乳房X光造影术的人工智能将攻击性乳腺癌减少了12%,并在瑞典的一项大型研究中改进了早期检测。 AI in mammography reduced aggressive breast cancers by 12% and improved early detection in a large Swedish study.

涉及100 000多名妇女的瑞典大型试验发现,在乳房X线造影检查中使用人工智能(AI),与放射科医生的标准重复阅读相比,将侵袭性、间歇性乳腺癌的发病率降低了12%。 A large Swedish trial involving over 100,000 women found that using artificial intelligence (AI) in mammography screening reduced the rate of aggressive, interval breast cancers by 12% compared to standard double-reading by radiologists. AI帮助早期发现更多癌症,81%在筛查期间发现,而74%在控制组中发现,并减少了入侵性肿瘤和大型肿瘤。 AI helped detect more cancers earlier, with 81% found during screening versus 74% in the control group, and reduced invasive and large tumors. AI系统对超过200 000次扫描进行了培训,为优化工作量和保持类似的假正率对案件进行分级处理,以优化工作量和保持类似的假正率。 The AI system, trained on over 200,000 scans, triaged cases to optimize review workload and maintained similar false positive rates. 研究人员说,AI支持放射学家,但不应取代他们,需要进一步研究,以确认不同人口群体的长期利益。 Researchers say AI supports radiologists but should not replace them, and further studies are needed to confirm long-term benefits across diverse populations.