Using AI, breast radiologists in Denmark have enhanced breast cancer screening performance and decreased the rate of false-positive results. The study’s findings were released in Radiology, a journal of the Radiological Society of North America (RSNA). Mammography has been successful in reducing breast cancer mortality, but it also poses the risk of false-positive findings. In recent years, AI has been used to improve the accuracy of mammography interpretations.ars, researchers have looked into the use of AI systems in screening. “We think AI has the potential to enhance screening performance,” said Andreas D. Lauritzen, Ph.D., a post-doctoral student at the University of Copenhagen and researcher at Gentofte Hospital in Denmark. When used to prioritize likely normal screening results or provide decision support, AI can also significantly decrease the workload of radiologists. “Population-based screening with mammography reduces breast cancer mortality, but it places a substantial workload on radiologists who must read a large number of mammograms, the majority of which don’t warrant a recaDr. Lauritzen emphasized the importance of managing the workload of radiologists in screening programs to ensure the overall well-being of the patients. The use of double reading in screening programs can lead to an increased workload for radiologists, but it is also beneficial in improving cancer detection rates and reducing false-positive recalls.
Dr. Lauritzen and his colleagues conducted a study to compare the workload and screening performance before and after the implementation of AI in two groups of women who underwent mammography screening. The study, which was retrospective, focused on two cohorts of women aged 50 to 69 who received biennial mammography screening in the Capital Region of Denmark.
The first group consisted of women whose mammograms were read by two radiologists.Between October 2020 and November 2021, mammograms were assessed without the use of AI. From November 2021 to October 2022, a group of women had their mammograms initially analyzed by AI. Mammograms identified as likely normal by AI were then reviewed by one of 19 specialized full-time breast radiologists, while the remaining mammograms were reviewed by two radiologists with AI-assisted decision support.
The AI system used for screening was trained to identify and evaluate suspicious lesions and calcifications in mammograms using deep learning models. All women were included in the study.A total of 60,751 women who had mammographic screening were followed for at least 180 days. Invasive cancers and ductal carcinoma in situ (DCIS) found through screening were confirmed through needle biopsy or surgical specimens.
Out of the total, 60,751 women were screened without AI, and 58,246 women were screened with the AI system. In the AI implementation group, 66.9% (38,977) of the screenings were single-read, and 33.1% (19,269) were double-read with AI assistance.
Compared to screening without AI, screening with the AI system detected significantly more breast cancers (0.82% versus 0.70%) and had a lower false-positive rate (1.63% versus 2.39%)
According to Dr. Lauritzen, in the AI-screened group, the recall rate decreased by 20.5 percent and the radiologists’ reading workload was reduced by 33.4 percent.
The positive predictive value of AI screening was also higher than that of screening without AI (33.5% versus 22.5%). In the AI group, a higher percentage of invasive cancers detected were 1 centimeter or less in size (44.93% vs. 36.60%).
Dr. Lauritzen stated that “All screening performance indicators improved except for the node-negative rate, which showed no evidence of change.”
According to Dr. Lauritzen, more research is necessary to fully understand the implications of these findings.Assess the long-term results and prevent overdiagnosis from increasing.
“Radiologists usually have access to the previous mammograms of the women, but the AI system does not,” he explained. “That’s something we hope to address in the future.”
It’s also worth noting that not all countries have the same breast cancer screening protocols and time intervals. The breast cancer screening protocols in the U.S. differ from those used in Denmark.
Â
