Artificial intelligence for breast cancer screening: breathtaking results and a word of caution

In The Lancet Oncology, Kristina Lång and colleagues report a prespecified clinical safety analysis of the MASAI trial, 1 Lång K Josefsson V Larsson AM et al. Artificial intelligence-supported screen reading versus standard double reading in the Mammography Screening with Artificial Intelligence (MASAI) trial: a clinical safety analysis of a randomised, controlled, non-inferiority, single-blinded, screening accuracy study. Lancet Oncol. 2023; 24: 936-944 Google Scholar investigating the use of artificial intelligence (AI) as a supplement to radiologist reading of mammograms in breast cancer screening. Conducted within the Swedish national breast screening programme, participants were randomly assigned (1:1) to either AI-supported screening (intervention group) or standard double reading without AI (control group). In the intervention group, the AI algorithm was used to triage screening mammograms to provide a level of risk of breast cancer from 1 to 10, with 10 having the highest risk. Screening examinations given a risk score of 1–9 were assigned to AI-assisted single reading, and those with a score of 10 were assigned to AI-assisted double reading. In the analysis, which was prespecified to occur after 80 000 women had been enrolled, there were 244 screen-detected cancers (6·1 per 1000 participants) and 46 345 screen readings in the intervention group (n=39 996 participants), and 203 cancers (5·1 per 1000) and 83 231 screen readings in the control group (n=40 024). Compared with the control group, the screen-reading workload of radiologists was reduced by 44·3% using AI-assisted techniques. Artificial intelligence-supported screen reading versus standard double reading in the Mammography Screening with Artificial Intelligence trial (MASAI): a clinical safety analysis of a randomised, controlled, non-inferiority, single-blinded, screening accuracy studyAI-supported mammography screening resulted in a similar cancer detection rate compared with standard double reading, with a substantially lower screen-reading workload, indicating that the use of AI in mammography screening is safe. The trial was thus not halted and the primary endpoint of interval cancer rate will be assessed in 100 000 enrolled participants after 2-years of follow up. Full-Text PDF Digital solutions for early breast cancer detectionBy 2040 , more than 3 million new cases of breast cancer and 1 million deaths are estimated to occur each year worldwide. Identifying those at greatest risk of developing life-threatening breast cancers is a public-health priority. In this issue of The Lancet Digital Health, Ash Kieran Clift and colleagues developed and evaluated four prognostic models to predict 10-year risk of breast cancer mortality in women aged 20–90 years. The best performing model could be used to identify women at high risk of mortality, including those too young for breast cancer screening. Full-Text PDF Open Access