When Korea’s go master Le Sedol suffered a shocking defeat at the hands of Google DeepMind’s AlphaGo, the Korean society fell into a massive panic. The media talked about “jobs to disappear because of Artificial Intelligence (AI)” daily, and voices began to pour out stressing the need to prepare for the age of AI. The phrase “AlphaGo shock” was no exaggeration at all.
The medical field was no exception. Watson Oncology made its local debuts at Gachon University Gil Medical Center (GIMS) and Pusan National University Hospital (PNUH), opening the age of AI healthcare. It made headlines recently when a patient, faced with a difference of views between his doctor and Watson, chose the latter’s advice.
In the vortex of new development, however, there are people and places quietly focusing on developing related services at the industrial forefront. One such place is VUNO Korea, a startup that provides AI-based medical imaging service, and its CEO Lee Ye-ha.
|Lee Ye-ha, CEO of VUNO Korea, believes medical industry and artificial intelligence should, and can, establish a win-win relationship.|
Lee is an expert in data recognition and analysis tool, who started his career at the Samsung Advanced Institute of Technology. In 2014, Lee quit Samsung and jumped into the world of startups with a dream of creating “services that help the world.” Already, this is the fourth year he’s been in this game. VUNO’s technology has gained global recognition by placing itself high at ILSVRC 2015 — an international convention for visual cognitive function technologies. Products that grafted its technology to medical care have also begun to knock on the markets.
Lee didn’t find the medical field very attractive from a business viewpoint. But he chose this area and emphasized he has no regrets whatsoever about his choice.
“I wanted to use VUNO technologies meaningfully. Hospitals have a lot of data as they have to store them for five to 10 years under medical laws, and I thought hospitals were not using them.” Lee said in a recent interview with The Korea Bio Medical Business Review. “There’s a ton of data that we can analyze but have failed to do so. I thought it would be meaningful to take it out of the cave and making it valuable again. I believed that if our technology could help advance medical industry, it could ultimately contribute to my, and my families, health and happiness.”
VUNO Korea’s aspirations have borne fruits one by one. The company has developed the software that sharply decreased the time for assessing bone age, from five minutes to 20 seconds. It has also created a program that aids diagnosis of epilepsy through chest CT analysis. The program uses a case search system that matches a given CT scan with a similar CT scan in the database and then compares them with a patient’s clinical feature.
“We’re currently focusing on medical images, but ultimately we want to create AI software that comprehensively considers various biological signals to aid diagnosis,” Lee said. “We intend to develop technology to prevent diseases by analyzing changes in a patient’s heart rate, temperature, blood pressure and other biological signals.”
Doctors’ anxiety about AI
Not all people welcome the rapid rise of artificial intelligence. Among the skeptics are medical workers. More than a handful of medical professors are telling their students “to do this or that not to lose the competition with AlphaGo” in the aftermath of Lee Sedol’s defeat. These professors are treating AI as a “potential threat for doctors.” Lee cautioned against such a viewpoint.
“AI cannot replace doctors, nor should they do so,” he said. “No matter how much AI progresses, it can’t take the place of doctors’ authority as the ultimate decider, and will remain as just a supplemental tool to help physicians improve the quality of medical diagnoses.”
He compared AI to a new weapon, or a better magnifying glass, for doctors.
“We’ve moved from a time of using stethoscopes and X-rays to using CT scans and MRIs in diagnosis,” Lee said. “But the invention of the MRI has not replaced doctors. In fact, it provided patients with new information and improved the accuracy of diagnoses. I see AI as just another new and accurate diagnostic modality.”
Lee emphasized the phrase “ultimate decision-maker.” As an expert of AI, he seemed to face up to the technology’s limitations. As medical industry deals with human lives, it can run into various legal and ethical issues when AI makes misjudgments. Lee, well aware of this problem, said VUNO would continue to produce technologies that “assist physicians’ judgment.”
‘All workers in health industry need to know medical science well’
Unlike medical personnel’s anxieties, VUNO Korea and doctors are in a cooperative relationship. VUNO has doctors among its staff and the company, whenever it plans to develop a new technology, finds the way through communicating with doctors in university hospitals.
“Doctors’ perspectives are crucial since doctors will be the ones using our services,” he said. “The doctors within VUNO assess whether our ideas will aid doctors in practice. Only people who have diagnosed patients can make the judgment. They also help communicate with medical professors. Engineers like me don’t understand less than half of their conversations.”
VUNO is building organic relationships with many medical institutions. The company recently participated in Asan Medical Center’s “AI Medical Image Project” in efforts to promote academy-industry cooperation.
Lee also offered advice for medical students interested in artificial intelligence technology.
“We need people with thorough knowledge of the medical field and ability to ‘define its problems.’ It’s important that you focus on your medical studies,” she said. “Many people think that you need a lot of non-medical knowledge to work in the medical industry. In reality, that’s not true.”
Lee believes building trust between the AI and medical fields is a top priority since there still are few cases of grafting AI to healthcare. One could feel his pride in developing “technology that saves lives,” despite many trials and errors in the process, and anticipates VUNO Korea’s continuous leadership in Korea’s medical AI area.
<© Korea Biomedical Review, All rights reserved.>