Research On The Level Of Participation And Attention Of The Patient During Robot-assisted Neurorehabilitation

Stroke survivors suffer various deficits in brain that generate disability in their motor,perceptual,and cognitive functioning.These disabilities have a large impact on their independence of living.Traditional robot-assisted rehabilitation can deliver motor therapy according to the movement patterns of human body by imitating conventional therapy provided by physical therapists.However,results from clinical experiments with various rehabilitation robots have shown that force and movement set by the robotic device and repetitive motor training are not enough to achieve expected rehabilitation outcome.Focus in robot-assisted rehabilitation research has shifted to motivating the patient to actively participate and engage in the training exercise and providing adaptive treatment according to patient’s status and interest.These affordances of rehabilitation system could stimulate the neural system and facilitate neural plasticity,thereby promoting robotic devices in clinical rehabilitation.The objective of this thesis is to improve the efficiency of robot-assisted rehabilitation.The principal idea in this thesis is to provide personalized training and to identify the factors influencing patient’s engagement.The thesis mainly studies how the level of participation and attention change during the training process and how to interpret the changes.The research in this thesis include: i)investigating the factors influencing engagement in the context of rehabilitation and the indicators for evaluating engagement;ii)validating the effectiveness of the indicators in evaluating subject’s engagement during upper limb rehabilitation exercises;iii)investigating the influence of rehabilitation exercise integrated with video games on patient’s engagement;iv)demonstrating a smart rehabilitation system with the capability of enhancing patient’s engagement.Specifically,it first reviews engagement models proposed in various fields and identifies the common influencing factors of engagement from these models.Then the approaches to maintain engagement are proposed considering the context of rehabilitation.Based on the review of the indicators for evaluating engagement in the literature,the concept of participation and attention are identified,as well as the parameters to measure these indicators.Then,an experiment was conducted to validate the effectiveness of identified indicators in evaluating the subject’s engagement during rehabilitation exercises.An upper limb rehabilitation robot was used to deliver exercises under different training modes,including passive mode,active mode without restriction,and active mode with restriction.Then the differences in the engagement in different modes were analyzed in order to validate the indicators.Next,it investigates the influence of video games on patient’s engagement during rehabilitation training exercises.The rehabilitation system integrated with a Tetris game was designed and implemented.The system was tailored to fit stroke patients’ ability with adjustable difficulty level both in motor and cognitive aspects.Patient’s participation and attention were monitored during the game exercises with different difficulty levels.The results showed that the level of participation and attention decreased when the patient kept playing the game at the same difficulty level for a long period,and adjusting the difficulty level to a proper level can increase the patient’s participation and attention.Since video games exercises cannot maintain the patient’s participation and attention,rehabilitation system should have the capability of providing adaptive and personalized stimulations to maintain patient’s engagement.Last but not least,this thesis demonstrates the concept of a smart rehabilitation system that monitors the patient’s physiological and psychological status,and delivers personalized treatment using a learning mechanism.The learning mechanism can learn the effects of different stimulations on the patient’s participation and attention,so that it can provide personalized stimulations based on the learned contents.The effectiveness of the learning mechanism was validated based on the data collected from previous clinical experiment and simulated data respectively.