Initial assessment of facial nerve paralysis based on motion analysis using optical flow method

• Main Author: Wan Syahirah, W. Samsudin
• Other Authors: Associate Professor Ir. Dr. Kenneth Sundaraj
• Format: Thesis
• Language: English
• Published: Universiti Malaysia Perlis (UniMAP) 2019
• Subjects: Facial nerve paralysis; Facial paralysis; Motion analysis; Rehabilitation; Facial nerve
• Online Access: http://dspace.unimap.edu.my:80/xmlui/handle/123456789/61990
• Institution: Universiti Malaysia Perlis

Patients with facial nerve paralysis suffer serious functional, cosmetic and psychological problems with impaired ability to communicate both verbally and nonverbally. Rehabilitation for facial paralysis begins with a thorough clinical evaluation in accessing the degree of paralysis. Generally, for assessment of facial nerve dysfunction on patient’s face in daily clinical application, observation-based subjective grading systems by clinicians is been employed. Subjective assessments of facial nerve function are accomplished by inferring nerve function through the observation of voluntary facial movement. These methods usually yield a single value which hypothetically corresponds to the severity of facial paralysis. To date, there are no subjective assessment methods that can reliably produce quantitative information regarding facial nerve function. Previous works have proposed a wide range of methods to obtain objective data to quantify and determine the severity of facial paralysis. However, there is no standardized quantitative method for objective description of facial nerve function. Acknowledging the importance of this facial assessment, this research was conducted to develop an initial assessment method that can classify normal and patient subjects based on facial motion analysis using Optical Flow algorithm as well as categorizing the severity of patient into six levels according to House-Brackmann (HB) system. Prior to initial assessment method been developed, two experiments were conducted to find the best parameter and best measurement for assessing facial paralysis. Distance and area were selected as two measurement parameters which were investigated in this research. A number of mathematical and statistical analyses were performed to determine the best measurement of these parameters. The results indicate that the distance is the best parameter and the value of initial exercise frame is the most important measurement in tracking the changes of facial movement. This thesis also presents the initial facial assessment method which contains the individual scores for each exercise involved in this research and also grading the paralysis for each patient corresponds to HB system. This research showed satisfactory results which was validated by a medical professional in Otorhinolaryngology. This initial facial assessment method may play a pivotal role in initiation of final assessment system for facial paralysis and develop better rehabilitation program for patients.