Several indices have been developed that use bleeding and color changes as indicators of early gingival pathology. In the presence of gingivitis, vascular proliferation and reduction of keratinization owing to increase redness in gingiva. Descriptions of healthy gingiva are numerous, ranging from pale pink and coral pink to deep red and violet. This terms are not objective.

Because of perception of color depends on a lot of factors such as light source, object, observer and so on. It is difficult to make an objective expression. Therefore the using of mechanical equipment is recommended to exclude these variables and observer's vias. The purpose of this study was to evaluate gingival color change after scaling & subgingival root planing. The other purpose of this study was to research the correlation of pocket depth, P.B.I. score and gingival color change. After photo-taking and storaging the image of gingival color into a computer, color change was examine with an image analysis program.

Results were as follow;

Color of healed gingiva after scaling & subgingival root planing was significantly differ from color of inflamed gingiva(p<0.01).

Color of healed gingiva after scaling was similar to color of healed gingiva after subgingival root planing(p<0.05).

There was statistically significant correlation between color change of red component and pocket depth after scaling & subgingival root planing(p<0.01)

There was no correlation between color change of green, blue component and pocket depth after scaling & subgingival root planing(p<0.01)

There was statistically significant correlation between between color change of red component and P.B.I. score after scaling & subgingival root planing(p<0.01).

There was no correlation between color changes of green, blue component and P.B.I. score after scaling & subgingival root planing(p<0.01)

Increase of pocket depth and P.B.I. score were significantly correlated to the amount of color change(p<0.01).

P.B.I. score had a higher correlation with color change than pocket depth(p<0.01).