该文应用近红外光谱和类别相关残差约束的非负表示分类(class-specific residual constraint non-negative representation base classification, CRNRC)方法快速识别被毒素污染的贻贝。受腹泻贝类毒素(diarrheal shellfish toxins, DST)污染的贻贝,其组织变化可通过近红外光谱曲线反映出来,可利用CRNRC模型对健康贻贝和受DST污染的贻贝进行分类。在CRNRC模型中引入类相关残差项和协同表示,将编码与分类联系起来。研究了CRNRC的编码向量,通过实验确定CRNRC模型的最优参数。实验结果表明,CRNRC模型在平均准确率、F-measure、1-specificity等评价指标上均优于协同表示和非负表示分类模型;近红外光谱与CRNRC相结合,能有效地鉴别被DST污染的贻贝,该检测方法具有智能、无损、准确、不需要化学试剂等优点。可将CRNRC模型的近红外光谱检测方法扩展到其他海鲜产品的检测(如检测海鲜产品核污染程度),以确保人类摄入健康的海鲜产品。
Human beings are up against serious health hazards when ingesting toxins-contaminated shellfish.It is badly required to identify shellfish contaminated by toxins.Near-infrared spectroscopy and a class-specific residual constraint nonnegative representation classification (CRNRC) were applied to recognize toxins-contaminated mussels rapidly.The changes in the tissue of mussels contaminated with diarrhea shellfish toxins (DST) could be reflected in the near-infrared spectral curves.The CRNRC model was used to classify healthy and DST-contaminated mussel samples with the preprocessed near-infrared spectra of mussels as input.Class-specific residual terms and collaborative representation were introduced into the CRNRC model to relate the coding with classification.The coding vectors of collaborative representation were shown for CRNRC.The optimal parameters affecting the performance of the CRNRC model were determined through experiments.The experimental results showed that CRNRC model was superior to collaborative representation classification, and non-negative representation classification (NRC) for the evaluation indexes of average accuracy, F-measure, and 1-specificity.The study indicated that NIRS combined with the CRNRC could distinguish DST-contaminated mussel samples, which had the advantages of intelligence, non-destruction accuracy, and without chemical reagents.The detection method of CRNRC with NIRS would be extended to detect other seafood products, for example, testing the level of nuclear contamination in seafood products, which could ensure human beings ingest healthy seafood products.
[1] REGUERA B, RIOBÓ P, RODRÍGUEZ F, et al.Dinophysis toxins:Causative organisms, distribution and fate in shellfish[J].Marine Drugs, 2014, 12(1):394-461.
[2] FARABEGOLI F, BLANCO L, RODRÍGUEZ L P, et al.Phycotoxins in marine shellfish:Origin, occurrence and effects on humans[J].Marine Drugs, 2018, 16(6):188.
[3] GERSSEN A, POL-HOFSTAD I E, POELMAN M, et al.Marine toxins:Chemistry, toxicity, occurrence and detection, with special reference to the dutch situation[J].Toxins, 2010, 2(4):878-904.
[4] HOLLOWAY T, RIOS A, OKORO C, et al.Replacing high-performance liquid chromatography (HPLC) with high-resolution thin layer chromatography (TLC) for rapid radiopharmaceutical analysis[J].Nuclear Medicine and Biology, 2021, 96:S63.
[5] RODELL R, NING T S, GANGULY A, et al.DNA Damage Responses[M].New York:Springer, 2022:125-140.
[6] LEIS D, VISKIĆ M, JEROMEL A, et al.Analysis of varietal thiols in sauvignon blanc wines—Optimization of a solid-phase extraction gas chromatography Tandem mass spectrometry method[J].Food Analytical Methods, 2022, 15(6):1591-1605.
[7] MOLNAROVA K, DURIS A, JECMEN T, et al.Comparison of human IgG glycopeptides separation using mixed-mode hydrophilic interaction/ion-exchange liquid chromatography and reversed-phase mode[J].Analytical and Bioanalytical Chemistry, 2021,413 (16):4321-4328.
[8] KARGI F, TOPRAK H.Rational design of metal mesh particles for biological fluidized bed reactors[J].Journal of Chemical Technology & Biotechnology, 1994, 59 (2):201-204.
[9] KUMAR R, SHARMA V.A novel combined approach of diffuse reflectance UV-Vis-NIR spectroscopy and multivariate analysis for non-destructive examination of blue ballpoint pen inks in forensic application[J].Spectrochimica Acta Part A:Molecular and Biomolecular Spectroscopy, 2017, 175(3):67-75.
[10] GUO X H, CAI R, WANG S S,et al.Non-destructive geographical traceability of sea cucumber (Apostichopus japonicus) using near infrared spectroscopy combined with chemometric methods[J].Royal Society Open Science, 2018,5(1):170714.
[11] SHIM K, JEONG Y.Freshness evaluation in chub mackerel (Scomber japonicus) using near-infrared spectroscopy determination of the cadaverine content[J].Journal of Food Protection, 2019, 82(5):768-774.
[12] HUAI Q Y, GAO C L, MIAO J L, et al.Fast detection of saxitoxin using laser tweezers surface enhanced Raman spectroscopy[J].Analytical Methods, 2013, 5(23):6870-6873.
[13] VARRÀ M O, FASOLATO L, SERVA L, et al.Use of near infrared spectroscopy coupled with chemometrics for fast detection of irradiated dry fermented sausages[J].Food Control, 2020, 110(4):107009.
[14] PUERTAS G, VÁZQUEZ M.Fraud detection in hen housing system declared on the eggs′ label:An accuracy method based on UV-VIS-NIR spectroscopy and chemometrics[J].Food Chemistry.2019, 288(2):8-14.
[15] GHIDINI S, CHIESA L M, PANSERI S, et al.Histamine control in raw and processed tuna:A rapid tool based on NIR spectroscopy[J].Foods, 2021, 10(4):885.
[16] CHENG J H, SUN D W.Rapid and non-invasive detection of fish microbial spoilage by visible and near infrared hyperspectral imaging and multivariate analysis[J].LWT - Food Science and Technology.2015, 62 (2):1060-1068.
[17] TITO N B, RODEMANN T, POWELL S M.Use of near infrared spectroscopy to predict microbial numbers on Atlantic salmon[J].Food Microbiology,2012, 32(2):431-436.
[18] LIU Y, XU L L, ZENG S G, et al.Rapid detection of mussels contaminated by heavy metals using near-infrared reflectance spectroscopy and a constrained difference extreme learning machine[J].Spectrochimica Acta Part A:Molecular and Biomolecular Spectroscopy, 2022, 269(3):120776.
[19] LIU Y, QIAO F, XU L L, et al.Fast detection of diarrhetic shellfish poisoning toxins in mussels using NIR spectroscopy and improved twin support vector machines[J].Frontiers in Marine Science, 2022, 9:907378.
[20] YIN H F, WU X J.Class-specific residual constraint non-negative representation for pattern classification[J].Journal of Electronic Imaging.2020, 29 (2):1.
[21] ZHANG L, YANG M, FENG X C. Sparse representation or collaborative representation: Which helps face recognition?. Barcelona: International Conference on Computer Vision, 2011: 6-13.
[22] XU J, AN W P, ZHANG L, et al.Sparse, collaborative, or nonnegative representation:Which helps pattern classification?[J].Pattern Recognition, 2019, 88 (4):679-688.
[23] MISHRA P, LOHUMI S, Improved prediction of protein content in wheat kernels with a fusion of scatter correction methods in NIR data modelling[J].Biosystems Engineering, 2021, 203:93-97.
[24] BOYD S, PARIKH N, CHU E, et al.Distributed optimization and statistical learning via the alternating direction method of multipliers[J].Foundations and Trends® in Machine Learning, 2010, 3(1):1-122.
[25] HU K J, YIN H F, SUN J.Discriminative non-negative representation based classifier for image recognition[J].Journal of Algorithms & Computational Technology.2021, 15:174830262110449.
