Recognition of Postures and Freezing of Gait in Parkinson’s Disease Patients Using Microsoft Kinect Sensor

Abstract

Freezing of Gait (FOG) is a disabling symptom and movement disorder, typically associated with the latter stages of Parkinson’s disease. In this paper, we propose a novel approach for real-time FOG, tremor monitoring and fall detection, consisting of a 3D camera sensor based on the Microsoft Kinect architecture. The system is capable of recognizing freezing episodes (FOG) in a standstill state, tremors and fall incidents, commonly seen in Parkinson’s disease patients. In case of an incident, it automatically alerts relatives and healthcare providers. The system was tested on seven simulated subjects in 12 events indicating that the design was able to detect 99% of the falling incidents, 91% of tremor and 92% of the freezing of gait episodes with an average latency of 300 milliseconds. The performance of the system can be further improved with the deployment of the recently released version of Kinect, capable of providing even higher levels of accuracy.