• First-ever randomized controlled trial regarding deep learning-based AI in radiology provides strong evidence for the clinical value of AI
• Large-scale study with 10,476 patients published in Radiology, the top journal in radiology

SEOUL, South Korea, Feb. 13, 2023 /PRNewswire/ -- Findings from a recent study demonstrated the clinical effectiveness of Lunit's AI solution for chest x-ray image analysis, with significant improvement in the detection of lung nodules in patients undergoing routine health checkups. The results of the study were published in Radiology.^[1]

While past retrospective studies have repeatedly indicated promising performance by AI in chest x-ray screening, evidence from a prospective trial assessing the impact of AI-based CAD software in real-world populations is highly warranted. Thus, researchers conducted a pragmatic, randomized controlled trial to investigate the clinical utility of AI in detecting actionable lung nodules among health checkup participants using Lunit INSIGHT CXR, Lunit's CE-marked AI solution for chest x-ray analysis.

Lunit INSIGHT CXR detects suspicious lesions in chest x-ray images, helping radiologists distinguish disease areas by providing the location of the lesion with an abnormality score that reflects the AI's confidence level. The AI solution can detect 10 of the most common chest abnormalities, including tuberculosis, with 97-99% accuracy.^[2]

Researchers included 10,476 adult patients, who had undergone chest x-rays at a health screening center between June 2020 and December 2021. The patients were randomly divided evenly into two groups—AI or non-AI. The first group's x-rays were analyzed by radiologists aided by AI while the second group's x-rays were interpreted without the AI results.

Lung nodules were identified in 2% of the evaluated patients. Analysis showed that the detection rate for actionable lung nodules on chest x-rays more than doubled when aided by AI (0.59%) than without AI assistance (0.25%). Moreover, the detection rate of malignant lung nodules on chest radiographs was higher in the AI group (0.15%) than in the non-AI group (0%). There was no significant difference in the false-referral rates between AI and non-AI-interpreted groups. Furthermore, health characteristics including older age and a history of lung cancer or tuberculosis did not have an impact on the efficacy of Lunit's AI solution, suggesting the algorithm's consistency across different populations.

"This is the first real-world evidence proving that AI for chest radiography can improve actionable nodule detection without increasing false positives," said Brandon Suh, CEO of Lunit. "We believe that this prospective study will lay the groundwork for AI to eventually become the standard of care for chest radiography."

"As our trial was conducted with a pragmatic approach, almost all enrolled participants were included, which is a real clinical setting," said study co-author Jin Mo Goo, M.D., Ph.D., from the Department of Radiology at Seoul National University Hospital. "Our study provided strong evidence that AI could really help in interpreting chest radiography. This will contribute to identifying chest diseases, especially lung cancer, more effectively at an earlier stage."