A novel approach to quantify palpebral fissure morphology

Purpose. To use image processing routines to acquire precise information on the palpebral fissure morphology. Methods. A CCD camera with 100 mm 1:34 zoom lenses and a Power Mackintosh equipped with a Scion frame-grabber were used to acquire palpebral fissure images. The images were processed by (he NIH IMAGE 1.55 software. The software was customized for contrast enhancement and to accept the user interaction in order to define a set of points delineating the eyelid contour. An automatic procedure to define landmarks was accomplished by image binarization using thresholding followed by a line thinning procedure. Results. The upper eyelid contour is well adjusted by a second degree polynomial function (ax2 + bx + c). The lower eyelid contour is more complex and well adjusted by a third degree polynomial function (ax' + bx: + ex - d). A full analysis of these data is presented. Conclusions. A new approach to standardize image acquisition for palpebral fissure parameters quantification was developed with encouraging results. The eyelid contours have been fitted to well-known curves with a spatial resolution of 0.07mm. The results show that this approach could useful for clinical purposes.