[Computerized analysis of facial motions--objective evaluation of facial palsy].

The severity of facial palsy is often diagnosed using a scoring system. However, there have been some problems as to the objectivity and reproducibility of evaluations by this method. To solve these problems, a variety of new methods have been introduced. Compared to the simple scoring method, none of these methods seem to have gained wide-spread application in routine clinical practice. The development of our diagnostic system was inspired by the idea that numerical quantitation of shifting of marks on the face might be possible by means of digital image-processing. Using this method, only marks adhering to the face are extracted, and the shifting position of these marks is quantitatively analyzed. Detailed analysis of the facial movement requires full motion video images. We found that such video images were easily obtainable by application of a Quick Time Movie developed by Apple computer. Twelve healthy subjects served as normal controls. Seven patients were analysed as cases of facial palsy. The facial movements examined in this study consisted of eye-closing and whistling. A total of 20 black marks were used for adherence to the face. The image processing called thresholding leaves only black marks on the face and has made it possible to trace and measure these marks on the face. The shifting positions of the marks moving in accordance with facial movement, from the stationary phase until the maximum stage, were measured and traces were plotted on the coordinate axis. The shifting of the marks adhering to the face was numerically expressed as a trajectory investigation.(ABSTRACT TRUNCATED AT 250 WORDS)