Facial Expression Analysis with Kinect for the Diagnosis of Paralysis Using Nottingham Grading System

Abstract

Facial palsy is a neuromuscular disorder caused by damage to motor and sensory pathways of the facial nerve, which causes a movement disorder in the facial muscles. In clinical diagnosis of facial nerve palsy, simple and quantitative assessment of facial motion is strongly required. A systematic evaluation of facial nerve paralysis allows the clinician to determine objectively the severity of disability and record this information, in order to evaluate the response to therapy. The proposed method evaluates the degree of facial paralysis by using Kinect as a feature extraction tool and Nottingham grading system. The proposed grading system measure the distance between the eyebrows to infra orbital and the distance between the edges of mouth and lateral canthus. We divided the grading scale into three levels: normal, mild and severe by using the estimated parameters.