Fewer false positives, earlier detection: How AI is making mammograms more accurate

The results from a 2023 trial in Sweden, published in The Lancet journal, revealed that AI-supported breast cancer screening identified more women with clinically relevant cancers without a higher rate of false positives

Representative Image for Breast Cancer Representative Image

Artificial intelligence (AI) has become a significant pillar in modern medicine. The use of AI in Sweden has actually proven effective in the early detection of breast cancer and other diseases. 

The AI-supported mammography conducted in Sweden's national breast cancer screening programme revealed that it is more effective than the standard ones. 

The results from a 2023 trial published in The Lancet journal revealed that AI-supported breast cancer screening identified more women with clinically relevant cancers without a higher rate of false positives.  

The study also revealed that the women who underwent AI-supported screening were less likely to be diagnosed with more aggressive and advanced breast cancer in the following two years. 

The Mammography Screening with Artificial Intelligence (MASAI) showed that the AI use resulted in the detection of 20 per cent more cancers compared with standard screening. 

The workload for radiologists also reduced by 44 per cent. 

The study was conducted by researchers from Lund University and others in Sweden, Denmark, Norway and the Netherlands. 

For your daily dose of medical news and updates, visit: HEALTH

AI-supported screening improves the early detection of clinically relevant breast cancers, which leads to fewer aggressive or advanced cancers diagnosed in between screenings, said Dr Kristina Lang, a breast radiologist and clinical researcher from Lund University, Sweden. 

The AI system was trained, validated, and tested with more than two lakh examinations from multiple institutions from over 10 countries. 

Between April 2021 and December 2022, more than 105,900 women participated in a trial where they were randomly assigned to either receive AI-integrated breast cancer screenings or the standard procedure of having two radiologists review their results manually.

During the follow-up period of two years, 1.55 interval cancers ( these are tumours detected between scheduled screenings) per 1,000 women were detected in the AI-supported mammography group compared to the 1.76 per 1,000 women in the standard group.

Also, 81 per cent of cancer cases in the AI-supported mammography were detected at screening, compared to 74 per cent of cancer cases in the standard group.