In a groundbreaking study published in the Lancet Oncology journal, researchers have found that artificial intelligence (AI) can detect breast cancer in mammograms at a similar rate to that of two experienced radiologists. The study compared the performance of AI-supported breast cancer screening with standard care, presenting promising results that could revolutionise breast cancer detection and diagnosis.
The interim study involved more than 80,000 women from Sweden, with an average age of 54. Half of the participants underwent standard care, with their mammograms assessed by two radiologists. The other half received AI-supported screening, followed by interpretation by one or two radiologists. The results showed that the AI-supported group detected 244 cases of cancer, while the standard care group recalled 203 women with cancer.
Furthermore, the use of AI did not lead to an increase in false positives, where scans are incorrectly diagnosed as abnormal. The false-positive rate was consistent at 1.5% for both the AI-supported group and the radiologist-assessed group.
One of the most significant findings of the study was that the use of AI could significantly reduce the workload of radiologists. Researchers observed a staggering 44% reduction in the screen-reading workload for radiologists in the AI-supported group, with 36,886 fewer screen readings compared to the standard care group.
Lead author Dr. Kristina Lang, from Lund University in Sweden, emphasised that the results are highly promising, but additional research and trials are needed before implementing AI in mammography screening fully. While AI-supported screening requires at least one radiologist for detection, it has the potential to eliminate the need for double reading of most mammograms. This would alleviate the burden on radiologists, allowing them to focus on more advanced diagnostics and reducing waiting times for patients.
The implications of this study are profound, particularly in countries facing a radiologist shortage. AI has the potential to assist radiologists in accurately detecting breast cancer, leading to earlier diagnoses and potentially saving more lives. The NHS has already taken note of the study’s findings and is actively exploring how AI can improve breast screening by rapidly analysing complex images on a large scale.
More news from Reading
