不同预处理方法对苯丙酮尿症FTIR/ATR光谱筛查模型的影响

建立苯丙酮尿症的FTIR/ATR光谱筛查模型，比较基线校正、平滑、求导、傅里叶退卷积等光谱预处理方法对模型精度的影响。利用多模型共识偏最小二乘法(cPLS)建立干血片中苯丙氨酸浓度的校正模型，以相关系数(r)、预测均方根误差(RMSEP)、平均相对误差(MRE)和预测准确率(Acc)等指标，考察不同预处理方法对建模效果的影响。结果 一阶微分9点平滑处理方法效果最好。与原始光谱相比，模型的r，RMSEP，MRE和Acc分别从0.822 7，115.8，0.395和94.6改善到0.889 9，102.2，0.286和100。本方法直接快速、 不消耗试剂、 不产生污染，有望成为PKU大人群快速筛查的简便、 绿色新技术。

Effects of Different Pretreatment Methods on the Phenylketonuria Screening Model by FTIR/ATR Spectroscopy

To establish a phenylketonuria screening model by FTIR/ATR spectroscopy, and to compare the effects of different pretreatment methods, such as baseline correction, smoothing, derivation, Fourier deconvolution, on the model quality. A consensus partial least squares regression method (cPLS) was used to build the quantitative model of phenylalanine in dried blood spots. The effects of different pretreatment methods on the model performance were investigated, using the correlation coefficient (r), root mean square error of prediction (RMSEP), mean relative error (MRE) and predictive accuracy (Acc). The nine-point smoothing coupled with the first differential was found to perform the best. Compared with the model by the original spectra, its r, RMSEP, MRE and Acc were improved from 0.822 7, 115.8, 0.395 and 94.6 to 0.889 9, 102.2, 0.286 and 100, respectively. With the advantages of fast speed, easy process, no reagents consumption and environmental protection, the present method is expected to become a simple and green technology for rapidly screening the neonatal phenylketonuria in a large population.