Enhancing engagement of patients during stroke rehabilitation exercises are in the focus of current research. Various methods and computer supported tools have been developed for this purpose, which try to avoid mundane exercising that is prone to become a routine or even boring for the patients and leads to ineffective training. This paper introduces an engagement enhancing cyber-physical stroke rehabilitation system (CP-SRS) aiming at enhancing the patient's engagement during rehabilitation training exercises. This paper focuses on introducing the implementation and validation of the engagement monitoring subsystem (EMS) in the CP-SRS. The EMS is expected to evaluate the patient's actual engagement levels in motor, perceptive, cognitive and emotional aspects. Experiments in these four aspects were conducted separately, in order to characterize the range and accuracy of the engagement indicators by influencing the subjects into different engaged states. During the experiments, different setups were created to mimic the situations in which the subject was engaged or not engaged. The subjects involved in the experiments were healthy subjects. Results showed that the measurement in motor, perceptive, cognitive, and emotional aspects can represent the corresponding engagement level. More experiments will be conducted in the future to validate the efficiency of the CP-SRS in enhancing the engagement with stroke patients.

@en

Maintaining and enhancing engagement of patients during stroke rehabilitation exercises are in the focus of current research. There have been various methods and computer supported tools developed for this purpose, which try to avoid mundane exercising that is prone to become a routine or even boring for the patients and leads to ineffective training. This paper proposes a strategy bundle-based smart learning mechanism (SLM) to increase the efficiency of rehabilitation exercises. The underpinning strategy considers motor, perceptive, cognitive and emotional aspects of engagement. Part of a cyber-physical stroke rehabilitation system (CP-SRS), the proposed SLM is able to learn the relationship between the actual engagement levels and applied stimulations. From a computational point of view, the SLM is based on multiplexed signal processing and a machine learning agent. The paper presents the mathematical concepts of signal processing, the reasoning algorithms, and the overall embedding of the SLM in the CP-SRS. Regression and classification are two possible solutions for this learning mechanism. Computer simulation is conducted to investigate the limitations of the proposed learning mechanism and compare the results of different machine learning methods. We simulate regression with artificial neural network (ANN), and classification with ANN and Naive Bayes (NB). Results show that classification with NB is more promising in practice since it is less sensitive to the deviations in the inputs than the applied version of ANN.

@en

Maintaining and enhancing engagement of patients during stroke rehabilitation exercises are in the focus of current research. In the preceding phase of our research, an entry-level cyber-physical stroke rehabilitation system (CP-SRS) has been developed, with the aim of enhancing patients' overall engagement during rehabilitation exercises. As a follow up on the evaluation of the proposed engagement enhancing method and the smart learning mechanism based on the simulated data, this paper presents the validation results of the proposed CP-SRS system based on real-life data. Validation included two aspects: (i) validation of the effectiveness of the applied stimulation strategies (SSs), and (ii) validation of the accuracy of the suggestions of the smart learning mechanism. Methodologically, a within-subject experiment was designed and completed. Eighteen subjects were recruited to participate in the experiments, based on convenience sampling. During the completed game exercises SSs were applied individually as well as in combination. The engagement levels of the participants were evaluated and recorded after applying the SSs individually and combined. The results were processed by within-subject ANOVA in order to test if there was a significant difference between the influences of the different SSs and combinations. In addition, training and testing of the smart learning mechanism (SLM) was also executed in MATLAB. The results indicated that several SSs significantly increased the engagement of the subjects, and that both neural network-based SLM and the Naive Bayes-based SLM were able to learn and discriminate the effects of the various SSs. Our conclusion is that they both can be used to assist making decision on effective application of SSs. However, applying neural network-based SLM is more appropriate in the context of increasing engagement.@en