/resistance, number of distractors) continuously vary according to a series of real-time performance indicators (mean success rates, UL kinematics, UL force and reaction time). This study investigated whether this dynamic control mechanism is optimal to deliver rapidly adapting and individualized UL therapy after stroke. Methods: 24 stroke survivors were enrolled in a 5-day protocol. Detailed clinical and robotic assessments were conducted during initial evaluation. Then, each participant completed a daily 45-minute session of robot-assisted therapy (RAT) playing with ROBiGAME during three consecutive days. Relationships between clinical/robotic assessments and game difficulty parameters were studied, regarding motor and cognitive aspects, throughout therapy sessions. Results: Game difficulty parameters oscillated towards a steady state half-way through the first session (after 10 to 15 completed tasks) displaying a rapid adaptation of the system. Mean success rates averaged 75.0% (SD: 7%) indicating an optimal challenge point to maintain motivation and enhance learning during therapy sessions. Finally, motor (Fugl-Meyer Assessment, Action Research Arm Test, Box and Block Test) and cognitive (Bells test, Apples test, MonAmour test) scores strongly correlated to respective difficulty parameters (FMA: r=0.76 p<0.001; ARAT: r=0.71 p=0.001; MonAmour test: r=-0.82 p=0.001) showing that task-difficulty was well adapted to patients’ individual needs. Conclusions: This study illustrates that ROBiGAME’s dynamic difficulty adjustment, during combined motor and cognitive rehabilitation, is tailored to each patient and remains optimal throughout therapy. Future work should evaluate effectiveness of this type of control mechanism for robotic UL rehabilitation after stroke.
Motor and cognitive dynamic difficulty adjustment for robot-mediated upper limb rehabilitation after stroke
2020-01-01
Conference paper
Electronic Resource
English
DDC: | 629 |
Upper Limb Robot Mediated Stroke Therapy-GENTLE/s Approach
British Library Online Contents | 2003
|