本論文提出一種應用多頻譜影像再現技術於辨識食道癌細胞分期之方法,研究內容分為兩個部份:第一部份為利用多頻譜影像系統分析模擬的食道細胞頻譜;第二部份則應用主成份分析圖將不同分期的食道細胞頻譜做分類。本系統為利用多頻譜影像技術以及色彩再現並搭配主成份分析圖實現本研究。實驗結果表明,在主成份分析圖分類裡,使用第一主成份以及第二主成份將比第一主成份以及的三主成份更符合需求。根據這項特點,經由主成份分佈點的運算,我們可以將白種人正常食道細胞(OE21)、白種人第一期食道癌(OE21-1)、黃種人第一期食道癌(CE81T2-1)以及黃種人第四期食道癌(CE81T2-4)各別分類。 未來我們希望能夠利用此技術,有效地提升醫生診斷效率,幫助病患進行早期治療。
This paper presents an application of multi-spectral image reproduction technology in methods of identification staging esophageal cancer cells, the research divided into two parts: The first part is the use of multi-spectral imaging system analysis and simulation of the esophagus cell spectrum; the second part is principal component analysis diagram of the different stages of esophageal cell spectrum to classification. This system is the use of multi-spectral imaging technology and color reproduction and with the principal component analysis diagram realization of this research. Experimental results show that the principal component analysis chart points classification, the use of the first principal component and the second principal component than the first principal component as well as the three main ingredients more in line with demand. According to the characteristics of the principal components of distributed computing via the point, we can Caucasians normal esophageal cells (OE21), the first stage of Caucasian esophageal cancer (OE21-1), the first stage of yellow esophagus cancer (CE81T2-1) and the fourth stage of yellow fourth esophagus cancer (CE81T2-4) with individual classification. In the future we hope to be able to use this technology to effectively improve the efficiency of medical diagnosis, early treatment to help patients.