In 2017, Vinod Khosla he told CNBC that the occupation of “radiologist will be obsolete in five years.” Although the founding father of Khosla Ventures was later modified this schedule as much as 15 yearshe maintained that image recognition through artificial intelligence will soon give you the chance to diagnose diseases from scans higher than human doctors.
Seven years later, radiologists still have to interpret most scans (even with AI software helping them); a more immediate challenge is the shortage of those doctors in United States AND Around the World.
While Khosla Ventures has backed several imaging startups including Vista.ai AND P. Biothe last company selected a company that makes the work of radiologists easier by reducing the time spent documenting reports, slightly than trying to switch the doctor with a machine.
On Tuesday, Khosla led a $50 million Series B round Council AI, which has developed a tool enabling the generation of reports for radiologists. World Innovation Lab and returning investors ARTIS Ventures, OCV Partners, Kickstart Fund and Gradient Ventures (Google’s artificial intelligence fund) also participated in the round. The financing raised the company’s total capital to over $80 million.
Rad AI was founded in 2018 by Dr. Jeff Chang, who accomplished his medical training as a radiologist at age 16 and later earned an MBA from UCLA, and serial entrepreneur Dr. Gurson.
Because Chang knew from his own experience as a practicing physician that most of radiologists’ time was spent documenting results slightly than analyzing images, the two got down to develop a proprietary LLM tool trained on radiology report datasets to automate physician findings and impression documentation.
While tech corporations didn’t widely use generative AI until OpenAI’s ChatGPT arrived in 2022, Rad AI is proud to be the first to introduce the technology. “I’m confident we’re the first company in the radiology industry to start using LLM,” Gurson, CEO of Rad AI, told TechCrunch. “We began this work in 2018, around the same time as open A.I. [first] models.”
Six years later, Rad AI products are used by about one-third of U.S. health systems and nine of the 10 largest radiology groups in the country, Gurson said.
The fresh capital might be used to build a team to implement Rad AI’s newest product: a standalone radiology reporting solution.
“We’re very interested, but there’s only so much we can implement at a time,” Gurson said, adding that Rad AI is hiring individuals who can install and maintain the software.
Some incumbents have attempted so as to add GenAI functionality to their radiology reporting software over the past 18 months, but Rad AI does not yet consider these corporations to be true competitors.
“At this point, probably 99-100% of the market is using our products,” he said. “If it’s any indication, we haven’t lost a single customer since we started.”