Stroke is the second largest cause of death worldwide and one of the most common causes of disability. However, several approaches have been proposed to deal with stroke patient rehabilitation like robotic devices and virtual reality systems, researchers have found that the brain-computer interfaces (BCI) approaches can provide better results. In this study, the electroencephalography (EEG) dataset from post-stroke patients were investigated to identify the effects of the motor imagery (MI)-based BCI therapy by investigating sensorimotor areas using frequency and time-domain features and to select particular methods that help in enhancing the MI-based BCI systems for stroke patients using EEG signal processing. Therefore, to detect the imagined movements that are typically required within conventional rehabilitation therapy with good identification accuracies, the conventional filters and wavelet transform (WT) denoising technique was used in the first stage. Next, attributes from frequency and entropy domains were computed. Finally, support vector machine (SVM) classification techniques were utilized to test the motor imagery (MI)-based BCI rehabilitation. The results demonstrate the capability of the WT denoising technique together with the used features and SVM classifier to discriminate the tested classes of the left hand, right hand and foot MI-based BCI rehabilitation. This study will help medical doctors, clinicians, physicians and technicians to introduce a good rehabilitation program for post-stroke patients.