Title: Detection of Campylobacter Colonies using Hyperspectral Imaging Authors
|Siragusa, Gregory -|
Submitted to: Sensing and Instrumentation for Food Quality and Safety
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: February 24, 2010
Publication Date: March 12, 2010
Citation: Yoon, S.C., Lawrence, K.C., Line, J.E., Siragusa, G.R., Feldner, P.W., Park, B., Windham, W.R. 2010. Detection of Campylobacter Colonies using Hyperspectral Imaging. Sensing and Instrumentation for Food Quality and Safety. Vol. 4, Issue 1. pg. 35-49. Interpretive Summary: Campylobacter infections in humans are the leading cause of bacterial diarrhea in the United States and other developed countries. Isolation and detection for identification of Campylobacter in foods like chickens involve time-consuming or complicated laboratory tests taking several days to a week. Direct culturing on agar plates has been an effective method to isolate Campylobacter; however, distinguishing Campylobacter from non-Campylobacter contaminants that frequently grow on many existing agars is difficult. The incubation time for growing Campylobacter colonies in agar media is typically 24 hours to 48 hours. Hence, the objective of this research was to develop an imaging technique for detecting Campylobacter colonies in as early as 24 hours after the incubation. Hyperspectral imaging utilizing both imaging and spectroscopy provided information for characterizing the absorption features of the tested micro-organisms. Classification algorithms were compared in terms of detection accuracy. A band ratio method using continuum removed reflectance spectra was the most effective way for early detection of Campylobacter colonies in blood agar. The detection accuracy of the Campylobacter colonies was over 99% whereas the detection accuracy of the non-Campylobacter colonies was 90%. This research has implications for early detection of presumptive Campylobacter colonies. Also, the developed hyperspectral imaging protocol is applicable to other pathogen detection studies.
Technical Abstract: Isolation and detection of Campylobacter in foods via direct plating involves lengthy laboratory procedures including enrichments and microaerobic incubations, which take several days to a week. The incubation time for growing Campylobacter colonies in agar media is typically 24 hours to 48 hours. For detection and identification, presumptive Campylobacter colonies are picked up from Petri dishes and examined under a microscope, followed by confirmatory laboratory tests. However, oftentimes it is difficult to visually differentiate Campylobacter colonies from non-Campylobacter contaminants that frequently grow together with Campylobacter on many existing agars. A new pre-screening technique using non-destructive and non-contact hyperspectral imaging was developed to detect Campylobacter colonies grown in Petri dishes. A reflectance spectral library of Campylobacter and non-Campylobacter contaminants was constructed for characterization of absorption features in wavelengths from 400 nm to 900 nm. Blood agar and Camp-Cefex agar were used as culture media. Classification algorithms were developed for accurate early detection of Campylobacter colonies as early as 24 hours of incubation time. For detection of Campylobacter, single-band thresholding was compared to two-band ratio thresholding. Single-band thresholding was applied to a band at 503 nm for blood agar plates and 501 nm for Campy-Cefex plates. Band ratio thresholding was applied to two bands at 433 nm and 461 nm chosen from continuum removed spectra. The band ratio method using continuum removed spectra provided more accurate detection results with less false positives than the algorithm based on a single spectral band. Both classification algorithms performed equally well with over 99% detection of Campylobacter but with different detection accuracies for non-Campylobacter contaminants: 90% (band ratio) versus 86% (single band). Blood agar was the better culture medium than Campy-Cefex agar in terms of Campylobacter detection accuracy in hyperspectral imaging. This research has profound implications for early detection of Campylobacter colonies with high accuracy. Also, the developed hyperspectral imaging protocol is applicable to other pathogen detection studies.