AI-based software scours literature to help physicians get to a diagnosis faster
Kahun, backed by Waze co-founder Uri Levine, says platform can go through text and point doctors to a recommended course of action based on medical sources
Shoshanna Solomon was The Times of Israel's Startups and Business reporter
Avid followers of the TV medical series “Grey’s Anatomy” will be familiar with the idea of interns scouring piles of medical literature in an attempt to be the first to find a solution to the mysterious problems afflicting their patients.
If only they had a smart system that could do all of that reading for them and point them in the right direction, one can imagine them thinking. That is exactly what Israeli startup Kahun hopes to do with its AI-based software, which is meant to ease doctors’ paths from facing huge amounts of text to making the right diagnostic call.
“Ever since the first medical article was ever published on a papyrus in 2,000 BCE,” medical knowledge has been documented, but until 2020 this knowledge has always been in written form, said Eitan Ron, a co-founder of Kahun, in a phone interview. “We are now trying to turn all that text around” into something physicians can use more efficiently.
The startup, by the way, is named for the Egyptian village of Kahun where that first medical paper, an OB/GYN-related article, was discovered, he explained.
After physicians submit their patients’ symptoms to Kahun’s website, the software scours the wealth of information available in medical texts and creates a “knowledge graph” that can map out 5 million relations between symptoms and diseases, findings, labs, complications and risk factors, said co-founder Dr. Michal Tzuchman-Katz, a pediatrician and a software engineer.
Kahun has also developed an AI engine that uses the knowledge graph created to “mimic” a physician’s thinking, she said. It assists them in making their diagnosis, points them toward a recommended course of action, and “backs them up with medical sources” by linking the information provided to the original medical sources consulted.
In addition to suggesting the most likely diseases, the software offers a plan for next steps, including the best questions to ask, tests to perform, or further lab work and imaging that is required to narrow down the diagnosis.
The web-based interface core application is used by hundreds of physicians and medical students in Israel, and the company on Thursday announced the release of an open beta testing phase of its product.
The company hopes the final product will be available commercially — initially for free, in the second half of this year, said Ron. Eventually, the company will charge users a subscription fee.
Ron added that the software was given the United States Medical Licensing Examination, and passed with an 80% score.
Kahun’s technology gained prominence during the COVID-19 pandemic as it used its AI capabilities to tap into more than 2,000 papers and articles found in the medical library PubMed studying the new coronavirus.
Using Kahun’s knowledge graph, doctors were able to screen patients and assess their risk for severe illness from the coronavirus. A joint study with the Rabin Medical Center, HaSharon Campus, showed Kahun’ efficacy in helping doctors make their risk assessments, the company said.
Serial entrepreneur Uri Levine, who co-founded the navigation app Waze — which was acquired by Google in 2013 for over $1 billion — is an investor in Kahun and is its chairman.
“All medical knowledge in the world today is in text and the only way to make sense out of it is if a doctor or medical staff is reading it, but what if we could convert that into data and build apps to use it? This is exactly what we do,” Levine said in a statement.
In the statement, a medical intern at a hospital in central Israel, Yohai Shraga, said that he uses Kahun whenever he found himself at a loss for a diagnosis. Kahun “has become my go-to gear to have in a tense hospital shift. I can use it to make sure I don’t miss anything, or to consult with just before turning to a senior physician.”
Over the past few years a number of AI-based applications have been released to help patients figure out their symptoms. One of them is Israel’s K Health, which this week said it raised $132 million in its third year of funding in under a year. K Health has developed a free mobile app that enables users to insert their age and gender and ailments to get an immediate initial diagnosis of what could be afflicting them.
These kinds of apps are geared to patients and most are limited to basic symptom checking for a number of common diseases, said Ron, adding that Kahun taps into a far greater pool of knowledge and is geared toward medical professionals. “We see no face-to-face competition with other players,” he said.
The Tel Aviv-based startup, which employs 22 workers, was co-founded by Tzuchman-Katz, Ron and Tal Goldberg in 2018. The three met when working together at HumanClick Ltd., an Israeli startup founded by Ron and Goldberg and acquired by LivePerson, a maker of chatbots for businesses, in the early 2000s. Tzuchman-Katz, who was a software engineer at LivePerson, went to medical school and became a pediatrician, and the three got together in 2018 to start Kahun.
