- 期刊
- OpenAccess
近紅外光技術應用於穿山甲中藥成份之檢測
Identification and Quantitation Assays for Chinese Medicines Containing Pangolin Scales Using Near Infrared Spectroscopy
摘要
本研究利用近紅外光檢測技術，針對以穿山甲鱗片粉為主方之四種中藥處方進行光譜掃瞄，建立中藥成份檢測之近紅外光定性及定量分析模式，並建立光譜資料庫以進行成份鑑定比對。以波長空問之相關性、波長空間之最大距離、主成份空問之馬氏距離與主成份空間之殘差變異等四種定性模式搭配馬氏距離、最大距離及隨機分組之三種分組方式進行分析。定量分析之研究以多重線性迴歸及部份最小平方迥歸等兩種模式來進行探討。結果顯示，在定性分析方面，同一處方之不定比例背景成份分析中，以主成份空間之殘差變異定性模式搭配最大距離分組方式，所得之最佳辨識成功率為100%，其辨識最高精度為1%。四種處方之不同背景混合處方分析中，以主成份空間之殘差變異定性模式搭配隨機分組方式，可達到86%~96%之辨識成功率，並具有2%之微量檢測精度。在定量分析方面，以部份最小平方迥歸之定量模式進行分析，最佳分析結果顯示，判定係數可達0.999，其校正標準誤差SEC為0.082。
並列摘要
Near Infrared (NIR) Spectroscopy was used to scan Chinese medicines containing pangolin scales; identification and quantitation models were developed, and a library of NIR spectra was constructed. Four prescriptions of Chinese medicines taking pangolin scales as the principal constituent were examined in this study. Identification analyses were conducted by using correlation in wavelength space (CWS), maximum distance in wavelength space (MDWS), Mahalanobis distance in principal components space (MDPCS) and residual variance in principal components space (RVPCS); and three grouping schemes were investigated: Mahalanobis Distance (MhD), Maximum Distance (MD) and Random Selection (RS). Two models including multiple linear regression (MLR) and partial least square regression (PLSR) were used in quantitation analyses. The best qualitative results indicated that RVPCS model with MD grouping scheme reached a discrimination rate of 100% with 1% accuracy in the randomly assigned background constituents of the same prescription, and RVPCS model with RS grouping scheme gave a discrimination rate of 86%~96% with 2% accuracy in the randomly assigned background constituents of 4 mixed prescriptions. The best quantitative result indicated that PLSR model reached a determination coefficient of 0.999 with standard error of calibration (SEC) of 0.082.