'Lunit Insight CXR' Shows Superior Benign Pulmonary Nodule Diagnosis Rate Compared to Specialists

Medical AI company Lunit announced on the 18th that a study comparing multiple AI software and radiologists' interpretation abilities for detecting pulmonary nodules was recently published in the international journal Radiology. Radiology is recognized as the most prestigious journal in the field of radiology, with a citation index of 19.7.

Lunit's chest X-ray AI image analysis solution 'Lunit Insight CXR'
[Photo by Lunit]

The study was led by the research team of Kiki J. van Rijmenam from the ‘Project AIR (AI for Radiology)’ team at Radboud University Medical Center in the Netherlands. The team utilized chest X-ray data from 386 patients collected from seven hospitals in the Netherlands between January 2012 and May 2022.

The researchers conducted a comparative analysis by combining Lunit’s chest X-ray AI image analysis solution, 'Lunit Insight CXR,' with other overseas AI solutions for pulmonary nodule detection, and calculated evaluation scores for each. Additionally, the average interpretation results of 13 radiology specialists and 4 radiology residents were used for comparison.

According to the study results, based on the AUC value, a representative performance evaluation metric for AI models, Lunit Insight CXR demonstrated a pulmonary nodule detection ability with an AUC of 0.93, the highest among all AI solutions analyzed in the study. Generally, an AUC closer to 1 indicates superior performance, and models with an AUC above 0.8 are classified as high-performance models.

Lunit Insight CXR showed a statistically superior sensitivity of 89% compared to other AI products and radiology specialists, and its specificity was 80%, showing no statistically significant difference from that of radiology specialists (80%).

This study focused on independently evaluating the pulmonary nodule detection capabilities of commercial AI solutions and demonstrating the applicability and effectiveness of AI in real medical environments. Notably, four out of seven AI solutions outperformed the average interpretation ability of radiology specialists. The company explained that this suggests AI could soon become an important tool in the field of radiology.

Seobum Seok, CEO of Lunit, stated, "This study is significant as it is the first paper to provide third-party validation of the performance of Lunit Insight CXR. In 2020, through a third-party validation paper published in the top medical journal JAMA Oncology, 'Lunit Insight MMG' was recognized as superior to other products. Accordingly, Lunit has once again proven world-class performance across all Lunit Insight products currently held."

This content was produced with the assistance of AI translation services.