The study led by a research team from
Study Highlights
Lunit INSIGHT CXR achieved the highest AUC (Area Under the Curve) of 0.93 in lung nodule detection, surpassing all other AI vendors and human readers (Mean AUC 0.81).
Lunit INSIGHT CXR achieved the highest AUC for all categories of nodule conspicuity and size (diameter).
The sensitivity of Lunit INSIGHT CXR reached 89%, outperforming all other AI algorithms, while maintaining specificity comparable to human readers at 80%.
Study Details
The study used radiographs from 386 patients, including 144 who had at least one nodule according to the reference standard CT image and were therefore considered true positive nodule cases; the remaining 242 were considered controls.
17 human readers, consisting of radiologists and radiology residents with varying experience levels, participated in the study. The mean AUC for human readers stood at 0.81, with a mean sensitivity of 71% and a mean specificity of 80%.
Through the evaluation of seven commercially available CE-marked lung nodule detection algorithms on chest radiographs, Lunit INSIGHT CXR emerged as the benchmark for excellence in standalone lung nodule detection, with the highest AUC and sensitivity, and a comparable specificity to human readers.
'These results are a reflection of
About
Founded in 2013,
As a medical AI company grounded on clinical evidence, our findings are presented in major peer-reviewed journals, such as the
After receiving FDA clearance and the CE Mark, our flagship Lunit INSIGHT suite is clinically used in approximately 3,000+ hospitals and medical institutions across 40+ countries.
Contact:
Email: contact@lunit.io
Email: ir@lunit.io
(C) 2024 Electronic News Publishing, source