 |
- 13 oral presentations included, reaffirming the clinical value of AI-based medical image analysis
- Highlighting key topics such as early risk stratification based on AI abnormality scores and interval cancer classification
SEOUL, South Korea, March 4, 2026 /PRNewswire/ — Lunit (KRX:32813), a leading provider of AI for cancer diagnostics and precision oncology, today announced that 21 studies featuring its AI solutions will be presented at the European Congress of Radiology 2026 (ECR 2026), taking place March 4-8 in Vienna, Austria.
At this year’s congress, independent studies evaluating the clinical value of Lunit INSIGHT MMG, Scorecard, and Lunit INSIGHT CXR will be presented. Of the 21 accepted abstracts, 13 have been selected for oral presentations, which represent the congress’s main scientific sessions, while eight will be presented as posters.
One of the key studies featuring Lunit’s solutions to be presented at ECR 2026 is an early breast cancer risk assessment study conducted by Dr. Claudia Weiss and her team at AULSS n.2 “Marca Trevigiana”, a regional health authority in Treviso, Italy.
The researchers analyzed mammography data from 67,686 women to evaluate whether a risk score derived using Lunit INSIGHT MMG could identify women who were initially assessed as normal but were at higher risk of a subsequent breast cancer diagnosis.
The analysis showed that among 451 women who were ultimately diagnosed with breast cancer, the average score rose sharply from 15.4 at the first screening to 73.9 at the second screening. In contrast, among 67,235 women who were assessed as negative at both screenings, the average score showed little change, decreasing slightly from 6.7 to 6.4. This difference was observed regardless of breast density, demonstrating the potential of using Lunit INSIGHT MMG as a tool for early identification of women at high risk of developing breast cancer.
A study examining the potential role of AI in the interval cancer audit process will also be presented. A research team led by Professor Yan Chen at the University of Nottingham evaluated the applicability of AI in the interval cancer classification process of the UK’s NHS Breast Screening Programme (NHSBSP). Currently, within the NHSBSP, two expert readers retrospectively review interval cancer cases and classify them into Category 1, 2, or 3[1].
The researchers applied Lunit INSIGHT MMG to 409 interval cancer cases and assessed whether AI scores could distinguish Category 1 cases from Category 2 and 3 cases. Using a predefined approach in which cases with AI scores below a given threshold were classified as Category 1 and those above the threshold as Category 2 or 3, the AI correctly classified 63 of 65 cases as Category 1 at a threshold of 0.5, and 206 of 229 cases at a threshold of 10. Notably, under both thresholds, no Category 3 cases were incorrectly classified as Category 1. These results suggest that AI could be used as a supportive tool to prioritize Category 1 cases, which account for the majority of interval cancers, and to help specialists focus on cases requiring more detailed evaluation.
The results of a large-scale randomized controlled trial (RCT) using the breast density quantification solution Scorecard from Lunit International (formerly Volpara) will also be presented. A research team led by Dr. Carla van Gils at UMC Utrecht followed women who had negative mammography results but were classified as having extremely dense breasts by Scorecard, to assess whether supplemental MRI screening could reduce the incidence of advanced breast cancer.
The researchers followed 8,061 women in the MRI screening group and 32,312 women in the control group who received mammography only across three screening rounds. At the third screening round, the incidence of advanced breast cancer in the MRI group was statistically significantly lower than in the control group, by 2.6 cases per 1,000 women. This study suggests that strategies to more precisely screen women with extremely dense breasts through identification with quantitative density assessment offered by Scorecard, and to link them with appropriate additional screening, may lead to real clinical benefits.
“These studies demonstrate that AI can contribute beyond simple reading support, extending to early risk assessment, screening quality management, and identification of high-risk populations.” said Brandon Suh, CEO of Lunit. “We will continue to build clinical evidence that can be applied in real-world global screening environments through ongoing collaboration with leading medical institutions around the world.”
ECR is one of Europe’s leading radiology congresses and is widely recognized as a major international medical imaging conference. Held under the theme of “Rays of Knowledge”, ECR 2026 is expected to bring more than 20,000 radiologists, researchers, and industry professionals from around the world. Lunit has participated in ECR every year since 2020, consistently presenting its research and clinical results.
Join Lunit at booth AI-10 in the hall Expo X1 to discover how our clinically validated AI solutions support radiologists in daily practice.
ECR 2026 Lunit Abstract Information
|
No. |
Session |
Session Title |
Session Type |
|
1 |
RPS 1402 |
Oral Presentation |
|
|
2 |
RPS 1402 |
Oral Presentation |
|
|
3 |
RPS 1002 |
Oral Presentation |
|
|
4 |
RPS 1905 |
AI performance on an interval cancer mammography dataset and its role in audit triage |
Oral Presentation |
|
5 |
RPS 1905 |
Oral Presentation |
|
|
6 |
RPS 1102 |
Oral Presentation |
|
|
7 |
RPS 1102 |
Oral Presentation |
|
|
8 |
RPS 1102 |
Oral Presentation |
|
|
9 |
RPS 1102 |
Oral Presentation |
|
|
10 |
RPS 1105 |
Oral Presentation |
|
|
11 |
RPS 1905 |
Oral Presentation |
|
|
12 |
RPS 805 |
Oral Presentation |
|
|
13 |
RPS 2102 |
Oral Presentation |
|
|
14 |
Automated Mammography Breast Positioning Assessment After Surgical Intervention |
Poster |
|
|
15 |
AI on the Front Line: Real-World Performance of Lunit INSIGHT CXR in a high-volume Tertiary Centre |
Poster |
|
|
16 |
Evolution of Breast Cancer Screening: From Human-Only Double Reading to AI-Integrated Single Reading – flaggings, recall rates and increased cancer detection rate |
Poster |
|
|
17 |
Real world monitoring of the threshold for an AI-algorithm in breast cancer screening and different mammography equipments |
Poster |
|
|
18 |
More Than Hot Air: Enhancing Pneumothorax Detection with AI-Assisted Interpretation |
Poster |
|
|
19 |
Commercial artificial intelligence tools for chest radiography: diagnostic performance and workflow effects in a prospective crossover |
Poster |
|
|
20 |
Normal-Flagging for Chest Radiography: pre-clinical evaluation of a dual-engine AI for safe triage and workflow efficiency |
Poster |
|
|
21 |
Evaluating BI-RADS 4 Mammographic Lesions with Artificial Intelligence: Accuracy and Risk Stratification |
Poster |
About Lunit
Founded in 2013, Lunit (KRX: 328130) is a global leader on a mission to conquer cancer through AI. Our clinically validated solutions span medical imaging, breast health, and biomarker analysis—empowering earlier detection, smarter treatment decisions, and more precise outcomes across the cancer care continuum.
Lunit offers a comprehensive suite spanning risk prediction and early detection to precision oncology. Our FDA-cleared Lunit INSIGHT suite and breast health solutions support cancer screening in thousands of medical institutions worldwide, while the Lunit SCOPE platform is used in research partnership with global pharma and laboratory leaders for biomarker research, and companion diagnostic development.
Trusted by over 10,000 sites in more than 65 countries, Lunit combines deep medical expertise with continuously evolving datasets to deliver measurable impact—for patients, clinicians, and researchers alike. Headquartered in Seoul with global offices, Lunit is driving the worldwide fight against cancer. Learn more at lunit.io/en.
|
[1] In the NHS Breast Screening Programme (NHSBSP), interval cancer cases are retrospectively reviewed and classified into three categories: Category 1, difficult to detect at the time of screening; Category 2, potentially detectable in retrospect; and Category 3, visible in retrospect and considered to have been missed. |
