台灣部分人口由於有嚼檳榔的習慣，因此口腔癌的發生率比其他國家還要高。為了要降低口腔癌死亡率，口腔癌篩檢是一個非常重要的議題。目前口腔癌篩檢法中，病理切片是最重要的依據。然而病理切片會造成病人不適，為了避免讓病人感到疼痛，我們希望能藉由自體螢光影像來輔助口腔癌檢測，並且分析其可行性。
自體螢光由組織基質和細胞內的螢光物質產生。在癌症發展的階段，組織結構還有細胞代謝的改變，會讓螢光物質或光學特性產生變化。我們研究主要針對兩個自體螢光物質：還原態菸鹼醯胺腺嘌呤雙核苷酸 (NADH) 以及黃素腺嘌呤二核苷酸 (FAD)，觀察其在口腔癌的變化。除此之外我們將NADH螢光訊號作為分子，NADH以及FAD的螢光訊號總和作為分母，相除後計算出Redox Ratio，用來代表組織的代謝情況。並且我們採用灰階共生矩陣方法，分析NADH、FAD螢光影像還有Redox Raio紋理上的變化。我們採用倉鼠作為研究對象，將11隻倉鼠注射倉鼠腫瘤細胞作為癌症觀測對象；另外將36隻倉鼠塗抹致癌劑，作為癌前病變到癌症的觀測對象。
實驗結果發現，NADH和FAD在癌前病變還有腫瘤處的螢光訊號都比正常組織低；而Redox Ratio的計算結果，腫瘤則比正常組織來得高。紋理分析的結果發現，正常組織、癌前病變和腫瘤區域的一些紋理特徵，亦有所差異。接著我們使用支援向量機 (support vector machine)利用分析過後的特徵做影像分類，可以辨別出正常和病變影像。在倉鼠癌症實驗中，敏感度(sensitivity)和特異度(specificity)分別為82.78%和90.51%。而在倉鼠癌前病變到癌症的實驗中，將癌前病變和癌症視為有問題的樣本，良性增生做為正常樣本，結果敏感度和特異度分別高達92.83%和91.95%。因此將自體螢光影像，作為非侵入式口腔癌檢測法，具有一定潛力。

The incidence rate of oral cancer in Taiwan is higher than other countries due to betel nut chewing. Therefore, it makes oral cancer screening an important issue to reduce mortality rate. Current oral cancer screening method is by utilizing tissue biopsy. However, this method is time consuming and causes patients feeling uncomfortable and even painful. To overcome this situation, we proposed a non-invasive, faster method by using autofluorescence image analysis to assist oral cancer screening.
Autofluorescence is produced by fluorophores in tissue matrix or living cells. It varies during cancer development due to tissue structural changes and metabolism differences. Our targeting fluorophores are NADH and FAD. In this research, a metabolic indicator named Redox Ratio, which is the autofluorescence signal of NADH divided by the sum of NADH and FAD autofluorescence signals, is considered as well. For texture features extraction, Gray Level Coocurrence Matrix (GLCM) was applied to autofluorescence and Redox Ratio images. Hamster was chosen to be our animal model. Eleven hamsters were injected by hamster squamous-cell carcinoma cell line HCDB1 and observed their cancer development. Thirty-six hamsters were treated by carcinogen DMBA and observed their precancerous to cancer development.
Our research shows that, both NADH and FAD signals from precancerous and tumor regions are lower than normal tissue. After calculating the pixel-wise Redox Ratio, the value in tumor region is higher than normal region as we expected. Furthermore, analyzing autofluorescence and redox ratio images respectively; Texture difference between normal and tumor tissues also exists. By using all features mentioned above for SVM classification, the sensitivity and specificity of cancer animal model are 82.78% and 90.51% respectively. And the sensitivity and specificity of precancerous animal model are 92.83% and 91.95% respectively. These results indicate that autofluorescence image has its potential for developing non-invasive oral cancer screening.

[1] S. Warnakulasuriya, "Global epidemiology of oral and oropharyngeal cancer," Oral Oncol., p. 309–316, 2009.
[2] J. A. Woolgar, S. Rogers, C. R. West, R. D. Errington, J. S. Brown and E. D. Vaughan, "Survival and patterns of recurrence in 200 oral cancer patients treated by radical surgery and neck dissection," Oral Oncol., p. 257–265, 1999.
[3] M. W. Lingen, J. R. Kalmar, T. Karrison and P. M. Speight, "Critical evaluation of diagnostic aids for the detection of oral cancer," Oral Oncol., p. 10–22, January 2008.
[4] J. E. Bouquot, P. Suarez and N. Vigneswaran, "oral precancer and early cancer detection in the dental office–Review of New Technologies," The Journal of Implant & Advanced Clinical Dentistry, pp. 47-63, April 2010.
[5] M. A. Onofre, M. R. Sposto and M. C. Navarro, "Reliability of toluidine blue application in the detection of oral epithelial dysplasia and in situ and invasive squamous cell carcinomas," Oral Surg Oral Med Oral Pathol Oral Radiol Endod., pp. 535-40, May 2001.
[6] T. J. Potter, D. Summerlin and J. H. Campbell, "Oral malignancies associated with negative transepithelial brush biopsy," J Oral Maxillofac Surg, pp. 674-677, 2003.
[7] D. C. De Veld, M. J. Witjes, H. J. Sterenborg and J. L. Roodenburg, "The status of in vivo autofluorescence spectroscopy and imaging for oral oncology," Oral Oncol., p. 117–131, July 2005.
[8] N. Kollias, G. Zonios and G. N. Stamatas, "Fluorescence spectroscopy of skin," Vibrational Spectroscopy, pp. 17-23, 2002.
[9] G. T. Thomas, M. P. Lewis and P. M. Speight, "Matrix metalloproteinases and oral cancer," Oral Oncol., p. 227–233, 1999.
[10] J. H. Ostrander, C. M. McMahon, S. Lem, S. R. Millon, J. Q. Brown, V. L. Seewaldt and N. Ramanujam, "Optical redox ratio differentiates breast cancer cell lines based on estrogen receptor status," Cancer Res, pp. 4759-4766, 1 June 2010.
[11] R. M. Haralick, K. Shanmugam and I. Dinstein, "Textural features for image classification," IEEE Systems, Man, and Cybernetics Society, pp. 610-621, November 1973.
[12] C. Cortes and V. Vapnik, "Support-Vector Networks," Machine Leaming, pp. 273-297, 1995.
[13] B. Mognetti, F. Di Carlo and G. N. Berta, "Animal models in oral cancer research," Oral Oncol., p. 448–460, 2006.
[14] A. E. Profio and D. R. Doiron, "A feasibility study of the use of fluorescence bronchoscopy for localization of small lung tumours," PHYS. MED. BIOL., pp. 949-957, 1977.
[15] K. Onizawa, N. Okamura, H. Saginoya, H. Yusa, T. Yanagawa and H. Yoshida, "Analysis of fluorescence in oral squamous cell carcinoma," Oral Oncology, vol. 38, pp. 343-348, 2002.
[16] C. S. Betz, M. Mehlmann, K. Rick, H. Stepp, G. Grevers, R. Baumgartner and A. Leunig, "Autofluorescence imaging and spectroscopy of normal and malignant mucosa in patients with head and neck cancer," Lasers in Surgery and Medicine, vol. 25, p. 323–334, 1999.
[17] P. M. Lane, T. Gilhuly, H. Zeng, C. F. Poh, P. M. Williams, L. Zhang, M. P. Rosin and C. E. MacAulay, "Simple device for the direct visualization of oral-cavity tissue fluorescence," Journal of Biomedical Optics, vol. 11, no. 2, p. 024006, March/April 2006.
[18] D. Roblyer, C. Kurachi, V. Stepanek, M. D. Williams, A. K. El-Naggar, J. J. Lee, A. M. Gillenwater and R. Richards-Kortum, "Objective detection and delineation of oral neoplasia using autofluorescence imaging," Cancer Prev Res, vol. 2, no. 5, pp. 423-431, May 2009.
[19] M. S. Rahman, N. Ingole, D. Roblyer, V. Stepanek, R. Richards-Kortum, A. Gillenwater, S. Shastri and P. Chaturvedi, "Evaluation of a low-cost, portable imaging system for early detection of oral cancer," Head & Neck Oncology, vol. 2, no. 10, 2010.
[20] D. Roblyer, C. Kurachi, V. Stepanek, R. A. Schwarz, M. D. Williams, J. J. Lee, A. M. Gillenwater and R. Richards-Kortum, "Comparison of multispectral wide-field optical imaging modalities to maximize image contrast for objective discrimination of oral neoplasia," Journal of Biomedical Optics, vol. 15, no. 6, p. 066017, November/December 2010.
[21] B. Chance, ,. B. Schoener, ,. R. Oshino, ,. F. Itshak and Y. Nakase, "Oxidation-Reduction Ratio studies of mitochondria in freeze-trapped samples," THE JOURNAL OF BIOLOGICAL CHEMISTRY, pp. 4764-4771, 10 June 1979.
[22] Q. Liu, G. Grant, J. Li, Y. Zhang, F. Hu, S. Li, C. Wilson, K. Chen, D. Bigner and T. Vo-Dinh, "Compact point-detection fluorescence spectroscopy system for quantifying intrinsic fluorescence redox ratio in brain cancer diagnostics," Journal of Biomedical Optics, March 2011.
[23] L. Z. Li, R. Zhou, H. N. Xu, L. Moon, T. Zhong, E. J. Kim, H. Qiao, R. Reddy, D. Leeper, B. Chance and J. D. Glickson, "Quantitative magnetic resonance and optical imaging biomarkers of melanoma metastatic potential," PNAS, vol. vol.106, no. 16, p. 6608–6613, April 2009.
[24] K. Cai, H. N. Xu, A. Singh, M. Haris, R. Reddy and L. Z. Li, "Characterizing prostate tumor mouse xenografts with CEST and MT-MRI and Redox scanning," Adv Exp Med Biol., pp. 39-45, 2013.
[25] H. N. Xu, S. Nioka, J. D. Glickson, B. Chance and L. Z. Li, "Quantitative mitochondrial redox imaging of breast cancer metastatic potential," Journal of Biomedical Optics, vol. Vol.15, no. 3, p. 036010, May/June 2010.
[26] J. M. Benavides, S. Chang, S. Y. Park and R. Richards-Kortum, "Multispectral digital colposcopy for in vivo detection of cervical cancer," OPTICS EXPRESS, pp. 1223-1236, 19 May 2003.
[27] G. A. Wagnieres, W. M. Star and B. C. Wilson, "ln vivo fluorescence spectroscopy and imaging for oncological applications," Photochemistry and Photobiology, pp. 603-632, 1998.
[28] Y. He, N. Khanna, C. J. Boushey and E. J. Delp, "Specular highlight removal for image-based dietary assessment," in IEEE International Conference on Multimedia and Expo Workshops, 2012.
[29] L. Coghlan, U. Utzinger, R. Drezek, D. Heintzelman, A. Zuluaga, C. Brookner, R. Richards-Kortum, I. Gimenez-Conti and M. Follen, "Optimal fluorescence excitation wavelengths for detection of squamous intra-epithelial neoplasia: results from an animal model," OPTICS EXPRESS, vol. 7, no. 12, pp. 436-446, December 2000.
[30] S. R. McDougall, A. R. Anderson and M. A. Chaplain, "Mathematical modelling of dynamic adaptive tumour-induced angiogenesis: Clinical implications and therapeutic targeting strategies," Journal of Theoretical Biology, p. 564–589, 2006.
[31] P. P. Hsu and D. M. Sabatini, "Cancer cell metabolism:Warburg and beyond," Cell, pp. 703-707, 5 September 2008.