Using commercial AI-based mammography analysis software to improve breast US interpretations

This study evaluated the use of commercial AI-based mammography software to improve breast ultrasound (US) lesion interpretations. The AI software, which provides an AI malignancy score ranging from 0 to 100, was tested on 1,109 breasts that underwent both mammography and US-guided breast biopsy. The AI software showed an area under the curve (AUROC) of 0.79 for distinguishing benign from cancerous cases. When integrated into prediction models, it improved accuracy, with the integrated model achieving an AUROC of 0.85 in the development cohort, outperforming non-AI models. The study suggests that AI-enhanced mammography analysis could be a valuable tool in clinical decision-making for managing US-detected breast lesions.

Key points:

• Breast US has high rates of false-positive interpretations.
• A commercial AI-based mammography analysis software could distinguish mammograms having benign outcomes from those revealing cancers after US-guided breast biopsy.
• A commercial AI-based mammography analysis software may improve interpretations for breast US-detected lesions.

Article: Use of a commercial artificial intelligence-based mammography analysis software for improving breast ultrasound interpretations

Authors: Hee Jeong Kim, Hak Hee Kim, Ki Hwan Kim, Ji Sung Lee, Woo Jung Choi, Eun Young Chae, Hee Jung Shin, Joo Hee Cha & Woo Hyun Shim