====== Conclusions ====== The analysis of gait and the detection of the motor symptoms of a PWP are challenging problems, especially if they are to be done through the use of the motion and position sensors embedded in a smartphone, even if the measured values are improved through the use of sensor fusion. In the scope of this project, the step detector and step length estimator, when calibrated, have achieved, respectively, a step detector had a mean absolute percentage error of 1.85% and the step length estimator, when calibrated, had a mean absolute percentage error of 5.26%. Although a thorough analysis of the sound module wasn't possible, the preliminary results suggest that the developed sound synthesis and music synthesis submodules seem to be promising alternatives to the auditory cueing typically employed in the rehabilitation of the gait of PWP. During the tests with the PWP, a dataset of sensor and video data of 7384 steps, which corresponds to 66 minutes of PWP walking, was collected and validated. This dataset will be valuable for the development of possible future work, given that no such dataset, as far the study of the state of the art allowed to assess, exists, at least in the public domain. The biggest innovation of this project was the development of an Android application which generates synthesized sounds and synthesized music, according to the temporal gait parameters of the PWP, in real-time while the PWP is walking, which as far the study of the state of the art allowed to assess, such characteristics in an Android application didn't exist until the development of this project. ====== Future Work ====== This project could be continued, developed and enhanced further in several ways: * Using the collected dataset to continue to investigate the effect that the different types of auditory stimuli had on the PWP. * Expand the acquired dataset with more instances of PWP’s suffering from bradykinesia or even other symptoms in order to expand the diagnosis of symptoms. * Implement a new method for the estimation of the step length that works well for all types of subjects, especially PWP’s that are in the OFF motor state. * Extend the application in order to use sensors outside the smartphone in addition to the sensors embedded in the smartphone itself. * In order to better evaluate the performance of the application’s modules, perform more rigorous tests with stricter protocols and in environments with better conditions.