Title: Non-destructive prediction of low levels of melamine particles in milk powder using hyperspectral reflectance imaging and partial least square regression model Authors
|Lim, Jongguk -|
|Baek, Insuck -|
|Mo, Changyeun -|
|Lee, Hoyoung -|
|Kang, Sukwon -|
|Lee, Kangjin -|
|Kim, Giyoung -|
Submitted to: Food Engineering Progress
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: November 18, 2013
Publication Date: December 2, 2013
Citation: Lim, J., Kim, M.S., Baek, I., Mo, C., Lee, H., Kang, S., Lee, K., Kim, G. 2013. Non-destructive prediction of low levels of melamine particles in milk powder using hyperspectral reflectance imaging and partial least square regression model. Food Engineering Progress. 17(4):377-386. Interpretive Summary: Hyperspectral imaging methods, which combine spectroscopy and imaging, can provide rapid and non-destructive means to assess the quality and safety of agricultural products. In this study, near-infrared hyperspectral imaging combined with numerical analysis methods was used to detect relatively low melamine particle concentrations in dry milk powder. Melamine and milk powder mixture concentration levels, as low as approximately 0.05% by weight ratio, could be accurately predicted with the investigated analysis method. The techniques and results presented in this study are beneficial to food processors.
Technical Abstract: Melamine has been used in industrial manufacturing of numerous forms of plastics, fertilizer, adhesives and laminates. In 2008, dairy products tainted with melamine have been reported to be responsible for kidney stones and renal failure among infants and children in China. Some dairy farmers and manufacturers illegally added water in order to increase the volume of milk. In addition, melamine was intentionally added to boost the protein content. The detection of melamine in food is mainly accomplished using a variety of chemical assay methods such as High-Performance Liquid Chromatography (HPLC), Gas Chromatography–Mass Spectrometry (GC-MS). The detection limits for these methods are excellent trace amounts of the contaminant, but there are some major disadvantages in that these methods are time consuming, expensive, and labor-intensive requiring complicated sample pretreatment procedures. In addition, well-trained technicians are needed to operate the equipment. Hyperspectral imaging methods, which combine spectroscopic and imaging, can provide rapid and non-destructive means to assess the quality and safety of agricultural products. In this study, near-infrared hyperspectral reflectance imaging combined with partial least square regression (PLSR) analysis was used to predict the melamine particle concentration in dry milk powder. Melamine and milk powder mixture concentration levels ranging from 0.02% to 1% by weight ratio (g/g) were evaluated. The optimal PLSR result for predicting melamine concentration in milk powder was obtained using a 1st order derivative pretreatment with RV2 =0.974, SEP=±0.055%, and F=6, respectively.