Automated grading of facial paralysis using the Kinect v2: A proof of concept study

Grading of facial paralysis has been the subject of numerous works in the past half century and is thus essential in order to quantify the extent of the condition and to evaluate the effect of treatment. However there is still a need for an accurate quantitative grading system that is easy to use, inexpensive and has low inter-observer variability. This work presents such a grading system based on the newly released Kinect for Windows v2 and accompanying SDK 2.0. This new system automatically detects facial landmarks and is designed to compute and grade resting facial symmetry and the extent of facial paralysis for several voluntary movements. The system also offers the option to report grading according to three widely used traditional grading systems. The system developed so far shows promising results and is a work in progress. More facial landmarks and more voluntary movements are being added to have a comprehensive quantitative grading tool which can be widely used by clinicians. With further work, it can be extended to become a virtual rehabilitation tool for facial paralysis.