发酵乳乳酸菌数近红外快速检测方法的研究

Rapid detection method of lactic acid bacteria number in fermented milk by near infrared spectroscopy

用近红外光谱技术对发酵乳乳酸菌数进行快速检测。采用平板计数法测定67批次含活乳酸菌发酵乳样品的乳酸菌数实测值,并同时采集近红外光谱数据。以校正集根均方误差(RMSEC)、预测集根均方误差(RMSEP)、校正集交互验证的根均方误差(RMSECV)及其相关系数Rc、Rp、Rcv为评价指标建立最优的乳酸菌数定量模型。利用模型对10批次含活乳酸菌发酵乳样品的乳酸菌数进行预测。结果表明,双歧杆菌和乳杆菌数的RMSEC、RMSEP、RMSECV和Rc、Rp、Rcv分别为0.128、0.172、0.179和0.9466、0.9364、0.8932。嗜热链球菌数的RMSEC、RMSEP、RMSECV和Rc、Rp、Rcv分别为0.127、0.180、0.176和0.9478、0.9241、0.8982。乳酸菌总数的RMSEC、RMSEP、RMSECV和Rc、Rp、Rcv分别为0.111、0.194、0.154和0.9533、0.9100、0.9097。经外部验证后,双歧杆菌和乳杆菌数、嗜热链球菌数、乳酸菌总数模型预测值和实测值的最大绝对差值分别为0.35、0.40、0.33,满足不超过0.45的要求。说明模型具有良好的预测性能,可用于发酵乳中乳酸菌数的快速检测。

Lactic acid bacteria number of fermented milk was rapidly analyzed and detected by near infrared spectroscopy technology.The measured lactic acid bacteria number of 67 fermented milk samples containing live lactic acid bacteria were determined by using plate counting method,near infrared spectroscopic data of the samples were collected simultaneously.Using RMSEC(root mean square error of calibration),RMSEP(root mean square error of prediction),RMSECV(root mean square error of cross validation)and their correlation coefficents Rc,Rp,Rcv as the evaluation index,the optimal lactic acid bacteria numble quantitative model was established.The lactic acid bacteria number of 10 fermented milk samples was predicted by the model.The results indicate that the RMSEC,RMSEP,RMSECV and Rc、Rp、Rcv of Bifidobacteria and Lactobacillus spp number were 0.128,0.172,0.179 and 0.9466,0.9364,0.8932,respectively,the RMSC,RMSEP,RMSECV and Rc、Rp、Rcv of Streptococcus thermophilic number were 0.127,0.180,0.176 and 0.9478,0.9241,0.8982,respectively,the RMSC,RMSEP,RMSECV and Rc、Rp、Rcv of lactic acid bacteria number were 0.111,0.194,0.154 and 0.9533,0.9100,0.9097,respectively.After external validation,the maximum relative error between the predicted and measured values of Bifidobacteria and Lactobacillus spp number,Streptococcus thermophilic number and lactic acid bacteria number were 0.35,0.40 and 0.33,meeting the requirement of no more than 0.45.The model has good predictive performance and can be used for the rapid detection of lactic acid bacteria number in fermented milk.