Lunit’s artificial intelligence software for breast cancer screening demonstrated promising results in detecting cancer risks years before traditional diagnosis, according to two large-scale European studies.
The research, conducted at Norway’s Cancer Registry and Denmark’s Odense University Hospital, analyzed the Korean company’s Lunit INSIGHT MMG technology across different screening scenarios. The Norwegian study, examining data from over 116,000 women, found the AI system could identify potential cancer development up to six years before conventional detection methods.
The software assigns risk scores from 0 to 100 to mammograms, with higher scores indicating greater cancer probability. Women who later developed cancer showed consistently higher AI scores years before diagnosis, with the score gap widening over time.
In Denmark, researchers tested the AI system’s potential to reduce radiologist workload. Using nearly 250,000 mammograms, the study showed the technology could cut reading volume by almost 50% while maintaining accuracy levels.
Lunit CEO Brandon Suh said the findings demonstrate concrete benefits for healthcare providers, though he didn’t provide specific pricing or implementation timelines for the technology.
The study results mark a step forward in AI-assisted cancer screening, but broader clinical validation and cost-effectiveness analysis may be needed before widespread adoption.
