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 #310147

Title: High-throughput Raman chemical imaging for rapid evaluation of food safety and quality

item QIN, JIANWEI - University Of Maryland
item Chao, Kuanglin - Kevin Chao
item CHO, BYOUNG-KWAN - Chungnam National University
item Kim, Moon

Submitted to: Transactions of the ASABE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/1/2015
Publication Date: 6/1/2015
Citation: Qin, J., Chao, K., Cho, B., Kim, M.S. 2015. High-throughput Raman chemical imaging for rapid evaluation of food safety and quality. Transactions of the ASABE. 57:1783-1792.

Interpretive Summary: Food safety incidents in recent years due to milk adulterations have brought increased interest in developing rapid and accurate screening method for authenticating milk products. Current commercial Raman imaging instruments generally perform measurements at subcentimeter scales, which cannot be used for imaging whole surfaces of large food items. In this study, a newly developed line-scan hyperspectral system enables high-throughput macro-scale Raman chemical imaging for food safety and quality research. A high-power 785 nm line laser deployed in an optical configuration based on a dichroic beamsplitter provides a 24 cm long excitation line normally incident on the sample surface to generate Raman signals. The components in detection module were selected for optimal signal-to-noise ratio and quantum efficiency for rapid Raman imaging applications using 785 nm laser excitation. The Raman shift range and the spectral resolution of the imaging system are adequate for spectral analysis of most food and agricultural products. The system is flexible in collecting spatial information to accommodate different imaging applications requiring either large field of view or high spatial resolution. The long laser line and the comparable wide instantaneous field of view allow fast inspection of large sample areas using a push-broom method, resulting in a typical sampling time measured by minutes. In-house developed system software can generate chemical images in real time for visualizing different targets for different food safety and quality applications. An example application for authenticating flour was used to demonstrate the performance of the developed system. Chemical image was created using a simple image classification method for detecting azodicarbonamide particles mixed in unbleached flour. The high-throughput macro-scale Raman chemical imaging method and system has great potential for tackling existing and new challenges in the area of food safety and quality evaluation. The technique would benefit food processors in ensuring the safety and quality of their products and also regulatory agencies (e.g., FDA and USDA FSIS) with an interest in enforcing standards of food safety and quality.

Technical Abstract: High-throughput macro-scale Raman chemical imaging was realized on a newly developed line-scan hyperspectral system. The system utilizes a custom-designed 785 nm line laser with maximum power of 5 W as an excitation source. A 24 cm × 1 mm excitation line is normally projected on the sample surface using a 45° dichroic beamsplitter. A detection module consisting of a dispersive imaging spectrograph and a CCD camera is used to acquire Raman signals along the laser line. A motorized positioning table moves the samples transversely through the laser line, and hyperspectral data is obtained using a push-broom method. System software was developed using LabVIEW to fulfill various functions such as parameter setting, data transfer, and image processing. The system covers a Raman shift range of –691.5–2841.2 cm–1 with a spectral resolution of 14 cm–1. The system can collect spatial information requiring either large instantaneous field of view (e.g., 23 cm) or high spatial resolution (e.g., 0.07 mm). The system performance was demonstrated by an example application for authenticating flour. The system acquired a 320×250×1024 hypercube (80,000 spectra) from three powder samples placed in three Petri dishes (two with a diameter of 47 mm and the other 90 mm) in 9 min. Azodicarbonamide particles mixed in unbleached flour can be detected based on the chemical image generated using a simple image classification method. The method and system for high-throughput macro-scale Raman chemical imaging is promising to inspect other food and agricultural products for safety and quality evaluation.