Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Environmental Microbial & Food Safety Laboratory » Research » Publications at this Location » Publication #385114

Research Project: Sensing Technologies for the Detection and Characterization of Microbial, Chemical, and Biological Contaminants in Foods

Location: Environmental Microbial & Food Safety Laboratory

Title: Quantitative detection of benzoyl peroxide in wheat flour using line-scan short-wave infrared hyperspectral imaging

item KIM, GEONWOO - Orise Fellow
item LEE, HOONSOO - Chungbuk National University
item BAEK, INSUCK - Us Forest Service (FS)
item CHO, BYOUNG-KWAN - Chungnam National University
item Kim, Moon

Submitted to: Sensors and Actuators B: Chemical
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/23/2021
Publication Date: 10/27/2021
Citation: Kim, G., Lee, H., Baek, I., Cho, B., Kim, M.S. 2021. Quantitative detection of benzoyl peroxide in wheat flour using line-scan short-wave infrared hyperspectral imaging. Sensors and Actuators B: Chemical. 352:130997.

Interpretive Summary: Due to the use of benzoyl peroxide (BPO) as a bleaching agent for wheat flour and potential detrimental health effects from accumulative or long term consumption of excessive amounts, varying regulations have been established worldwide regarding permissible BPO levels in wheat flour, ranging from zero (banned use) to as much as 300 parts per million. Consequently, rapid and nondestructive methods to detect and measure BPO residues are needed to help enforce the allowable limits of BPO in flour. This study tested a hyperspectral imaging (HSI) system, image processing procedures, and detection model for using shortwave-infrared (SWIR) imaging to quantitatively assess BPO concentration in flour. Pure unbleached all-purpose wheat flour, pure BPO, and samples mixed at eight concentrations (by weight) of BPO in flour were prepared and imaged. The mixture concentrations (prepared by weight) ranged from 0.005% to 0.64% (50 to 6400 parts per million). The sample images spanned 1000 to 2500 nm, and were processed and analyzed to develop a detection model based on the method of partial least squares regression. The model was able to detect BPO in wheat flour at all eight concentrations with good correlation between predicted actual concentrations. The results demonstrated that this SWIR imaging method has high potential for use in industrial food processing to quantitatively detect BPO particles in wheat flour, to help processors and regulators ensure that flour products are in compliance with permissible BPO limits.

Technical Abstract: The addition of benzoyl peroxide (BPO) to wheat flour as a bleaching agent has been widely recognized as an important food safety issue due to its negative effects on human health. To address this issue, various nondestructive optical-based techniques have been developed to screen for BPO, such as Raman spectroscopy and hyperspectral imaging (HSI). In this study, a shortwave infrared (SWIR) HSI system was developed for the rapid detection of BPO particles in wheat flour. The SWIR HSI system, detailed hyperspectral image processing procedures, and optimal model to detect BPO particles were evaluated. The model was developed using the partial least square regression (PLSR) method. To improve performance of the model, effective wavelength regions were selected, and various pre-processing methods were applied to the PLSR analysis. The developed model was able to detect BPO in wheat flour at 50–6400 ppm with a high determinant coefficient (> 0.985) between predicted and actual values. The developed SWIR HIS system and optimized model demonstrated a high potential for discriminating BPO particles in wheat flour and allowed for its quantitative evaluation.