Applicability of a computer-assisted image analysis system for the evaluation of gingival status in subjects from Asian and Caucasian backgrounds

Abstract

We have recently developed a computer program to allow the automatic evaluation of the degree of gingival inflammation. The aim of the present study was to determine the efficiency of our system, which consisted of a digital camera, computer and our original program, in a sample population of various ethnic backgrounds. Digital images of frontal views of maxillary incisors were taken for 141 subjects and then subjected to computer-assisted analysis to determine the degree of redness (colorimetric assessment) and swelling (morphometric assessment) and was compared with the visual assessment of the PMA gingival score. The age of subjects ranged from 19 to 65 (Mean 33.9 SD12.8), and consisted of slightly more females (81) than males (60). Within the study group 78% of subjects classified themselves as Caucasian (European, South American, Middle Eastern), 21% were Asian or Pacific Islander, and 1% was African. The subjects were separated into 9 groups according to their PMA score. In addition, colorimetric and morphometric score were plotted against each other. The colorimetric scores ranged between 55.1 to 64.8 (arbitrary units) and morphometric scores ranged between 0.175 to 0.313 (arbitrary units). There were no ethnic differences with respect to colour and contour when two groups of European (90) and Asian (17), who were scored as 000 using the PMA index, were compared. There was no significant correlation between PMA score and morphometric score (P = 0.0711), but the correlation coefficient between PMA score and colorimetric score was 0.130 (P = 0.0122). Comparison of the highly scored image samples and the lower scored images with the visual assessment codes showed very good correlation. These preliminary findings suggest that the developed computer-assisted image analysis system shows promise for assessing gingival inflammation via an automated method.