利用生物分子独特的拉曼光谱特征进行生物组织分类研究。利用研制的拉曼探针穿刺生物组织,获得生物分子的拉曼光谱信号数据,并对数据进行基线校正和滤波预处理;利用主成分分析法,提取拉曼光谱数据的关键特征;通过反向传播(BP)神经网络算法对这些特征进行组织分类;利用动物组织样品上采集的拉曼光谱数据,进行自动分类实验研究。结果表明,BP神经网络能够实现不同生物组织的分类,且准确率达到95%。 The biomolecular unique Raman spectroscopy is used to study the biological tissue classification. Raman spectroscopy signal data of biomolecules were obtained by puncturing the biological tissue with the developed Raman probe, and baseline correction and filtering preprocessing were performed on the data. Principal component analysis was used to extract the key features of Raman spectroscopy data; Propagation (BP) neural network algorithm was used to classify these features; the automatic classification experiment was conducted by using the Raman spectrum data collected from animal tissue samples. The results show that BP neural network can realize the classification of different biological tissues with the accuracy of 95%.