胃粘膜组织共焦显微拉曼光谱分类研究

采用基于小波变换的光谱去噪和背景扣除预处理技术,对32例(其中:13例正常,19例癌变)胃粘膜组织拉曼光谱进行分析,克服了手动背景扣除的一些缺点,并观察到一些新的光谱特征。基于此,给出能对所有正常组织和癌变组织进行有效分类的特征量和判据。通过研究这些特征量与胃粘膜组织癌变发展阶段之间的定量关系,可望实现胃癌的早期诊断。

Classification of normal and malignant human gastric mucosa tissue with confocal Raman microspectroscopy

32 samples from the human gastric mucosa tissue, including 13 normal and 19 malignant tissue samples were measured by confocal Raman microspectroscopy. The low signal - to -background ratio spectra from human gastric mucosa tissues were obtained by this technique without any sample preparation. A preprocessed algorithm based on wavelet analysis was used to reduce noise and eliminate background/baseline of Raman spectra. Comparing preprocessed spectra of malignant gastric mucosa tissues with those of counterpart normal ones, there were obvious spectral changes, including intensity increase at -1156 cm^-1 and intensity decrease at 1587 cm^-1 The quantitative criterion based upon the intensity ratio of the -1156 and 1587 cm^-1 was extracted for classification of the normal and malignant gastric mucosa tissue samples, This could result in a new diagnostic method, which would assist the early diagnosis of gastric cancer.