肖志明 宋榮 賈錚 李陽 樊霞
摘要以不同廠家阿莫西林膠囊及其內(nèi)容物近紅外(Near infrared, NIR)光譜為例,尋找評價分段直接標(biāo)準(zhǔn)化算法(Piecewise direct standardization, PDS)進行光譜校正是否成功的量化指標(biāo)。本研究共涉及76批阿莫西林膠囊樣品,其中54批用于建立膠囊劑的定量模型。通過聚類分析,所有膠囊的NIR光譜分成5類,每類視為一個均質(zhì)樣本;分別計算每個均質(zhì)樣本的平均光譜,從該樣本中選擇10~15張光譜作為PDS校正的目標(biāo)光譜,對76批阿莫西林膠囊內(nèi)容物粉末光譜進行校正,利用阿莫西林膠囊定量模型對校正后的光譜進行含量預(yù)測;計算校正后的光譜與PDS校正中目標(biāo)光譜所屬均質(zhì)樣本的平均光譜的相似系數(shù),分析其與預(yù)測誤差的關(guān)系。結(jié)果表明,校正結(jié)果與所選擇的目標(biāo)光譜關(guān)系密切。PDS校正光譜與模型中不同均質(zhì)樣本平均光譜的相似系數(shù)(r)越大,通常校正效果越好;當(dāng)r<99%時,一般可判斷PDS校正失?。A(yù)測誤差>5%)。因此, 可以用PDS校正后光譜與校正時使用的目標(biāo)光譜所屬的均質(zhì)樣本的平均光譜的相似系數(shù)作為判斷PDS校正是否成功的標(biāo)志。
關(guān)鍵詞PDS算法; NIR定量模型; 預(yù)測結(jié)果; 誤差分析
1引言
利用近紅外(Near infrared, NIR)技術(shù)識別假劣藥品和進行藥品生產(chǎn)過程控制,已經(jīng)成為藥物分析的新熱點\[1~4\]。NIR技術(shù)的應(yīng)用與所采用的模型關(guān)系密切。NIR模型優(yōu)劣不僅與建模所選擇的譜段\[5,6\]、預(yù)處理方法\[7\]和算法\[6\]有關(guān),更與建模訓(xùn)練集樣本的代表性關(guān)系密切\[8,9\]。為表述NIR建模樣本的代表性問題,本研究組提出了均質(zhì)樣本概念\[10,11\]。NIR定量模型的訓(xùn)練集中包含有若干個不同的均質(zhì)樣本;當(dāng)模型遇到建模時未包括的新均質(zhì)樣本時,預(yù)測結(jié)果就可能出現(xiàn)較大偏差。這時可以通過加入新樣本進行模型更新或利用化學(xué)計量學(xué)算法對新樣本光譜進行校正,擴展原模型的適用范圍\[10~13\]。為了解決NIR技術(shù)在企業(yè)生產(chǎn)過程控制應(yīng)用之初代表性樣品收集困難、建模繁瑣問題,開展了對已建立的通用性模型經(jīng)校正后作為生產(chǎn)過程控制初始模型的研究\[13\],已采用分段直接標(biāo)準(zhǔn)化算法(Piecewise direct standardization, PDS)和斜率/截距(Slope/Bias, S/B)算法,用頭孢拉定膠囊定量模型直接預(yù)測生產(chǎn)過程的中間體膠囊內(nèi)容物的含量,取得了良好的預(yù)測效果。但如何合理快速地判斷這些校正方法的效果問題仍未解決。
本研究以阿莫西林膠囊定量模型為例,選擇不同的目標(biāo)光譜用于PDS校正,嘗試將生產(chǎn)過程中間體阿莫西林膠囊內(nèi)容物光譜經(jīng)過PDS校正后通過阿莫西林膠囊定量模型預(yù)測含量。通過探討PDS校正中,校正光譜和模型訓(xùn)練集中不同均質(zhì)樣本平均光譜的相似系數(shù)(簡稱校正光譜相似系數(shù))與預(yù)測誤差的關(guān)系,尋找用于判定PDS校正準(zhǔn)確性的量化指標(biāo),為實際應(yīng)用提供指導(dǎo)。
2方法原理
2.1光譜預(yù)處理方法
對光譜進行預(yù)處理可以提高定量分析中光譜數(shù)據(jù)與其對應(yīng)含量值之間的相關(guān)關(guān)系。本實驗主要采用的光譜預(yù)處理方法有:
2.1.1直線差減法直線差減法是一種針對傾斜光譜的經(jīng)典基線校正方法。首先對校正的波段用最小二乘擬合一條直線,然后從光譜中減去該直線,達到基線校正的目的\[14\]。
2.1.2矢量歸一化法在NIR固體樣本的定量分析中,一般假設(shè)測量時近紅外光在樣品中的有效路程一致。但樣品內(nèi)部的粒度、晶型及測量重復(fù)性等因素都易引起測量光程的變化。
轉(zhuǎn)換矩陣F就可以將轉(zhuǎn)化光譜Xs轉(zhuǎn)換成與目標(biāo)光譜相匹配的光譜Xs,std。
利用PDS可以進行不同儀器光譜間系統(tǒng)誤差的校正。本研究嘗試將阿莫西林膠囊光譜(目標(biāo)光譜)與其內(nèi)容物粉末光譜(轉(zhuǎn)化光譜)的差異看作是系統(tǒng)誤差(主要為膠囊殼的差異),采用PDS法對膠囊內(nèi)容物的光譜先進行校正,再利用阿莫西林膠囊含量預(yù)測模型預(yù)測膠囊內(nèi)容物的含量。
2.3均質(zhì)樣本
均質(zhì)樣本是指主成分含量不同,輔料以及制劑工藝相同或相近的一組樣品。該理念起源于通用性氧氟沙星注射液定量模型的研究\[11\],研究中發(fā)現(xiàn)模型訓(xùn)練集中,處方相同僅活性成分含量不同的樣本在其主成分得分圖中可明顯集中于一組,該組樣本被稱為一個均質(zhì)樣本;輔料處方不同的樣本屬于不同的均質(zhì)樣本;均質(zhì)樣本的NIR光譜具有高度的相似性。后來該理念被延伸至藥品固體制劑:認為NIR光譜具有高度相似性的一組樣本為一個均質(zhì)樣本;均質(zhì)樣本可通過聚類分析的方法進行劃分,用于建模時訓(xùn)練集樣本的選擇以及判斷模型是否需要更新。均質(zhì)樣本光譜相似性的閾值可利用相關(guān)系數(shù)確定\[10\]。定量模型的訓(xùn)練集可以認為由一個或幾個均質(zhì)樣本組成,建模時應(yīng)從不同的均質(zhì)樣本中選擇代表性樣品組成訓(xùn)練集,當(dāng)預(yù)測訓(xùn)練集未包含的均質(zhì)樣本樣品時,預(yù)測誤差變大,需要對模型更新或?qū)υ擃悩颖具M行校正。
3實驗部分
3.1儀器與試劑
Bruker MatrixF傅立葉變換NIR光譜儀,配有光纖探頭測樣附件,銦鎵砷(InGaAs)檢測器,Bruker公司OPUS 5.5光譜分析軟件。島津20A高效液相色譜分析系統(tǒng),配有自動進樣器,二極管陣列檢測器以及工作站。
76批阿莫西林膠囊(規(guī)格為0.25 g和0.50 g)為2010年全國評價性抽驗樣品,含量(mg/mg)范圍為84.0%~67.5%;阿莫西林對照品(批號:130409201011),由中國食品藥品檢定研究院提供。
3.2含量參考值的測定
按中國藥典2010版HPLC法測定阿莫西林含量\[18\]。色譜柱:Dikma Diamonsil C18 (250 mm×2.4 mm, 5 μm);流動相:0.05 mol/L KH2PO4溶液(用2 mol/L KOH調(diào)至pH 5.0)乙腈(97.5∶2.5, V/V);檢測波長254 nm;柱溫:30 ℃;流速:1.7 mL/min;進樣量:20 μL。
3.3樣品NIR光譜的采集
利用光纖探頭分別采集阿莫西林膠囊和膠囊內(nèi)容物光譜。光譜測量范圍為4000~12000 cm
背景掃描次數(shù)為32次,樣品掃描次數(shù)為32次,測定溫度為室溫(22±2)℃,濕度為20%~50%。
從每批樣品中隨機抽取6粒膠囊,將光纖抵在單層膠囊殼的一側(cè)掃描,每粒掃描3張光譜,計算平均光譜。再將膠囊內(nèi)容物分別傾倒至標(biāo)準(zhǔn)NIR測量瓶中,將光纖插入內(nèi)容物中,掃描3張光譜,計算平均光譜。
3.4建立NIR模型
用Bruker OPUS軟件中的Qunant 2模塊,參照文獻\[8,9\],采用PLS算法建立阿莫西林膠囊定量模型。按照參考文獻\[7\]選擇訓(xùn)練集樣本:首先對所有的樣品光譜經(jīng)矢量歸一化處理后在全譜范圍內(nèi)采用Wards算法進行聚類分析,從76批光譜中選擇出54張光譜作為訓(xùn)練集,其它光譜作為驗證集;建模譜段為5400~7100 cm
3.5PDS校正
利用OPUS 5.5軟件中“Setup spectra transfer method”模塊,采用PDS法對阿莫西林膠囊內(nèi)容物的光譜進行校正,再利用所建立的阿莫西林膠囊定量模型預(yù)測膠囊內(nèi)容物的含量。參照文獻\[13\],設(shè)定PDS校正過程中使用的參數(shù);目標(biāo)光譜(阿莫西林膠囊光譜)數(shù)量一般選擇10~15張,窗口大小選擇7個波長點。
根據(jù)76批阿莫西林膠囊光譜的聚類分析結(jié)果,全部樣本大致可分成5類,每一類被認為是一個均質(zhì)樣本。從5類光譜中分別選擇PDS校正的目標(biāo)光譜,分別稱之為類Ⅰ、類Ⅱ、…、類Ⅴ光譜,計算同一膠囊內(nèi)容物光譜經(jīng)不同的目標(biāo)光譜校正后得到的校正光譜與各均質(zhì)樣本平均光譜的相似系數(shù)
4結(jié)果與討論
4.1阿莫西林膠囊定量模型
阿莫西林膠囊、膠囊內(nèi)容物和膠囊殼的NIR光譜呈明顯差異
4.2PDS校正
由于膠囊內(nèi)容物光譜與阿莫西林膠囊光譜具有較大的差異,直接利用阿莫西林膠囊模型預(yù)測膠囊內(nèi)容物的含量,誤差均大于5%;經(jīng)PDS校正后可以改善預(yù)測的準(zhǔn)確性,但部分樣本的預(yù)測誤差較大(>5%)(表2)。
PDS校正系通過對膠囊光譜和對應(yīng)的膠
囊內(nèi)容物光譜進行關(guān)聯(lián),將內(nèi)容物光譜校正成膠囊光譜進行預(yù)測。分別用類Ⅰ、類Ⅱ、…、類Ⅴ光譜對每一個內(nèi)容物光譜進行PDS校正(得到的校正光譜分別稱類Ⅰ、類Ⅱ、…、類Ⅴ校正光譜),預(yù)測含量,計算預(yù)測誤差;再計算諸校正光譜與阿莫西林膠囊各均質(zhì)樣本平均光譜的相似系數(shù),簡稱校正光譜相似系數(shù)(r);將每一個校正光譜的預(yù)測誤差與對應(yīng)的r值作圖,以預(yù)測誤差5%為分界,分析預(yù)測誤差與r的關(guān)系。
分析類Ⅰ校正光譜的預(yù)測誤差和與之對應(yīng)的r之間的關(guān)系,發(fā)現(xiàn)預(yù)測誤差隨著r的增大而減??;誤差大于5%的樣本共有14個,其中r最大為98.78%,最小為93.92%;誤差小于5%的樣本共有62個,其中r最大值為99.87%,最小值為98.84%。由主成分得分圖可見(圖2),預(yù)測誤差大于5%的光譜均分布在訓(xùn)練集樣本范圍之外,說明其與訓(xùn)練集光譜存在較大的差異;預(yù)測誤差小于2%的樣本,則基本都分布在訓(xùn)練集范圍之內(nèi),說明其與訓(xùn)練集光譜的相似性較高;證明了校正光譜的相似性直接影響預(yù)測結(jié)果。
同法分別分析類Ⅱ、…、類Ⅴ校正光譜的預(yù)測誤差和其在模型訓(xùn)練集光譜主成分得分圖中的位置(圖3),結(jié)果均與類Ⅰ校正光譜相似,預(yù)測誤差隨著r的增大而減小。預(yù)測誤差和r匯總于表3。結(jié)果表明,當(dāng)PDS校正光譜與建模的均質(zhì)樣本光譜均存在較大差異時,模型將不能對其準(zhǔn)確預(yù)測。
繪制r的正態(tài)分布曲線(圖4),單側(cè)檢驗,計算出其95%的置信區(qū)間為0.9863~0.9994。由于此正態(tài)分布中的r所對應(yīng)的校正光譜的預(yù)測誤差均小于5%,故可以認為當(dāng)r>98.63%時,模型對校正光譜的預(yù)測誤差小于5%。即r=99%可作為閾值, 用于判斷PDS校準(zhǔn)成功與否。5結(jié)論
PDS校正可以擴展NIR模型的適用范圍,但校正的成功與否與所選擇的目標(biāo)光譜關(guān)系密切。利用均質(zhì)樣本概念,計算PDS校正光譜與模型中諸均質(zhì)樣本光譜的相似系數(shù)(r)。通常r越大,校正效果越好;當(dāng)r<99%時,一般可判斷PDS校正失?。A(yù)測誤差>5%),據(jù)此可以選擇適宜的目標(biāo)光譜進行PDS校正,也可以判斷PDS校正的成功與否。
References
1Moffat A C, Assi S, Watt R A. J. Near Infrared Spec., 2010, 18(1): 1-15
2FENG YanChun, HU ChangQin. Spectrosc. Spect. Anal., 2014, 34(5): 1222-1228
馮艷春, 胡昌勤. 光譜學(xué)與光譜分析, 2014, 34(5): 1222-1228
3Reich G. Adv. Drug Deliver. Rev., 2005, 57: 1109-1143
4Boyer C, Gaudin K, Kauss T, Gaubert A, Boudis A, Verschelden J, Franca M, Roussille J, Boucher J, Olliaro P, White N J, Millet P, Dubost J. J. Pharm. Biomed. Anal., 2012, 6768: 10-15
5Li H D, Liang Y Z, Xu Q S, Cao D S. Anal. Chim. Acta, 2009, 648: 77-84
6Liu Y Y, Hu C Q. Vib. Spectrosc., 2011, 55: 241-249
7Ni Z, Feng Y C, Hu C Q. J. Anal. Bioanal. Techniques, 2010, 1(3): 1-7
8Jia Y H, Liu X P, Feng Y C, Hu C Q. AAPS PharmSciTech., 2011, 12(2): 738-745
9Zou W B, Feng Y C, Song D Q, Hu C Q. J. Chin. Pharm. Sci., 2012, 21(5): 459-467
10Zou W B, Feng Y C, Dong J X, Song D Q, Hu C Q. Sci. China. Chem., 2013, 56(4): 533-540
11Hou S R, Feng Y C, Zhang X B, Hu C Q. J. Chin. Pharm. Sci., 2012, 21(1): 62-69
12Zhang X B, Feng Y C, Hu C Q. Anal. Chim. Acta, 2008, 630: 131-140
13LEI DeQing, HU ChangQin, FENG YanChun, FENG Fang. Acta Pharm. Sin., 2010, 45 (11): 1421-1426
雷德卿, 胡昌勤, 馮艷春, 馮 芳. 藥學(xué)學(xué)報, 2010, 45 (11): 1421-1426
14KE BoKe. Research on Application of Near Infrared Spectroscopy in Silicagel Column Chromatography Process. Hangzhou: Zhejiang University, 2007: 8
柯博克. 近紅外光譜在硅膠柱層析過程分析中的應(yīng)用研究. 杭州: 浙江大學(xué), 2007: 8
15NI Zhen,HU ChangQin, FENG Fang. Chin. J. Pharm. Anal., 2008, 28(5): 824-829
尼 珍, 胡昌勤, 馮 芳. 藥物分析雜志, 2008, 28(5): 824-829
16Candolfi A, Maesschalck De R, JouanRimbaud D, Hailey P A, Massart D L. J. Pharm. Biomed. Anal., 1999, 21: 115-132
17ZHANG XueBo, FENG YanChun, HU ChangQin. Chin. J. Pharm. Anal., 2009, 29(8): 1390-1399
張學(xué)博, 馮艷春, 胡昌勤. 藥物分析雜志, 2009, 29(8): 1390-1399
18Chinese Pharmacopoeia Commission. Chinese Pharmacopoeia. Beijing: Chinese Medical Science and Technology Press, 2010, 401-402
國家藥典委員會. 中國藥典. 北京: 中國醫(yī)藥科技出版社, 2010: 401-402
AbstractThe near infrared (NIR) spectra of 76 batches of the amoxicillin capsules from different manufacturers and their corresponding content powder without capsules cell were used to find some quantitative indicators to evaluate whether the piecewise direct standardization (PDS) algorithm succeeded in NIR quantitative model updating. 54 batches were used to construct the NIR quantitative model for capsule preparation. All the NIR spectra of amoxicillin capsules were divided into five classes by cluster analysis, and each class can be regarded as a homology sample set. The average spectrum for each homology sample set was calculated. Ten to Fifteen spectra were selected from each homology sample set as the corresponding master spectra of the PDS algorithm to correct all the NIR spectra of the amoxicillin content powder respectively. Then the corrected spectra were predicted by the constructed NIR quantitative model for amoxicillin capsules. The prediction error for each corrected powder spectrum, and the correlation coefficient between each corrected powder spectrum and the average spectrum of the corresponding homolog sample set which the PDS master spectra came from, were calculated. Finally, the relationship between the prediction error and its corresponding correlation coefficient were studied. It was found that the correction results correlated closely with the selected master spectra set in PDS algorithm. The bigger the correlation coefficient (r), the better the correction results. In general, when r is less than 99%, it can be judged that the PDS correction is failed. At this condition, the prediction error is often more than 5%. Therefore, the correlation coefficient between the corrected spectrum and its corresponding average spectrum of the homology sample set can be used as an indicator to evaluate the efficiency of the PDS correction.
KeywordsPiecewise direct standardization algorithm; Near infrared quantitative model; Prediction results; Error analysis
7Ni Z, Feng Y C, Hu C Q. J. Anal. Bioanal. Techniques, 2010, 1(3): 1-7
8Jia Y H, Liu X P, Feng Y C, Hu C Q. AAPS PharmSciTech., 2011, 12(2): 738-745
9Zou W B, Feng Y C, Song D Q, Hu C Q. J. Chin. Pharm. Sci., 2012, 21(5): 459-467
10Zou W B, Feng Y C, Dong J X, Song D Q, Hu C Q. Sci. China. Chem., 2013, 56(4): 533-540
11Hou S R, Feng Y C, Zhang X B, Hu C Q. J. Chin. Pharm. Sci., 2012, 21(1): 62-69
12Zhang X B, Feng Y C, Hu C Q. Anal. Chim. Acta, 2008, 630: 131-140
13LEI DeQing, HU ChangQin, FENG YanChun, FENG Fang. Acta Pharm. Sin., 2010, 45 (11): 1421-1426
雷德卿, 胡昌勤, 馮艷春, 馮 芳. 藥學(xué)學(xué)報, 2010, 45 (11): 1421-1426
14KE BoKe. Research on Application of Near Infrared Spectroscopy in Silicagel Column Chromatography Process. Hangzhou: Zhejiang University, 2007: 8
柯博克. 近紅外光譜在硅膠柱層析過程分析中的應(yīng)用研究. 杭州: 浙江大學(xué), 2007: 8
15NI Zhen,HU ChangQin, FENG Fang. Chin. J. Pharm. Anal., 2008, 28(5): 824-829
尼 珍, 胡昌勤, 馮 芳. 藥物分析雜志, 2008, 28(5): 824-829
16Candolfi A, Maesschalck De R, JouanRimbaud D, Hailey P A, Massart D L. J. Pharm. Biomed. Anal., 1999, 21: 115-132
17ZHANG XueBo, FENG YanChun, HU ChangQin. Chin. J. Pharm. Anal., 2009, 29(8): 1390-1399
張學(xué)博, 馮艷春, 胡昌勤. 藥物分析雜志, 2009, 29(8): 1390-1399
18Chinese Pharmacopoeia Commission. Chinese Pharmacopoeia. Beijing: Chinese Medical Science and Technology Press, 2010, 401-402
國家藥典委員會. 中國藥典. 北京: 中國醫(yī)藥科技出版社, 2010: 401-402
AbstractThe near infrared (NIR) spectra of 76 batches of the amoxicillin capsules from different manufacturers and their corresponding content powder without capsules cell were used to find some quantitative indicators to evaluate whether the piecewise direct standardization (PDS) algorithm succeeded in NIR quantitative model updating. 54 batches were used to construct the NIR quantitative model for capsule preparation. All the NIR spectra of amoxicillin capsules were divided into five classes by cluster analysis, and each class can be regarded as a homology sample set. The average spectrum for each homology sample set was calculated. Ten to Fifteen spectra were selected from each homology sample set as the corresponding master spectra of the PDS algorithm to correct all the NIR spectra of the amoxicillin content powder respectively. Then the corrected spectra were predicted by the constructed NIR quantitative model for amoxicillin capsules. The prediction error for each corrected powder spectrum, and the correlation coefficient between each corrected powder spectrum and the average spectrum of the corresponding homolog sample set which the PDS master spectra came from, were calculated. Finally, the relationship between the prediction error and its corresponding correlation coefficient were studied. It was found that the correction results correlated closely with the selected master spectra set in PDS algorithm. The bigger the correlation coefficient (r), the better the correction results. In general, when r is less than 99%, it can be judged that the PDS correction is failed. At this condition, the prediction error is often more than 5%. Therefore, the correlation coefficient between the corrected spectrum and its corresponding average spectrum of the homology sample set can be used as an indicator to evaluate the efficiency of the PDS correction.
KeywordsPiecewise direct standardization algorithm; Near infrared quantitative model; Prediction results; Error analysis
7Ni Z, Feng Y C, Hu C Q. J. Anal. Bioanal. Techniques, 2010, 1(3): 1-7
8Jia Y H, Liu X P, Feng Y C, Hu C Q. AAPS PharmSciTech., 2011, 12(2): 738-745
9Zou W B, Feng Y C, Song D Q, Hu C Q. J. Chin. Pharm. Sci., 2012, 21(5): 459-467
10Zou W B, Feng Y C, Dong J X, Song D Q, Hu C Q. Sci. China. Chem., 2013, 56(4): 533-540
11Hou S R, Feng Y C, Zhang X B, Hu C Q. J. Chin. Pharm. Sci., 2012, 21(1): 62-69
12Zhang X B, Feng Y C, Hu C Q. Anal. Chim. Acta, 2008, 630: 131-140
13LEI DeQing, HU ChangQin, FENG YanChun, FENG Fang. Acta Pharm. Sin., 2010, 45 (11): 1421-1426
雷德卿, 胡昌勤, 馮艷春, 馮 芳. 藥學(xué)學(xué)報, 2010, 45 (11): 1421-1426
14KE BoKe. Research on Application of Near Infrared Spectroscopy in Silicagel Column Chromatography Process. Hangzhou: Zhejiang University, 2007: 8
柯博克. 近紅外光譜在硅膠柱層析過程分析中的應(yīng)用研究. 杭州: 浙江大學(xué), 2007: 8
15NI Zhen,HU ChangQin, FENG Fang. Chin. J. Pharm. Anal., 2008, 28(5): 824-829
尼 珍, 胡昌勤, 馮 芳. 藥物分析雜志, 2008, 28(5): 824-829
16Candolfi A, Maesschalck De R, JouanRimbaud D, Hailey P A, Massart D L. J. Pharm. Biomed. Anal., 1999, 21: 115-132
17ZHANG XueBo, FENG YanChun, HU ChangQin. Chin. J. Pharm. Anal., 2009, 29(8): 1390-1399
張學(xué)博, 馮艷春, 胡昌勤. 藥物分析雜志, 2009, 29(8): 1390-1399
18Chinese Pharmacopoeia Commission. Chinese Pharmacopoeia. Beijing: Chinese Medical Science and Technology Press, 2010, 401-402
國家藥典委員會. 中國藥典. 北京: 中國醫(yī)藥科技出版社, 2010: 401-402
AbstractThe near infrared (NIR) spectra of 76 batches of the amoxicillin capsules from different manufacturers and their corresponding content powder without capsules cell were used to find some quantitative indicators to evaluate whether the piecewise direct standardization (PDS) algorithm succeeded in NIR quantitative model updating. 54 batches were used to construct the NIR quantitative model for capsule preparation. All the NIR spectra of amoxicillin capsules were divided into five classes by cluster analysis, and each class can be regarded as a homology sample set. The average spectrum for each homology sample set was calculated. Ten to Fifteen spectra were selected from each homology sample set as the corresponding master spectra of the PDS algorithm to correct all the NIR spectra of the amoxicillin content powder respectively. Then the corrected spectra were predicted by the constructed NIR quantitative model for amoxicillin capsules. The prediction error for each corrected powder spectrum, and the correlation coefficient between each corrected powder spectrum and the average spectrum of the corresponding homolog sample set which the PDS master spectra came from, were calculated. Finally, the relationship between the prediction error and its corresponding correlation coefficient were studied. It was found that the correction results correlated closely with the selected master spectra set in PDS algorithm. The bigger the correlation coefficient (r), the better the correction results. In general, when r is less than 99%, it can be judged that the PDS correction is failed. At this condition, the prediction error is often more than 5%. Therefore, the correlation coefficient between the corrected spectrum and its corresponding average spectrum of the homology sample set can be used as an indicator to evaluate the efficiency of the PDS correction.
KeywordsPiecewise direct standardization algorithm; Near infrared quantitative model; Prediction results; Error analysis