#South Korea #Seoul #Lunit #AI Diagnostics #cancer screening

Lunit Showcases Breakthrough AI Research in Cancer Screening at RSNA 2025

Lunit, a leading player in AI-driven cancer diagnostics and precision oncology, has unveiled plans to present a comprehensive collection of 14 studies at the upcoming Radiological Society of North America (RSNA) 2025 Annual Meeting. The vast array of research highlights Lunit’s commitment to enhancing cancer screening technologies, particularly focusing on areas such as screening mammography, digital breast tomosynthesis (DBT), breast density science, and sophisticated risk modeling.

Scheduled to take place in Chicago, Lunit’s exhibition will be located in South Hall, Booth #4100, where they will conduct expert-led sessions throughout the conference in the Lunit Education Room #1252. These presentations represent one of the most extensive evidence portfolios the company has showcased to date, reflecting their dedication to clinical excellence and innovation.

Unveiling Key Studies

Among the pivotal research featured, one of the studies conducted at Capio St Göran Hospital in Sweden analyzed over 193,000 screening examinations pre- and post-implementation of Lunit INSIGHT MMG. The findings revealed that integrating AI with a single radiologist significantly elevated the program's invasive cancer detection rate, improved the positive predictive value, and reduced unnecessary recalls compared to traditional human-only double reading.

This finding reinforces previous results from the ScreenTrustCAD interventional trial, indicating that AI can maintain or enhance accuracy while alleviating the workload on radiologists. Furthermore, a study from Massachusetts General Hospital assessed the performance of Lunit INSIGHT DBT on a sample of 1,000 retrospective examinations. The results demonstrated that the AI correctly localized 84.4% of true-positive cancers, showing exceptional efficacy for mass-presenting and invasive ductal carcinoma cases.

In another initiative, researchers from Elizabeth Wende Breast Care examined the influence of volumetric breast density quantified by AI on risk stratification, comparing it across leading models such as Tyrer-Cuzick and BOADICEA. Their analysis of 44,651 women revealed that Tyrer-Cuzick identified notably more women as high-risk due to heavier weighting on density and family history inputs compared to BOADICEA. Additionally, a longitudinal analysis of over 335,000 images illustrated that incorporating volumetric breast density over time significantly enhanced calibration and discrimination, while static or categorical density inputs often underestimated cancer incidence.

Expert-Led Educational Sessions

In addition to discussions about groundbreaking research, Lunit’s Education Room #1252 will host several expert-led sessions covering important aspects of AI in mammography and breast health. Highlights include:

• - Image-Based Risk (Nov 30, 1:30 PM - 2:00 PM CST): A discourse among experts including Drs. Graham Colditz, Joy Jiang, and Hari Trivedi on integrating image-based risk assessments with conventional models for personalized screening.
• - Interval Cancer Detection with AI (Dec 1, 11:30 AM - 12:00 PM CST): Presentation by Prof. Fiona Gilbert from the University of Cambridge on AI's capability to detect interval breast cancers typically missed in routine screenings, underscoring its importance in reducing the occurrence of undetected cancers.
• - Academic Leadership in AI Adoption (Dec 2, 10:30 AM - 11:00 AM CST): A comparative examination by Drs. Liz Morris and Elizabeth Burnside on implementing breast AI in educational institutions regarding evidence, ethics, and practical considerations.
• - Real-World DBT Performance (Dec 2, 4:30 PM - 5:00 PM CST): Dr. Manisha Bahl will present new clinical data about Lunit INSIGHT DBT’s performance, emphasizing advances in cancer detection and diagnostic consistency.

The Road Ahead for AI in Cancer Care

According to Lunit’s CEO Brandon Suh, “RSNA serves as a platform for the medical field to gauge the tangible clinical value delivered by AI. This year’s extensive program showcases how rigorously validated AI can enhance screening efficacy, assist radiologists, and fortify the foundations for risk-adapted care. We remain devoted to producing evidence that supports responsible and significant integration into clinical practices.”

Conclusion

Lunit's participation in RSNA 2025 promises to be a significant contribution to the dialogue on the intersection of technology and medicine, particularly in how AI can redefine cancer screening and risk modeling methodologies. As the company continues to push the boundaries of what’s possible in cancer diagnostics, the advancements showcased at this event will be crucial in shaping future practices in oncology care.

Lunit, founded in 2013 and trading under KRX 328130, is on a mission to outsmart cancer through AI, providing validated solutions across medical imaging and precision oncology. Their offerings aim to facilitate earlier detection and more informed treatment decisions, which are vital in the ever-evolving field of cancer care.