Researchers find Lunit's AI X-ray analysis solution effective enough to reduce human doctors' workload

By Kim Joo-heon Posted : April 8, 2024, 11:22 Updated : April 9, 2024, 01:03
Courtesy of Lunit
Lunit's AI breast cancer analysis solution 'Lunit INSIGHT MMG' [Courtesy of Lunit]

SEOUL, April 08 (AJU PRESS) - A study has shown that the adoption of an artificial intelligence-based breast cancer analysis solution can reduce the workload of human radiologists by about 70 percent. According to the research, the system called "Lunit INSIGHT MMG" can also significantly improve cancer detection accuracy.  

Lunit INSIGHT MMG compares and analyzes X-rays and medical images of breast cancer patients and healthy individuals. The AI solution alerts human supervisors if patients need further tests. It's been deployed in several countries, such as Taiwan, Sweden, and Saudi Arabia.

A research team led by Mustafa Ege Seker, a radiologist at Acibadem Mehmet Ali Aydinlar University in Türkiye, analyzed Lunit INSIGHT MMG's performance through 22,621 mammograms. The test was carried out from 2009 to 2019, with 8,825 women aged between 40 and 69 participating.

The research results showed that the AI solution could reduce the workload of human experts by 69.5 percent while increasing detection accuracy by approximately 30.5 percent. Around 50 percent of cancers missed in previous screenings were identified, along with 52 percent of interval cancers. Interval cancers occur between two consecutive regular screenings, even if there were no abnormalities found in the recent check-up.

The test result was published in a peer-reviewed journal "European Radiology" on Monday. "This research showcases how AI can streamline workflows, reduce the radiological burden, and potentially offer significant cost and time savings," said LUNIT CEO Brandon Suh. "We'll expand our research and development efforts to further integrate AI in ways that not only maximize the quality of care but also the sustainability of healthcare systems."
기사 이미지 확대 보기