- Lunit INSIGHT CXR software to be embedded in Agfa's MUSICA® Workstation to improve chest pathologies detection
- Lunit to showcase poster presentation on Lunit INSIGHT CXR at ECR 2023 as part of 6 presentations
SEOUL, South Korea, Feb. 27, 2023 /PRNewswire/ -- Lunit (KRX:328130.KQ) and Agfa Radiology Solutions today announced the integration of Lunit's AI solution for chest radiology, Lunit INSIGHT CXR, in the MUSICA® Workstation. Agfa will showcase a demonstration of the integrated product at the upcoming European Congress of Radiology (ECR) 2023, to be held in Vienna, Austria, on March 1-5.
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.[1]
The new critical findings workflow for clinicians by Lunit and Agfa provides automatic notifications to radiographers when a critical, life-threatening disease is detected, such as pneumothorax, pleural effusion, and pneumoperitoneum. Thus, hospitals can improve patient safety through timely communication with clinicians, especially in outpatient settings where patients return home after x-ray exams.
Agfa's MUSICA® workstation is an image processing software that offers a single intuitive interface covering general radiography, fluoroscopy, mammography, and full leg/full spine (FLFS). Lunit INSIGHT CXR's integration is the first in a range of new functionalities to be embedded in the MUSICA® Workstation to assist clinicians and radiologists with pathology detection.
"With the integration of Lunit INSIGHT CXR, we are continuing to deliver on our vision to transform digital radiology," said Paul Danini, Global Software Marketing and Product Manager at Agfa. "Lunit has a strong clinical validation track record and is registered in Europe and also more than 19 countries outside Europe, making it an excellent partner for developing solutions that bring our customers intelligent and definitive answers to their daily needs."
"We are thrilled to team with Agfa once again to create clinical value through AI and ensure a healthier life for patients," said Brandon Suh, CEO of Lunit. "We believe that the integration of Lunit's AI in Agfa's advanced medical imaging workstation will help facilitate more effective clinical workflows in various hospital situations."
The integrated solution will be publicly demonstrated for the first time at Agfa's booth #207.
Lunit also announced that it will showcase a poster presentation highlighting Lunit INSIGHT CXR's ability to differentiate between normal and abnormal chest radiographs—one of 6 presentations Lunit will be delivering at ECR 2023.
The study aimed to automate the detection of normal vs. abnormal chest radiographs—thus removing them from radiologists' worklist. Lunit INSIGHT CXR was used to retrospectively analyze 1,671 x-ray images from two large hospitals in the Netherlands. The data were collected consecutively—without any data enrichment, thus mimicking real-world use cases. The x-ray images were initially reviewed and classified by a radiologist according to 5 categories: normal (548), clinically irrelevant findings (281), clinically relevant findings (327), urgent findings (494), and critical findings (21).
Lunit INSIGHT CXR showed a very high AUC of 0.908 for detecting normal cases and at a high specificity of 95%, it was simulated that around half of the normal radiographs could be safely removed from the worklist. Therefore, the results show that the workload for the radiologist could be significantly reduced with the usage of AI.
Lunit will showcase a total of 6 abstracts at this year's ECR meeting, with 4 oral presentations and 2 poster presentations.
[1] https://www.lunit.io/en/products/cxr
Photo - https://mma.prnewswire.com/media/2010208/Lunit_INSIGHT_CXR_Lunit_s_AI_solution_chest_radiology.jpg
Logo - https://mma.prnewswire.com/media/1576401/Lunit_CI_Logo.jpg
Share this article