Evaluation of Asymmetry in Facial Palsy Images by Generating Facial Key Points and Contours

Abstract

Facial Asymmetry is an effect of several diseases that are infected due to neural disorders. This has been growing rapidly among people of all age groups irrespective of earlier medical history and factors. The patients affected by this are unable to contract their facial nerve which results in droopy faces and deformation. Due to this many patients are also affected psychologically as this creates a phobia and discomfort for expressing themselves. As the facial expressions play a vital role for direct communication the social life of those people is been adverse. Recent study also shows that facial deformation is also an early sign of having strokes, so detecting any changes or deformation in facial features is desirable, which helps in detecting strokes beforehand. The advancement in technology helps in providing such a system for evaluating the facial images with computer vision methods. In this work, the facial images were analysed and key points of essential parts of face generated using the media pipe framework, which were considered for evaluating the asymmetry or droopy faces by comparison using Partial Curve Mapping (PCM) and other mathematical formulae. This assessment gives the level of asymmetry and can be further used for future diagnosis.