A recent study compared the accuracy of breast ultrasound image analysis software with that of sonographers. The study aimed to assess the software’s capability to analyze breast nodule features and its potential to reduce sonographers’ workload. Breast ultrasound is crucial for early breast cancer detection in China, but the increasing imaging data volume poses challenges for sonographers. The study analyzed breast ultrasound images from two Shanghai hospitals, evaluating features like echotexture, echo pattern, orientation, shape, margin, calcification, and posterior echo attenuation.

Results showed a high agreement between the software and sonographers in assessing these features, with proportions of agreement ranging from 80.5% to 93.7%. The software significantly reduced analysis time compared to sonographers. While there was substantial agreement for echotexture, shape, and margin, moderate agreement was observed for echo pattern, orientation, and posterior echo attenuation. However, fair agreement was noted for calcification.

The study highlighted the potential of AI software to enhance healthcare delivery by automating feature analysis in breast ultrasound images, particularly in underserved areas lacking skilled sonographers. The software’s high accuracy and time efficiency make it a promising solution to alleviate sonographers’ workload and improve diagnostic accuracy. The next steps involve refining the software to analyze dynamic video, expanding training datasets, and exploring AI assistance for less experienced sonographers. This research underscores the role of AI in advancing breast cancer screening and diagnosis, especially in resource-limited settings.