Telerehabilitation defined as an alternative against the usual rehabilitation service which helps patients in rural areas to get access to a less service that is practical concerning logistics and costs in a controlled environment. But increase in usage of mobile phones or other wireless devices. Medical Interactive Rehabilitation Assistant (MIRA) software is one of the software capturing the data of patient's rehabilitation progress and data analysis. Physiotherapists are using a default setting in rehabilitation for patients in exergames. This can lead to lack of appropriate settings for the patients. The default setting includes level of difficulty, percentage of tolerance and maximum range. This paper focus on optimization of patients' average percentage (range of motion) from rehabilitation, as well as profiling setting according to the estimated optimal results. Moreover, the optimal results of parameters setting would be the profiling setting for the future rehabilitation. Regression analysis, and Genetic Algorithm (GA) are implemented for modelling and optimization process. According to the result, the maximum average percentage (40.325%) is found at the combination of easy level of difficulty, 10% of tolerance and 100% of maximum range.