Abstract
Various methods have been reported to evaluate the degree of facial palsy. They have, however, some problems concerning their objectivity and reproducibility, because they are based on the subjective assessment. In 1989 one of co-authors, Y. Naito, reported a new method to objectively quantitate the degree of facial palsy using computerized imagedefference system.
The purpose of this study is to devise this new objective method for the clinical evaluation of facial palsy. The patient's face is recorded on a video tape with a CCD camera. The video data are digitized with a video-digitizer (CVI 98 II, Canopus) and analyzed by personal computer (PC 9801XL2, NEC) with a frame-buffer (Hyper-Frame+, Digitalarts). The image of the face in motion, eye closing and grinning, is then subtracted from the image at rest, and the white pixels that indicate the difference of the two images are counted. The degree of the palsy is indicated by the ratio of the numbers of pixels on the paralyzed and normal sides.
Eight patients with facial palsy and 20 normal controls were analyzed by this method.
The ratio in the control group was very close to 1:1, while in the facial palsy patients it reflected the degree of the facial palsy. The data obtained with this method was more reprodusible than the facial scores of the same subjects tested with 40-points grading system of Japan.
This new method is considered to be objective and reliable for the evaluation of facial palsy.
