CIE色差公式之修正於臨床牙科比色之應用

Abstract

在現代審美牙科發展下，大家愈趨重視補綴物與填補物和自然牙齒之間顏色的擬真。由於人類辨識色彩包括物理及心理兩部份，因此必須以數字化方式呈現，才能把誤差減至最低，同時也便於傳遞訊息。顏色數字化除了可以標定顏色，還可以計算兩種顏色之間的差異程度。目前於色彩學領域中主要使用的色差公式包括CIEL*a*b* (CIELAB色彩空間)以及以CIELAB色彩空間為基礎推導而成的CMC(l:c)、CIE94、CIEDE2000等公式，而各產業則根據主要使用的材料與照明環境各自規定了適用的色差公式。目前牙科領域主要使用CIEL*a*b*公式，然而此公式已被證實有修正之必要性。 本實驗目的為修正色差公式，使色差公式計算出的色差值大小可以與人眼感受到的色差大小更加一致，並且當數位比色儀使用新的色差公式選取出的最佳匹配顏色(色差值最小) 能在臨床上得到更好的結果。實驗中利用市售牙科專用比色板 (Vitapan Classical and Vita Linearguide 3D-Master shade tabs)，先透過數位比色儀 (Vita Easyshade Advanced 4.0)取得每支比色板的色彩資訊，再以不同色差值與不同色差參數大小配成一組兩支，共30組比色板配對組別。接著請22位受測者在相關色溫D65及照度1000lux的對色燈箱，根據感受到的色差大小給予評價0-5級 (0級色差最小，5級色差最大)，最後綜合所有受測者的評分與個別比色板組別的不同公式色差值進行相關性分析。 實驗結果顯示，22位受測者針對30組比色板配對組別感受到的顏色差異大小，與使用上述四種色差公式中的任一種計算出的色差值，都沒有一致的趨勢 ，相關係數r2皆小於0.4。然而，把公式與受測者評分進行擬合處理，相關係數r2 最高可提升至0.72，即可得到與受測者感受評分更一致的趨勢。除此之外，從各公式進行擬合處理後的結果發現，無論在哪一個公式架構中，ΔL^*下方的權重值都是三個參數權重值中最大的，表示 ΔL^* 在以CIELAB為基礎的所有色差公式中皆為影響性最弱的參數，Δa^* 則是影響性最強的參數；若比較ΔL^*、ΔC^*、ΔH^*，ΔL^*仍是影響性最弱，ΔH^*影響性最強。因此，臨床上使用傳統比色法時，建議先選擇明度、再選擇彩度、最後選擇色相的順序，在色差範圍縮小的情況下需視情況調整。

In the art of esthetic dentistry, precise shade matching is one of the key factors affecting the final result of dental restorations. Shade or color is the combination process of physical and psychological systems. Numerical processing of shade is essential for notation and accounting color differences. In colorimetry, there are some color-difference formulas, such as CIEL*a*b* and CIELAB-based color-difference formulas CMC(l:c)、CIE94、CIEDE2000, used independently by different industries. In dentistry, we usually use CIEL*a*b*, that is not a uniform color space and needed to be modified. The purpose of this study was to try to modify these color-difference formulas. We want to get more consistency between the color-difference value and perceptibility by human. And also, we want to use the new formula to get better matching shade (the least color-difference value) for clinical usage. The Vitapan Classical and Vita Linearguide 3D-Master dental shade tabs was used to get shade matching pairs. 2 shade tabs per pair and 30 pairs totally. 22 subjects mark scores for 30 shade pairs separately under a view booth with D65 and 1000lux condition. Integrate all the scores and analysis them with different color-difference formulas. Experimental results show that the current color-difference formulas are inconsistent with perceptibility by human. The Adj. R-square are all smaller than 0.4. While we carry out formulas fitting procedure, the Adj. R-square raises to 0.72 in highest. Additionally, in formula fitting procedure, the weighting value of ΔL^* is always the highest one. It means that in all the CIELAB-based color-difference formulas, ΔL^* is the least important parameter, and Δa^* is the most important one. Comparing ΔL^*、ΔC^* and ΔH^*, ΔL^* is still the least important parameter and ΔH^* is the most important one. Accordingly, in conventional shade matching order clinically, value is the first, chroma is the second and hue is the last one. However, some modifications are needed for small color difference cases.