When Good Food Goes Bad

While researching the use of light-scattering sensors to detect microscale contaminants on semiconductor wafers in the 1990s, Dan Hirleman got to thinking: Could the same technology be used to identify tainted food?

At the time, the United States Department of Agriculture was encouraging interdisciplinary research. Hirleman, mechanical engineering professor and William E. and Florence E. Perry school head, got in touch with food science professor Arun Bhunia at Purdue’s Center for Food Safety Engineering and learned that fast identification of food pathogens was a goal for scientists there. A partnership formed. The world of mechanical engineering cross-pollinated with food safety science.

“We realized that rapid, non-contact identification of bacteria was a similar problem to ours and that there should be significant cross-fertilization of our semiconductor work to food safety, so we did some preliminary research,” Hirleman says.

In the spring of 2000, an initial development proposal went before the food safety group; the research was promising, and the full proposal to pursue the light-scattering technology was funded.

Bacteria Beware

Named BARDOT (Bacteria Rapid Detection using Optical-scattering Technology), the process may revolutionize the world of pathogen detection in food, paving the way for quickly testing chicken, spinach, or other foods to determine whether it is safe or laden with bacteria that could make a person quite sick.

Listeria. Staphylococcus. Salmonella. Vibrio. Escherichia. With the advent of the BARDOT system, these bacteria can be identified in less than a day, instead of the current five- to seven-day period with conventional testing that involves growing bacterial cultures in agar.

Society is crying out for a better system for detecting toxins and making sure food is safe. Anyone who pays attention to the news knows that putting your trust in food manufacturing processes brings the possibility, albeit it slight, of food poisoning.

A Centers for Disease Control (CDC) study determined that the United States had 10,421 food-borne disease outbreaks from 1973 to 2006. So detecting food pathogens, and doing so quickly in order to focus in on the source, is crucial. The just-patented BARDOT system may well be the answer.

In a report outlining BARDOT’s evolution, its creators say the method can detect and identify bacteria in food samples “essentially instantaneously.” Using electromagnetic waves, it can identify bacteria in a petri dish within seconds. Scientists tested hot dogs, spinach, chicken, tomatoes, ground beef, and oysters spiked with bacterial cultures ranging from E. coli 0157:H7 to Salmonella enterica.

“These products were challenged with various pathogens at very low doses and our system is very sensitive and was able to detect the pathogens,” Bhunia says. Final results that now take almost a week are available in less than a day. “For food processors, products must be tested for pathogens before they are shipped for retail distribution. Also, some products have short shelf lives; thus, fast test results would protect consumers.”

Shining a Light on E. coli

Hirelman, Bhunia, and other researchers know from previous research that the transmission and reflection of laser light is sensitive to minute changes in a sample. They decided to apply the research to their food safety project.

This is how the testing works: BARDOT shines a laser beam through a petri-dish sample of a suspected pathogen or colony. “The colony interferes with the laser beam, and the image on the screen is of the beam plus all the distortions imposed by the colony as the laser beam passes through it,” Hirleman explains. “We call this image a forward-light-scattering signature, because it turns out that each strain of bacteria has a unique signature or fingerprint.”

He continues painting a picture. “In effect, BARDOT is sensitive to the differences in physical properties of colonies of bacteria. BARDOT then uses a computer to analyze these forward-scattering signatures and determine whether or not the bacteria forming the colony are harmful.”

In the world of food pathogens, this all means that once the technology is in use, detecting bacteria that lead to sometimes deadly food poisoning in humans may be as simple as shining a light through a tainted piece of turkey, then taking a look at the resulting scatter patterns to identify the culprit.

The BARDOT system took years to develop and perfect. But soon, it will save time and lives.

A Joint Effort

Collaboration and creative thinking from all around Purdue University combined to take the food safety invention on the road to commercialization. This ranged from the creation of a database to the development of a marketing plan.

“As for all research projects, the BARDOT project had its ups and downs,” Hirleman says. “Nothing would have been achieved without the partnership between food science and mechanical engineering, since neither group had anything close to the expertise required to do the project alone.”

When it came time to create a database for the scatter signatures and to develop more sophisticated image processing, the team brought Purdue Cytomics Professor Paul Robinson’s expertise to the table. He and some students had been working on multispectral analysis for cell detection and were able to adapt their research to develop methods that worked well with the BARDOT system.

Financial help came from graduate students at Purdue’s Fort Wayne campus, who developed a business plan and marketing strategy that resulted in funding for construction of the BARDOT system prototype.

“It was clear that none of the groups had any chance of developing the instrument alone,” Hirleman observes of the pieces that are helping lead the project towards commercialization. “It was the synergy, and trans-disciplinary effort, that made it possible.”

–Laura Lane