Korean medical AI company Lunit said Thursday that a study published in the European Journal of Radiology (IF 3.2) demonstrated that its AI-powered chest X-ray solution, Lunit INSIGHT CXR, can reduce the time required for triaging emergency patients by 77 percent.

The study, conducted by Srinath Sridharan’s team at Singapore’s Changi General Hospital, analyzed 20,944 chest X-rays taken in the hospital’s emergency department between August and December 2023. The AI solution classified the images into three categories: normal, non-urgent, and urgent, and the results were evaluated by 43 radiology specialists.

The findings showed that the AI achieved a sensitivity of 89 percent and a specificity of 93 percent for normal cases. For non-urgent cases, the sensitivity was 93 percent and specificity 91 percent. In urgent cases, the AI recorded a sensitivity of 82 percent and a specificity of 99 percent, demonstrating strong performance in identifying emergency situations.

The study also found that using AI significantly reduced the workload for medical staff in emergency settings. On average, the AI solution reduced triage time by 77 percent compared to physicians. The AI’s minimum processing time was 0.2 seconds, significantly faster than the 1.7 seconds required by physicians.

“This study validates the effectiveness of AI in quickly and accurately classifying chest X-ray results in emergency room settings,” the research team said. “The AI’s performance will assist physicians in making immediate clinical decisions.”

Lunit CEO Brandon Suh added that the findings underscore the real-world utility of AI in medical environments. “Lunit AI’s 99 percent specificity in classifying urgent cases highlights its potential to improve healthcare efficiency,” he said.