This work presents a low-cost multimodal system that combine the use of serious video games for health and the analysis of biomedical signals with the purpose of optimizing the processes of physical rehabilitation and neurorehabilitation in Colombia. Portable and non-invasive sensors are used to register the motion capture data and the neurophysiological signals taken using brain computer interface (BCI). The final system is called brain-kinect interface and has been used with success in patients with multiple motor neuron lesions. The cinematic signal processing is done using a specialized software developed to synthetize the information in the files of motion captured using the Kinect sensor and to provide angular graphs; also the open source software OpenViBe is used to process the EEG signal using the BCI Emotiv EPOC system. Preliminary results show significant improvement in the therapies' rehabilitation adherence of the evaluated patients, quantitative results show improvement of more than 15% in the motion ranges, showing the potential of using video games as a complementary tool to the clinical practice.