As a graduate student at the Case School of Engineering, Shoshana Ginsburg excelled in an emerging area of biomedical science called medical image analysis. She joined the fledgling Center for Computational Imaging and Personalized Diagnostics in 2012 and conducted significant research using machine learning to characterize tumor types in prostate cancer.

Becoming the center’s first PhD graduate in 2015, she had a thought: Why not pair the same image processing technology with drone technology to help industries analyze aerial images?

Ginsburg and her husband, Meir, in 2017 founded Quanterra Software, a Cleveland company that uses the technology to count cattle. The company has successfully automated cattle counting for feedlots around the world.

“What I did for my company is take the medical imaging technology and just transfer it over to drone imaging,” she said in a video chat from her home in suburban Cleveland. “So far, by June we had counted almost 1 million cattle and our accuracy is 99 percent when ground conditions are normal.”

Auditing cattle is a tedious, time-consuming job, typically ordered by banks for loan collateral purposes. It usually requires boots on the ground to estimate heads of cattle that can number in the tens of thousands at large feedlots. 

Quanterra’s CattleQuantsTM technology uses algorithms to extract information. It can identify cattle characteristics such as breed, the same way similar technology detects tumor types and foretells how tumors will progress and respond to treatment.  

The technology has vast potential in industry. Ginsburg envisions expanding aerial inspections to pipelines, rail lines, roofs and runways; tracking wildlife populations; and helping with search and rescue missions at sea. 

The company landed a $225,000 National Science Foundation Small Business Innovation and Research grant to complete the development of this technology. It’s awaiting word on a follow-on grant of $1 million to extend the technology to ranches and pastures.

Besides compiling detailed inventory reports of cattle, Quanterra is also investigating the technology to detect bovine illness, using thermal sensors to capture body temperatures from the air.

Finding sick cattle is typically done by visual inspection. “The problem is that by the time symptoms are strong enough, visible enough to be picked up, the cow has infected many others,” Ginsburg explained. 

Consequently, antibiotics are often given to large numbers of healthy cattle to control bovine respiratory disease, the most common illness in cattle. The technology has potential to reduce use of antibiotics.

“Antibiotics are pricey, but even more importantly, there is lots of talk about antibiotics in the food chain causing resistance in humans, though that hasn’t been proven,” she said.

Quanterra is conducting experiments with Texas A&M AgriLife Research to determine how altitude, time of day and other factors affect aerial fever detection. 

Seeing a new image 

Ginsburg never dreamed of doing this work when she began her studies. Born and raised in Denver, Colorado, she was a math major pursuing a career as an actuary at the University of Colorado when she sort of stumbled into medical imaging.

She took a class during her last semester where students were tasked with solving medical imaging problems using different technologies. She was so taken with the work, she decided to pursue a master’s degree and became a research assistant for the radiation oncologist who conducted the class. 

When Ginsburg decided to pursue her doctorate in biomedical engineering, she looked across the country for a program.

“I wanted to work with someone who was a real star in this niche field,” she recalled. 

She found Professor Anant Madabhushi at Rutgers University. She turned down other opportunities, including three Ivy League universities, to study under him. He was impressed with her determination.

In 2012, Madabhushi left Rutgers to become the first director of the Center for Computational Imaging and Personalized Diagnostics at Case Western Reserve. Ginsburg didn’t think twice about following him to Cleveland.

Her lab work focused on prostate cancer, and she discovered that cancer in different parts of the organ have different appearances, attributes and properties. She developed models to identify the tumor type based on where it is located on the organ.

“She did extremely well at Case,” said Madabhushi, a professor of biomedical engineering. “She published high-impact papers in the prostate cancer space. Her papers have been gaining many citations.”

After being awarded her doctorate in 2015, she pondered potentially new applications for image analysis technology. She had always been interested in agriculture.

“I don’t know if there was a single ‘ah-ha moment,’” she said. “I thought about other areas that could benefit from the wealth of literature coming out in medical image analysis. I thought I could solve a problem.”

Her startup was lean, relying on about $45,000 in grants. In 2018, she pitched her concept to the Morgenthaler-Pavey Startup Competition. The competition, designed to boost startups with CWRU connections, had attracted 50 entries. Quanterra was named a finalist and Ginsburg was handed an outsized $10,000 check. She was on her way.

Madabhushi said he’s thrilled to see Ginsburg take her skills in a new direction to fulfill an unmet need. He urges student researchers to think boldly.

“She’s having an impact in a new area,” he said. “It’s wonderful to see her evolve.”

Spector, a Cleveland freelance journalist, is a former award-winning reporter for The Plain Dealer.

Comment on this story via casealum@casealum.org