Breast ultrasound, while effective in detecting mammographically occult cancers, suffers from high false-positive rates. This study introduces an AI system achieving radiologist-level accuracy in identifying breast cancer in ultrasound images. Trained on a vast dataset (288,767 exams, 5,442,907 images), the AI boasts an AUROC of 0.976, surpassing the average AUROC of ten board-certified radiologists (0.924). The AI, when used with radiologists, reduced false-positive rates by 37.3% and biopsies by 27.8% while maintaining sensitivity, demonstrating its potential to enhance breast ultrasound diagnosis.
Publisher
Nature Communications
Published On
Oct 26, 2021
Authors
Yiqiu Shen, Farah E Shamout, Jamie R Oliver, Jan Witowski, Kawshik Kannan, Jungkyu Park, Nan Wu, Connor Huddleston, Stacey Wolfson, Alexandra Millett, Robin Ehrenpreis, Divya Awal, Cathy Tyma, Naziya Samreen, Yiming Gao, Chloe Chhor, Stacey Gandhi, Cindy Lee, Sheila Kumari Subaiya, Cindy Leonard, Reyhan Mohammed, Christopher Moczulsk, Jaime Altabe, James Babb, Alana Lewin, Beatriu Reig, Linda Moy, Laura Heacock, Krzysztof J Geras
Tags
breast cancer
ultrasound
AI system
diagnosis
false positive
radiologist
Related Publications
Explore these studies to deepen your understanding of the subject.
