When Andrew Ninh couldn’t go to his high school graduation, the result was an innovative idea that may change the way cancers are diagnosed.
Ninh was just three months shy of his 18th birthday when he experienced a spontaneous pneumothorax—a collapsed lung without an apparent cause—and was rushed to a hospital’s intensive care unit. While the repair to his lung went smoothly, Ninh’s heart rate dropped dangerously low during his overnight stay. He was shocked when the nurses later revealed that the hospital’s medical detection system didn’t work.
“These alarms would go off at night and the nurses told me that they learned to ignore them,” Ninh said. “And being the computer scientist that I was—I knew there had to be a better way.”
His experience eventually morphed into founding Docbot, an artificial intelligence diagnostic system for detecting and determining cancer—with 98% accuracy in milliseconds—without ever taking a biopsy.
Ninh was one of five winners at the Business Journal’s Innovator of the Year awards on Sept. 25 (see other profiles, pages 1, 4, and 8).
Now 28, he was also the youngest of the five. The award came just a decade after “he started to build an artificial intelligence, deep learning software that efficiently captures data in real time,” Luis Vasquez, associate director of venture collaboration at UC Irvine Applied Innovation, told the audience of 400 when announcing the award.
The company has also impressed venture capital investors, raising $2 million in a seed round from renowned investors including Kleiner Perkins and Bold Capital Partners.
In July, Ninh was one of 22 people named as a Thiel Fellow, which awards $100,000 each to young people who want to build new things. The fellowship also includes support from the Thiel Foundation’s network of founders, investors, and scientists. Peter Thiel became a billionaire with early investments in startups like Paypal Holdings Inc. and Facebook Inc.
“Docbot has developed better computer vision software to save people from colon and stomach cancers by helping colonoscopists identify more precancerous polyps,” the Thiel Foundation said.
After Ninh’s stay in the hospital around age 18, he used AI to help the hospital improve its software.
Ninh, who dropped out of college, kept his eye on the health industry. He met two UCI doctors who are experts in gastroenterology, William Karnes, whose clinical interests includes colon cancer screening; and Jason Samarasena, who is focused on deep learning and polyp detection and image recognition. Along with Ninh’s high school friend, Tyler Dao, the four formed Docbot.
Ninh, who became the chief executive, took the company through the UCI Applied Innovation Wayfinder program and the Y Combinator accelerator program.
“What really kept us going was passion for the problem,” Ninh said in his acceptance speech at the Hotel Irvine event.
The company’s research evolved when gastroenterologists expressed great interest in how AI can help them. Dr. Karnes’ reputation in the industry helped create believers in the product, Ninh said.
“When we validated the AI, that’s when we knew this could really help patients,” Ninh said.
The company built a database that included data from 10,000 colonoscopies and 190,000 images of polyps. Called Qualoscopy, it is the largest database in the world for visual pathogenic polyp detection.
To determine if a polyp is precancerous, the traditional process normally takes about two weeks, whereas Docbot’s system optically detects it in fractions of seconds.
“It is up to 98% accurate and can take about 10 milliseconds to process an image,” Dr. Karnes said in a UCI press release.
Docbot is not yet approved by the Food and Drug Administration and is indicated for investigational use only. UCI Medical currently employs Docbot’s technology.
The American Cancer Society ranks colon cancer as the third most prevalent cancer with 97,000 new cases last year.
According to UCI, 90% of colon cancers could be prevented with a routine colonoscopy with 50% of the recommended screening population likely to have one of these precancerous polyps that Docbot can detect.
UCI estimated that $1 billion is spent annually to conduct these routine pathologies.
Dotbot’s creators said the company’s technology can be used in other areas besides colonoscopies.
“If a pulmonologist were doing a bronchoscopy, this [product] is something that could work really well for them,” Karnes said.