电子舌对中药滋味的区分辨识

Discrimination of traditional Chinese medicinal materials with different tastes based on electronic tongue

目的:利用电子舌对中药的滋味进行区分辨识。方法:对22种常用中药的水煎液进行电子舌检测,采用主成分分析、判别因子分析等统计分析方法对检测数据进行分析。结果:通过主成分分析可以区分不同滋味的药材以及相同滋味的不同药材;在主成分分析图上,22种中药按酸、苦、甜、咸的规律聚类分布;得到的酸、苦、甜、咸特征雷达图形状不同,可以很好的区分中药不同滋味;用判别因子分析建立滋味判别模型,并进行交叉验证,正确率为88.2%。结论:电子舌能够对中药滋味进行精确地区分辨识,结合统计方法,可以建立中药滋味判别模型。

Objective： To investigate the use of the electronic tongue in the evaluation of traditional Chinese medicinal materi- als with sour, bitter, sweet or salty tastes, and assess the possible application of the sensor in the evaluation of different tastes of tradi- tional Chinese medicinal materials. Method： Aqueous extracts of 22 traditional Chinese medicinal materials were measured by the electronic tongue. The data collected with the tongue was evaluated for discrimination of the samples with multivariate statistical meth- ods： principal component analysis （PCA） and diseriminant factor analysis （DFA）. Result： The electronic tongue was capable of dis- criminating between samples with different taste modalities and could also distinguish different samples eliciting the same basic taste. Twenty-two traditional Chinese medicinal materials could be classified into five clusters based on PCA. These differences seem to derive from the different tastes. DFA was applied to construct a model to discriminate traditional Chinese medicinal materials with different tastes. And the samples yielded about 88.2% accuracy for cross-validation. Conclusion： We confirmed that the electronic tongue may provide an analytical procedure for classification of the samples with respect to tastes of the traditional Chinese medicinal materials.