This project is to present a new approach for freezing fo gait (FoG) events in Parkinson's diseas (PD) patients, by means of using combined strategies of deep-learning and signal processing. The data being used is data from real PD patients, and is composed by three tri-axial sensors (i.e. giroscope, accelerometer and magnetometer). The goal of this project is to outperform the current state of the art methods, which achive arround 85 % using shallow machine learning algorithms (i.e. SVM).

Publishe in Knowledge-Based Systems: https://doi.org/10.1016/j.knosys.2017.10.017