OBJECTIVES:

Breast-cancer detection faces a ‘four-fold’ screening challenge in women with dense breasts. Firstly, greater breast density is associated with an elevated cancer risk. Secondly, X-ray mammography (XM) is less accurate at detecting breast cancer in women with dense breasts (Boyd et al. 2007). Thirdly, while guidelines recommend supplemental screening for women with dense breasts (Mann et al 2017), there is a lack of clarity on which supplemental modalities are preferred, and there is restricted access to supplemental modalities. Finally, assessment and reporting of breast density in XM is not performed, going against guideline recommendations across Europe (2015-2022, DenseBreast-info, Inc).

This analysis outlines the potential for artificial intelligence (AI) technologies to improve health economic outcomes when using supplemental screening modalities in women with dense breasts.

METHODS:

A decision tree linked to a Markov chain was developed to model the cost effectiveness of AI technologies within the breast cancer screening pathway. The impact of potential future AI technologies, which alter the screening accuracy values for XM and supplemental MRI, were evaluated.

RESULTS:

AI technologies that enhance the diagnostic accuracy of supplemental screening modalities have the potential to improve clinical outcomes and reduce costs. This potential is driven by reductions in the number of patients receiving false-positive or false-negative diagnoses and through the ability to identify small lesions that, when undetected during screening, could develop into cancerous tumors.

CONCLUSIONS:

Emerging AI technologies have the potential to enhance screening accuracy in women with dense breasts. This analysis suggests that the application of AI can play a valuable role in optimizing screening pathways and overcoming sources of inefficiency, for example through increased cancer detection for XM or through reduced false-positive and false-negative MRI results.