A study with the involvement of Sweden's Lund University has revealed that incorporating AI-assisted image interpretation into breast cancer screening programs can significantly bolster screening detection abilities and diminish the occurrence of 'interval cancer'. The research encompassed around 106,000 women aged from 40 to 74, with a follow-up span exceeding two years. In the group where AI-assisted image reading was implemented, doctors' workload for image interpretation dropped by 44%. Simultaneously, the breast cancer detection rate surged by 29%, and the number of 'interval cancer' cases dropped by 12% compared to the control group. Moreover, there was a lower proportion of severe cases in the AI-assisted group. Notably, AI assistance did not lead to an uptick in 'false positives' and instead enhanced the 'sensitivity' of the screening process. This innovative technology has already been put into practice in certain areas of Sweden and other European regions, and there are expectations for its wider adoption in more countries that conduct breast cancer screenings.
