早期發現齲齒早期治療是避免細菌透過牙齒感染身體的最好方法，臨床上鄰接面齲齒不易發現，針對牙齒鄰接面齲齒的檢查，只能使用眼睛觀察或探針探測，以及配合口內根尖或咬翼X光片來判讀找出鄰接面齲齒。近年來數位影像處理科學有長足的進步，因此我們希望藉由數位影像處理中的模糊分群法，從口內所拍攝的咬翼X光影像中找出牙齒鄰接面齲齒，以提高診斷敏感度。實驗方法：將一張咬翼X光片經掃描機轉換成數位化影像，使用Matlab軟體撰寫程式。首先去除影像背景像素值，再使用模糊分群法將像素自動分群，最後有齲齒的部分用顏色顯示出來。實驗結果：初步結果可以使用顏色標示咬翼X光片齲齒的位置。未來還需測試更多X光片，修正程式找出最適合的模式。

The best way to prevent bacterial infection of the body thro ugh the teeth is early detection of dental caries and early treatment. It is difficult to find the tooth approximal caries during oral examination, Clinically only the eye observation, detection probes, or bite-wing radiograph identify proximal caries. In recent years, digital image processing of science has made great progress, so we hope that using digital image processing in the fuzzy C-mean, taken from the mouth bite-wing X-plane images to identify the tooth proximal decay in order to improve the diagnostic sensitivity. Material and Method: First one bite-wing radiograph film through the scanner transferred to the digital images. The image processing coding was written by Mathlab program. In the begging the background image pixel removed, then the fuzzy C-mean grouped together the pixels automatically by the pixel value. Finally dental caries displayed in different parts of the color. Results: Preliminary results can be the use of color illustrations of the bite-wing X-ray dental caries position. Follow-up needed to test a large number of clinical cases, to revise the program in order to identify the most suitable model. We needed to test mo re X-ray film, to amend the code to find the most suitable mode in the future.