Abstrakt: |
Population-based breast cancer screening programs using mammography have been effective in reducing breast cancer mortality. However, these programs have limitations, such as false negatives and false positives. Artificial intelligence (AI) algorithms based on deep learning have been developed to assist in breast cancer screening. These algorithms can identify lesions, assign a degree of suspicion to findings, and classify studies based on the likelihood of cancer. Retrospective studies have shown that AI can safely reduce the reading workload without compromising cancer detection or increasing false-positive recalls. Prospective studies have also supported the use of AI in breast cancer screening, although there are limitations. The application of AI in screening has the potential to extend screening to populations with a lack of breast radiologists, but ethical and legal issues must be considered. [Extracted from the article] |