Integrated-Regression and Whale Optimisation Algorithms to post stroke rehabilitation analysis: A case study for serious games

Stroke is one of the leading causes of death and long-term disability worldwide. The primary goal of post-stroke rehabilitation is to maximize the independence of the affected individuals by facilitating both neurological and compensatory functional recoveries in their daily lives. This paper presents an integrated study, using multiple linear regression (MLR) and whale optimization algorithm, for optimizing the post-stroke rehabilitation performance using Medical Interactive Rehabilitation Assistant (MIRA). MIRA is one of the serious games using virtual reality. The research objective finding the best games settings for MIRA. Analysis of variance and F-test were performed to justify the goodness of fit of the developed mathematical models, as well as the significance of the individual MIRA input settings of the games. Results demonstrated that the developed MLR models for average acceleration and distance were significant, with P=0.018 and P<0.000, respectively. Data from 41 stroke patients revealed that the maximum muscle strength that can be achieved was 2.56 cm/s2 while the maximum distance was 104.83 cm. The results showed that all patients outperformed the optimal average acceleration value by between 6% and 29%, whereas only two patients outperformed the optimal distance value. The results from this study may be integrated using MIRA (exergames) to guide therapists in finding the best combination of input settings that can maximise the performance output of post-stroke patients.