In this project, we develop an objective, learning-based, application for the diagnosis and grading of facial nerve palsy (FNP), a facial movement dysfunction effecting functional, aesthetic and psychological disability. Grading the severity of FNP has prognostic and follow-up signiﬁcance. Traditional grading systems are based on subjective physician observation, and are prone to inter and intra observer variability. In MobileMed (Nov. 2014, Prague) VISL students presented a proof of concept of a mobile application for FNP diagnosis, based on tracking facial points marked by paper stickers. Now we are working on an enhanced application, based on state-of-the-art methods for tracking facial feature points. Combining color and depth images and making advantage of INTEL’s RealSense™ camera is being explored.
The project is a collaboration with Dr. Ofer Azoulay (KAPLAN Medical Center).