Wednesday, May 22, 2024
spot_img
HomeTechnologyAI & Machine LearningLunit presents Seven Study Results at ECR 2024

Lunit presents Seven Study Results at ECR 2024

– Lunit to unveil four oral presentations and three poster presentations at ECR 2024, highlighting Lunit INSIGHT’s expansive capability, ranging from adaptability in different use cases to the potential to replace a human reader in mammography double-reading settings 

Lunit (KRX:328130.KQ), a leading provider of AI-powered solutions for cancer diagnostics and therapeutics, today announced the presentation of seven studies at the European Congress of Radiology (ECR) 2024, to be held in Vienna, Austria, from February 28 to March 3. Four will be presented through oral presentations and three as E-posters.

In a featured oral presentation led by Dr. Carolyn Horst from King’s College London, UK, 16,996 chest radiographs were retrospectively analyzed by Lunit’s AI-powered chest X-ray analysis solution, Lunit INSIGHT CXR. With a maximum sensitivity of 0.94 and specificity of 0.99, among various abnormalities, researchers were able to better understand the accuracy of Lunit INSIGHT CXR at different thresholds for different use cases and abnormalities.

Specifically, the AI model with high sensitivity could be appropriate for emergency findings like pneumothorax. Meanwhile, a high specificity is preferable to triage low-risk studies for reporting without missing actionable pathology. This dual approach highlighted the adaptability of Lunit’s AI tool to different clinical needs in routine practice. The research also indicates that AI can offer a valuable balance between sensitivity and specificity in the analysis of common chest abnormalities, showcasing its potential to reduce radiologists’ workload and enhance efficiency in reporting.

In another oral presentation, a research team from Radboud University Medical Center, Netherlands, showcased their result of an independent validation of multiple commercial AI products from leading vendors.

Among seven AI-powered algorithms, Lunit INSIGHT CXR achieved the highest AUC (Area Under the Curve) of 0.93 in lung nodule detection, outperforming human readers (mean AUC 0.81). The sensitivity of Lunit INSIGHT CXR reached 89%, outperforming human readers while maintaining specificity comparable to that of human readers at 80%. Four out of seven AI products, including Lunit INSIGHT CXR, showed superior performance compared to human readers. This validation provides crucial comparative performance data for AI algorithms, contributing to the ongoing discussions about the integration of AI in clinical practice.

Lunit INSIGHT MMG, Lunit’s AI-powered solution for early breast cancer detection, was highlighted in an award-winning study by Odense University Hospital, Denmark. Based on 249,402 consecutive screening mammograms in three distinct AI-integrated scenarios, the study assessed the software’s ability to replace one or more human readers in mammography double-reading settings. Scenario 1: AI replaces the first reader, Scenario 2: AI replaces the second reader in case it agrees with the decision of the first reader, and Scenario 3: AI acts as a standalone reader.

All three scenarios showed that AI-integrated screening can partially or fully replace one or both readers without affecting screening accuracy. Specifically, in Scenario 2, AI-assisted screening indicated significantly higher specificity (+0.6%) and positive predictive value (+4.7%). This groundbreaking research points towards a future where AI seamlessly integrates into double screening processes, enhancing efficiency, optimizing workflow, and maintaining stable accuracy.

Additionally, Lunit plans to announce the launch of the Leeds prospective trial, a collaboration with Leeds Teaching Hospitals NHS Trust, UK. The trial, which commenced on February 19, represents the world’s first prospective study of AI-powered breast reading in a symptomatic breast clinic environment. The study, scheduled to run until the end of 2025, will encompass 5,500 mammography cases of women visiting the UK Leeds Teaching Hospitals NHS Trust breast clinic. By evaluating Lunit INSIGHT MMG’s efficacy in real-world clinical practice, the trial aims to revolutionize breast cancer detection and streamline diagnosis, addressing radiologist shortages. Lunit invites booth visitors at ECR 2024 to explore detailed information on the project and discuss potential future collaborations.

“We are proud to present our latest research findings at this year’s ECR, demonstrating the robust performance and versatility of our Lunit INSIGHT suite in various clinical settings and scenarios. Our studies also reveal that our AI can enhance the efficiency and accuracy of mammography double-reading settings, and even replace human readers in some cases,” said Brandon Suh, CEO of Lunit. “We hope that these study results become a cornerstone to broaden the scope of our AI to more medical practices across the globe, and empower radiologists and patients to achieve better outcomes.” 

Join Lunit at booth AI-15 in the EXPO X1 hall to witness how the Lunit INSIGHT suite is shaping the future of radiology across the globe.

Oral presentations featuring Lunit INSIGHT at ECR 2024 include:

“The performance of a commercial artificial intelligence algorithm in an external quality assurance scheme regularly used by humans in the NHS breast screening programme” (ACV 2024 Research Stage 2, February 28, 3:00-4:00 pm)“The Multi- Sixteen Thousand and Counting: Performance of an Artificial Intelligence Tool for Identifying Common Pathologies on Chest Radiographs and Report Prioritisation” (ACV 2024 Research Stage 1, February 29, 2:00-3:30 pm)“Recent development in AI for lung nodule detection” (ACV 2024 Research Stage 2, March 1, 8:00-9:00 am)“Integration of artificial intelligence (AI) in double-read population-based mammography screening: simulated replacement of one reader and beyond” (ACV 2024 Research Stage 1, March 3, 11:30-12:30 pm)

Poster presentations featuring Lunit INSIGHT at ECR 2024 include:

“Can artificial intelligence decrease the time to histological diagnosis of lung cancer – a retrospective-cohort study” (EPOS Area -2 Level)“Deep Learning for Chest Radiograph Evaluation in Children: Repurposed Use of a Commercially Available AI Tool Developed for Adults (EPOS Area -2 Level)“Comparison of three AI breast density tools with a human reader” (EPOS Area -2 Level)

Visit AITechPark for cutting-edge Tech Trends around AI, ML, Cybersecurity, along with AITech News, and timely updates from industry professionals!

 

Post Disclaimer

The information provided in our posts or blogs are for educational and informative purposes only. We do not guarantee the accuracy, completeness or suitability of the information. We do not provide financial or investment advice. Readers should always seek professional advice before making any financial or investment decisions based on the information provided in our content. We will not be held responsible for any losses, damages or consequences that may arise from relying on the information provided in our content.

RELATED ARTICLES

Most Popular

Recent Comments

error: Content is protected !!