AI breakthrough in breast cancer detection: New study claims over 17% improvement in screening accuracy

For biopsies ordered, as many as 65 per cent of AI-assisted readings confirmed cancer against 59 per cent without AI. The study also revealed that AI reduced workloads by enabling 43 per cent faster reading times without compromising on accuracy. 

Why is study important?

Early breast detection can save lives, and traditional screenings involve radiologists reviewing images—a time-consuming procedure prone to errors. With AI, radiologists can now identify suspicious cases and reduce the instances of missed detections. 

The participants were between the ages of 50 and 69, and they participated in mammography screenings between 2021 and 2023. Around half of the screenings used AI, while the rest relied on standard processes. The objective of the study was to see if AI could assist radiologists in finding more cancers without unnecessary recalls. In cancer detection, recall happens when mammogram findings need additional imaging or comparison to earlier mammograms. 

The role of AI 

Based on the study, AI helped with higher detection, fewer false alarms, and improved accuracy. Essentially, AI increased the success rate of identifying actual cancer cases among women who received recalled mammograms. The study shows how AI is designed to assist and not replace radiologists and can help with early detection. The AI-assisted screens flagged cases as normal and suspicious and relayed alerts for potential oversights, allowing radiologists to focus on critical cases and save time on routine cases.

The study shows that AI can make breast cancer screening faster, more accurate, and less stressful for the patients. It could potentially address the shortage of radiologists and improve the efficiency of the healthcare sector. 

The study is known as “Nationwide real-world implementation of AI for cancer detection in population-based mammography screening,” and it has been published in Nature Medicine.