Deep convolutional autoencoder for the simultaneous removal of baseline noise and baseline drift in chromatograms

被引:27
作者
Kensert, Alexander [1 ,4 ]
Collaerts, Gilles [1 ]
Efthymiadis, Kyriakos [1 ,2 ]
Van Broeck, Peter [3 ]
Desmet, Gert [4 ]
Cabooter, Deirdre [1 ]
机构
[1] Katholieke Univ Leuven, Univ Leuven, Dept Pharmaceut & Pharmacol Sci, Pharmaceut Anal, Herestr 49, B-3000 Leuven, Belgium
[2] Vrije Univ Brussel, Dept Comp Sci, Artificial Intelligence Lab, Pleinlaan 9, B-1050 Brussels, Belgium
[3] Janssen Pharmaceut, Dept Pharmaceut Dev & Mfg Sci, Turnhoutseweg 30, Beerse, Belgium
[4] Vrije Univ Brussel, Dept Chem Engn, Pleinlaan 2, B-1050 Brussels, Belgium
关键词
Deep learning; Machine learning; Autoencoder; Baseline noise; Baseline drift; Noise reduction; Denoising; PEAK DETECTION; LIQUID-CHROMATOGRAPHY; AUTOMATIC PROGRAM; DECONVOLUTION; STRATEGY; SIGNALS;
D O I
10.1016/j.chroma.2021.462093
中图分类号
Q5 [生物化学];
学科分类号
071010 ; 081704 ;
摘要
Enhancement of chromatograms, such as the reduction of baseline noise and baseline drift, is often essential to accurately detect and quantify analytes in a mixture. Current methods have been well studied and adopted for decades and have assisted researchers in obtaining reliable results. However, these methods rely on relatively simple statistics of the data (chromatograms) which in some cases result in significant information loss and inaccuracies. In this study, a deep one-dimensional convolutional autoencoder was developed that simultaneously removes baseline noise and baseline drift with minimal information loss, for a large number and great variety of chromatograms. To enable the autoencoder to denoise a chromatogram to be almost, or completely, noise-free, it was trained on data obtained from an implemented chromatogram simulator that generated 190.0 0 0 representative simulated chromatograms. The trained autoencoder was then tested and compared to some of the most widely used and well established denoising methods on testing datasets of tens of thousands of simulated chromatograms; and then further tested and verified on real chromatograms. The results show that the developed autoencoder can successfully remove baseline noise and baseline drift simultaneously with minimal information loss; outperforming methods like Savitzky-Golay smoothing, Gaussian smoothing and wavelet smoothing for baseline noise reduction (root mean squared error of 1.094 mAU compared to 2.074 mAU, 2.394 mAU and 2.199 mAU) and Savitkzy-Golay smoothing combined with asymmetric least-squares or polynomial fitting for baseline noise and baseline drift reduction (root mean absolute error of 1.171 mAU compared to 3.397 mAU and 4.923 mAU). Evidence is presented that autoencoders can be utilized to enhance and correct chromatograms and consequently improve and alleviate downstream data analysis, with the drawback of needing a carefully implemented simulator, that generates realistic chromatograms, to train the autoencoder. ? 2021 Elsevier B.V. All rights reserved.
引用
收藏
页数:15
相关论文
共 38 条
[1]  
Abadi M, 2016, PROCEEDINGS OF OSDI'16: 12TH USENIX SYMPOSIUM ON OPERATING SYSTEMS DESIGN AND IMPLEMENTATION, P265
[2]  
[Anonymous], PERFECT SMOOTHER ANA, DOI [10.1021/ac034173t, DOI 10.1021/AC034173T]
[3]  
BARTLETT MS, 1950, BIOMETRIKA, V37, P1
[4]   THE MEASUREMENT OF POWER SPECTRA FROM THE POINT OF VIEW OF COMMUNICATIONS ENGINEERING .1. [J].
BLACKMAN, RB ;
TUKEY, JW .
BELL SYSTEM TECHNICAL JOURNAL, 1958, 37 (01) :185-282
[5]   New background correction method for liquid chromatography with diode array detection, infrared spectroscopic detection and Raman spectroscopic detection [J].
Boelens, HFM ;
Dijkstra, RJ ;
Eilers, PHC ;
Fitzpatrick, F ;
Westerhuis, JA .
JOURNAL OF CHROMATOGRAPHY A, 2004, 1057 (1-2) :21-30
[6]  
Felinger A., 1998, DATA ANAL SIGNAL PRO, V21
[7]   Powerful Artificial Neural Network for Planar Chromatographic Image Evaluation, Shown for Denoising and Feature Extraction [J].
Fichou, Dimitri ;
Morlock, Gertrud E. .
ANALYTICAL CHEMISTRY, 2018, 90 (11) :6984-6991
[8]   An objective comparison of pre-processing methods for enhancement of liquid chromatography-mass spectrometry data [J].
Fredriksson, Mattias ;
Petersson, Patrik ;
Magnus, Joerten-Karlsson ;
Axelsson, Bengt-Olof ;
Bylund, Dan .
JOURNAL OF CHROMATOGRAPHY A, 2007, 1172 (02) :135-150
[9]   AntDAS: Automatic Data Analysis Strategy for UPLC-QTOF-Based Nontargeted Metabolic Profiling Analysis [J].
Fu, Hai-Yan ;
Guo, Xiao-Ming ;
Zhang, Yue-Ming ;
Song, Jing-Jing ;
Zheng, Qing-Xia ;
Liu, Ping-Ping ;
Lu, Peng ;
Chen, Qian-Si ;
Yu, Yong-Jie ;
She, Yuanbin .
ANALYTICAL CHEMISTRY, 2017, 89 (20) :11083-11090
[10]   A simple multi-scale Gaussian smoothing-based strategy for automatic chromatographic peak extraction [J].
Fu, Hai-Yan ;
Guo, Jun-Wei ;
Yu, Yong-Jie ;
Li, He-Dong ;
Cui, Hua-Peng ;
Liu, Ping-Ping ;
Wang, Bing ;
Wang, Sheng ;
Lu, Peng .
JOURNAL OF CHROMATOGRAPHY A, 2016, 1452 :1-9