The Application of Color Science to Dental Color Matching

The Application of Color Science to Dental Color Matching

REVIEW OF THE LITERATURE
A dental shade guide is the commonly accepted method of evaluating the colour of natural teeth to provide an artificial substitute (Preston 1985). However the use of such guides has proved extremely inaccurate and frustrating (Culpepper 1970).

Colour matching is difficult and the results inadequate (Clark 1931, Sproull 1973, McLean 1979, Preston and Bergen 1980, Preston 1985).

Dentists are not equipped to carry out the difficult task of colour selection (Clark 1931), nor do they understand how to alter colour to achieve a colour match (Preston and Bergen 1980) as they do not have a basic education in this field (Sproull 1974). The use of unsuitable shade guides (Preston 1985, Seluk and LaLonde 1985) for colour selection adds to the problems making consistent colour matching impossible. Existing shade guides are unsuitable (Sproull 1976) as the shades do not conform to tooth colour (Sproull 1973, Lemire and Burk 1975) and there is a lack of any organised distribution of the samples (Sproull 1973, McLean 1979, Preston 1985). As well they are constructed in different materials (Preston 1985) and the colours do not match the restorative material (Preston 1985).

An explanation of the lack of progress in colour selection is in the complexity of colour and of the dental colour matching problem. The introduction to the chapter on colour in the widely used text “Psychology of Visual Perception” (Haber and Hershenson 1974) commences with “Color is an impossibly large and complex topic.” It indicates that this very familiar and apparently simple subject is deceptively complex.

In dentistry there has been insufficient knowledge to understand the problem, or to comprehend its complexity. “It is probable that the enormity of the problem is not understood” (Preston 1985).

“Color science has not yet been successfully applied to dental colour matching.” (Clarke 1983) This article on dental colour problems was instigated by Dr. John McLean and the British Ceramic Research Association. The only previous paper by a colour scientist on any dental subject (opacity of silicate cements) was over 50 years ago (Judd 1937). Clarke (1983) recognised the lack of understanding of colour science within dentistry and the complexity of tooth colour measurement. His suggestion was a tooth colour atlas for colour measurement and matching.

PROBLEMS OF COLOR MATCHING
The whole of colour contains at least ten million detectably different colours (Judd and Wyszecki 1975). Tooth colour is much smaller yet still contains over 100,000 different colours. This is a large number of possible tooth colours compared with the limited number of samples available in dental shade guides. Any shade guide sample, compared with a tooth, has only one chance in 100,000 of achieving a colour match. This assumes there is an even distribution of tooth colour and each sample contains only one colour.

If each sample contains two colours both have to match the tooth and the chance reduces to one in 1010! As porcelain shade guide teeth normally contain five colours, opaque, body (dentine), enamel, gingival and effects, the chance is now one in 1025, which for all practical purposes is nil. It can be deduced that dental shade guides were never designed for the task they are supposed to perform.

The elimination of multiple colours from a shade guide increases its effectiveness from nil back to one in 100,000. This is illustrated effectively by experienced ceramic technicians using ‘key ring’ guides of single colour samples to match selected areas of the tooth. From practice they realise U is much more effective even though the guides are cumbersome and time consuming to use. They contain the entire colour range manufactured by each company, typically 50 colours.

To have any realistic chance of obtaining a colour match the number of samples would have to be increased to many thousands with 50 probably being an indication of the practical limit for a manufacturer.

Among the problems with existing shade guides identified in the literature were:

  1. Shades do not conform to tooth colour (Sproull 1973, Lemire and Burk 1975)
  2. There are insufficient samples to cover tooth color space
  3. There is a lack of organisation of the samples (Sproull 1973, McLean 1979, Preston and Bergan 1980)
  4. Shades are seldom constructed from the restorative material (Preston 1985)
  5. Samples do not match the restorative material (Preston 1985).

The last two can be rectified by constructing the color guide in the restorative material but the other problems remain.

CONFORMITY TO TOOTH COLOR Colour is three dimensional and for the gamut of colour to be described or represented requires a three dimensional space. The manner of arranging the colours in that space is prescribed by a colour order system such as the Munsell system (Munsell 1905).
The variables are: Value, Hue and Chroma.

Value (lightness)is represented by a vertical central achromatic core which is a series of greys extending from white at the top to black at the bottom. A horizontal cross section at any Value shows the Hues radiating from the central core forming the familiar colour wheel of the strongest colours at the periphery. The Chroma (strength) decreases toward the core (Sproull 1973, Billmeyer and Saltzman 1981).

Figure 1shows the recorded spaces of existing shade guides (Preston and Bergen 1980) superimposed on recorded tooth colour (Sproull 1973) in Munsell. The restricted coverage is immediately obvious. The amount of coverage of tooth colour by each guide can be quantified simply by estimating the amounts of coverage in each of the three variables and multiplying to obtain an approximate 3-dimensional coverage. For Vita,* Hue coverage is 30 percent, Value 50 percent and Chroma 40 percent. Bioform Hue is 35 percent, Value 55 percent and Chroma 63 percent.


Figure 1

The three dimensional coverage of Vita (0.3 x 0.5 x 0.4) is six percent, and Bioform† is 12 percent, with the total coverage by both only 14 percent. These calculations are only approximate and overestimate the actual space occupied by at least 25 percent because it ignores the rounded and irregular nature of the spaces in 3-dimensions.

This represents very poor coverage of the represented tooth colour. At least 94 percent of tooth colour is not covered at all by the Vita samples, and 86 percent is not covered by any samples. Colour matching of the natural teeth in these regions is absolutely impossible with these shades.

* Vita Zahnfabrik, H Rauter GmbH & Co. KG, Bad Sackingen, Germany
† Caulk Dentsply Int Inc., York PA, USA

COLOR ATLASES
A tooth color atlas has been suggested as the logical answer to selecting tooth colour (Clarke 1983). A colour atlas is a systematic arrangement of colour samples usually for the whole of colour, spaced regularly according to stated criteria (Billmeyer and Saltzman 1981). The three dimensional arrangement criteria is specified by the colour order system.

Color atlases have been constructed for realisable colour as surface colours (paint samples) arranged for easy reference. Any colour can be compared with the painted samples and identified in relation to them. This colour can then be described in these terms to anyone with a duplicate colour atlas. The various atlases have different arrangements in three dimensions, and the spacing is dependent on the method of construction, such as subtractive pigment mixing.

The ideal arrangement for accurate colour measurement is based on perceptually uniform spacing of the samples as used in the universally accepted Munsell system. It is organised so that each sample is surrounded by other samples in three directions, with all samples spaced an equal colour distance apart.

To classify a colour the nearest sample in the Munsell Book of Color is located and this has a Munsell Notation of three numbers representing Hue, Value and Chroma. The object colour can be located exactly in relation to the surrounding colours in each of the three directions and these numbers interpolated and recorded. The samples also have known specification in colorimetrics such as CIELAB (CIE 1978) where the object colour can be recorded and compared with measured colour.

TOOTH COLOR ATLAS
Tooth color can be specified in the same manner by a tooth colour atlas arranged in a colour order system designed specifically for tooth colour (Clarke 1983). In a tooth colour atlas with perceptually even spacing, the object tooth colour will always occur between two samples which are a known colour distance apart, allowing a simple judgement of the correct mix of the sample colours required. This interpolation is known as the ratio method (Newhall 1939) which allows the eye to perform efficiently within its capability without any additional ability or training required. In contrast, the normal dental method of estimating the colour distance from a single shade sample with no other shade to compare with the tooth, is a judgement beyond the capability of the observer (Billmeyer and Saltzman 1981).

A perceptually uniform tooth colour atlas would indicate the nearest sample to the tooth colour and by interpolation, the mixture of the next nearest colour or colours required to obtain an exact match (Clarke 1983). This would increase the chance of an exact match from the previous one in 100,000, to one in one (0.001 - 100 percent).

COLOR ORDER SYSTEM
A system for representing colour is required for tooth colour. A tooth colour atlas can then be produced as coloured samples organised according to this specified arrangement.

Colour, perceived when light reflected from an object enters the eye, consists of quantity and quality (Clarke 1983).

Quantity is the variable lightness (value) and is the total amount of light reflected from an object. It is indicated by the relative nearness of a colour to white or black on an achromatic grey scale such as a black and white photograph. The rods in the retina of the eye respond only to lightness.

Quality is the chromatic component of colour consisting of two variables which can be represented in two dimensions. Tooth colour should be covered as efficiently as possible with the least number of samples with the accuracy dependent on the even visual spacing of the samples.

Polar coordinate systems, such as Munsell should be avoided (Judd and Wyszecki 1975) due to the inherent chromatic distortions where the distance between colours of the same hue increase as the chroma increases. Spacing the samples requires a non distorting colour space such as the more recently developed internationally accepted colorimetric for reflective materials, CIELAB (CIE 1978). It was designed for small colour differences as a cartesian coordinate system of opponent colours (Fig. 3), conforming to the accepted opponent colour theory of colour vision.

Colour vision is recorded in the retina in two stages. The first stage is the white-black (lightness) response of the rods and the three types of cones exhibiting sensitivity to red, green and blue wavelengths. The second stage is the ganglion cells transforming the chromatic responses to two channels of opponent colours, yellow-blue and red-green. The signals are transmitted to the visual cortex as white (+ve), black (-ve); yellow (+ve), blue (-ve); and red (+ve), green (-ve), representing the three variables L (white-black), a (red-green) and b (yellow-blue).

Colour can be represented in three dimensions by horizontal two-dimensional chromaticity planes spaced one above the other with lightness increasing vertically (Fig. 2).


Figure 2 - Chromatic planes in colour space

Figure 3 - Tooth colour (Hall 1988) in CIELAB

A dental colour atlas following this arrangement would enable the appropriate lightness level of the tooth to be ~61ected first. Lightness is the easiest variable to select initially. Then on closer scrutiny the instinctive use of the fovea of the retina, which contains only cones, enables the efficient selection of the chromatic differences (Clarke 1983).

ARRANGEMENT OF SAMPLES
The required dimensions of tooth colour space were established by instrumental and visual measurement (Hall 1988). In CIELAB tooth colour is contained in a relatively small shape best described as an ellipsoid (Fig. 3).

To rationalise the distribution of samples, tooth colour can be represented as cylindrical with an oval cross section. This eliminates very little of the extremities, mostly at high and low lightness, which contain a very low incidence of tooth colour. An efficient distribution would then be achieved by using an evenly spaced grid pattern, with the oval shape dictating a grid of equilateral triangles at each lightness level. Placing the material colours at the coordinates indicated by the line intersections (Fig. 4) allows the most efficient coverage of tooth colour with the minimum number of samples.

The spacing between samples dictates the number of samples. Nine samples on each of three lightness levels (Fig. 3) creates a chromatic spacing of four CIELAB units and a lightness spacing of eight, which is in proportion to the relative sensitivity to chromatic and lightness differences. This establishes three layers each of nine samples, a total number of 27 samples, which is a feasible number to organise into a functioning tooth colour guide. The accuracy and repeatability of the measurement of tooth colour depends on the spacing of the samples. Up to four visible steps are discernible between two samples spaced correctly (Judd and Wyszecki 1975). Viewing tooth colour between two samples spaced four units apart allows an accuracy of one unit which was intended to represent the discernible colour spacing in CIELAB.

The ideal is for each sample to be surrounded by other samples in three directions representing the three variables. The resultant arrangement has each internal sample surrounded by six samples equidistant from it on each chromatic plane. Even the peripheral samples at the chroma extremes (LR1, LY3, DR1 and DR3, Figs 3 and 4) have three samples equally spaced from them, two on their own lightness level and one on the Medium level.

The variables of tooth colour are lightness (lighter and darker), hue (more yellow and red) and chroma (weaker and stronger). Lighter, darker, yellow, red, weaker and stronger are readily understandable terms with international approval (Hunt 1977).


Figure 4 - Arrangement of colour samples on each chromatic plane.

Figure 5 - The three indicators of the colour guide.

TOOTH COLOR GUIDE
The samples need to be organised in a workable configuration to simplify clinical use.

The 27 colour samples have three levels of lightness (Light, Medium and Dark) with nine samples on each chromatic plane. Each group of nine samples is assembled in a holder forming a tooth colour indicator for that level of lightness. When the three indicators are held by the base horizontally above one another, with the lightest on top, they form the tooth colour guide (atlas) with the samples in their relative positions in tooth colour space.

Each indicator divides readily into three sections of three (Fig. 5), (more) yellow, central, and (more) red. Each group of three samples is attached to one arm which allows the comparison of tooth colour to more than one sample making accurate colour selection possible. The samples of each group (Y, C and R) on each arm are numbered one to three in ascending chroma i.e. weaker to stronger, identifying the colours (Fig. 4).

Selecting the lightness (L, M or D) reduces the choice to the nine samples of that lightness level. The appropriate group (yellow, central, or red) is selected reducing the selection quickly to which one of those three samples is closest to the tooth colour.

The configuration of the samples make them very manageable and easy to compare with teeth. The selection becomes a logical progression of three simplified choices.

CONSTRUCTION
The samples were produced from pigmented metal ceramic porcelains with each of the 27 colours having an opaque carefully matched to it so that the colour remained constant with varying thicknesses of dentine over opaque. Each sample was one millimeter of dentine porcelain over 0.5 min of opaque giving an accurate indication of the achievable colour of metal ceramic crowns.

In practice the colour indicators of the colour atlas were very practical with the organisation of the indicators simplifying their use.

Decreasing the chromatic spacing of the samples to less than three units produced confusion. The spacing needed to be at least four units to allow the interpolation of three intermediate colours and then observers obtained accurate repeatable results quickly.

Intermediate lightness levels were tried but also produced confusion. They were seldom required and when necessary were easily interpolated from the light, medium and dark lightness indicators.

The tooth colour indicators were used extensively in tooth measurement to obtain distribution data on tooth colour.

CONCLUSION
Existing shade guides have been shown to be inefficient and incapable of aiding accurate colour selection (Culpepper 1970).

The existing approach to colour selection is extremely primitive, having not changed in principle for 200 years. Shade guides offer a number of unrelated shades based on popularly selected porcelain teeth which have been shown to be mostly outside tooth colour. The shade guides only indicate the available material shades.

The system developed is fundamentally different with the colour guide indicating the material colours necessary to accurately reproduce tooth colour. The restorative material is then produced in these colours.

The application of colour science resulted in the logical organisation of the colours to allow the minimum number to most effectively cover tooth colour. The system allows the quick accurate selection of tooth colour without specialised knowledge or training. The system is self explanatory with the arrangement of the samples facilitating quick easy use. A tooth colour guide or colour atlas has been constructed which is capable of accurate tooth colour measurement. The accumulation of data will make possible the greater understanding of in vivo tooth colour and its distribution. It also offers the potential of precise colour matching with restorative materials produced in these colours.

REFERENCES

  • Billmeyer FW, Saltzman M. Principles of color technology. New York: J Wiley Sons, 1981.
  • CIE. Commission Internationale de l’Eclairage. Recommendations on uniform color spaces, colordifference equations, psychometric color terms, Supplement No.2 to CIE Publication No.15, (E-1.3.1) 19719 (TC-1.3) 1978, Bureau Central de la CIE, Paris, 1978. Clark EB. The color problem in dentistry. Dent Dig 1931; 37: 499-509.
  • Clark EB. An analysis of tooth color. J Am Dent A, 193 1; 18: 2 093.
  • Clark EB. The Clark tooth colour system, Part 3. Dent Mag Oral Top 1933; 50: 249-258.
  • Clarke FJJ. Measurement of colour of human teeth. In: McLean JW ed. Dental Ceramics: Proceedings of the First International Symposium on Ceramics. Chicago: Quintessence Publishing Co Inc. 1983: 441-488.
  • Culpepper WD. A comparitive study of shade-matching procedures. J Pros Dent 1970; 24(2): 166-173.
  • Crum F, Bartleson CJ eds. Optical radiation measurements, Vol 2. Colour measurement. New York: Academic Press, 1980.
  • Haber RN, Hershenson M. The psychology of visual perception. London: Holt Rinehart, 1974: 60.
  • Hall NR. Unpublished data. (MDS thesis) 1988.
  • Hall NR. Tooth colour selection. Aust Pros J 1991; 5: 41-46.
  • Hunt RWG. The specification of colour appearance. 1. Concepts and terms. Color Res Appl 1977; 2(2): 55-68.
  • Judd DB. Optical specification of lights-cattering materials. J Res NBS 1937; 19:287-317.
  • Judd DB, Wyszecki G. Color in Business Science and Industry, 3rd ed. New York: J Wiley Sons, 1975: 353.
  • Lemire PA, Burk AAS. Color in dentistry. Hartford: JM Ney Co, 1975.
  • McLean JW. The science and art of dental ceramics, Vol 1: The nature of dental ceramics and their clinical use. Chicago: Quintessence Publishing Co Inc, 1979.
  • Miller L. Round table discussion. In: McLean JW` ed. Dental Ceramics: Proceedings of the First International
  • Symposium on Ceramics. Chicago: Quintessence Publishing Co Inc. 1983: 523-524.
  • Munsell AH. A color notation, 1st ed. Baltimore: Munsell Color Co, 1905.
  • Newhall SM. The ratio method in the review of the Munsell colors. J Am J Psycholl939;52:394.
  • Preston JD, Bergen SF. Color science and dental art. St. Louis: CV Mosby, 1980.
  • Preston JD. Current status of shade selection and color matching. Quint Int 1985; 1: 47-58.
  • Ruyter IE et al. Colour stability of composite resins for C&B veneers. Dent Mater 1987; 3: 246-251.
  • Seluk WL, LaLonde TD. Esthetics and communication with a custom shade guide. Dent Clin Nth Am 1985; 29(4): 741-51.
  • Sproull RC. Color matching in dentistry, Part I. J Pros Dent 1973; 29: 416-424.
  • Sproull RC. Color matching in dentistry. Part II. i Pros Dent 1973; 29: 556-566.
  • Sproull RC. Color matching in dentistry. Part III. J Pros Dent 1974; 31: 146-154.
  • Sproull RC. Color matching problems in dentistry. In: ISCC Subcommittee for Problem 35 Report. Color Res Appl 1976; 1(3):134-136.
  • Wyszecki G, Stiles WS. Colour science - concepts and methods, quantitative data and formulas. New York: J Wiley Sons, 1982.
  • <<
  • >>

Comments

-->