| Summary: | 碩士 === 慈濟大學 === 醫學資訊研究所 === 94 === Dental shade matching is an important but a difficult task in clinical practice.Traditionally, a shade guide is manually placed adjacent to the matching teeth and thenreadout the results by naked eyes. Only until recently, digital shade analysis by mean of colorimeter or spectrophotometer provides an alternative. However, the color features of the shades do not only closely tight to each other, but also exhibit overlapping, they are difficult to be classified. The successful matching rate of computerized methods only
reach about 50%, while that of naked eyes is 48%. Thus, it is not surprising that these high-technology devices, though reported to be stable in repeated matches, have been revealed no significant difference from traditional visual matching.
Consumers’ level mega-pixels digital cameras become popular. They share some
merits and dis-merits to dentists. They are convenient and helpful in roducing sharp and colorful images in clinical practices only if calibration for shade selection is not
involved. Besides, consumers’ digital cameras when exposed in different illumination
conditions can hardly produce photographs that retain color features. Furthermore,
dental photos often tell exactly the shape and the color distribution of the teeth, but not
the color itself. In order to modify a consumer digital camera so that it is helpful in
clinical shade selection, this research proposes a possible mechanism. An algorithm is
devised to normalize the illumination so that shade features are stable and retainable
when capturing under different light sources.
The objective of this research is to successfully match a content of a shade from a
shade guide that ignoring the illuminant condition using a commercialized digital
camera. The color spaces of teeth shades are being analyzed and compared. The stable
and representative color features will then be selected. New measurements have
V
established in this study to measure difference between contents of similar shades.
Different sampling methods are also discussed. The shade metrics are composed of
effective measurements, features and sampling methods. After comparing the results, it
shows that, with out proposed metrics, over 50% matching rate in top one match that is
a little bit higher than the previous research which reported about 50% by computerized
match. The top five matches have near or higher than 90% that is very helpful in clinical
application. Those measurements counting for color gradation along the contents, in
turn along the crown cervico-incisal axis, will have better matching results which means
that spatial relations within the color contents play an important role in shade matching
and cannot be neglected. It is also told that sample mean calculated from the whole
sample group is no longer useful in shade matching due to the spatial relationship of the
features that they cannot be averaged. E of Lab color space has △ s long been used for
measuring dental color difference, but is shown to be an impropriate measurement in
this study. In addition, S feature from HSV color space together with a* and b* features
from Lab color space are shown to be suitable features for shade matching in clinical
condition using digital camera.
|
|---|
