AI Software Enhances Breast Cancer Screening Accuracy and Efficiency

Breast cancer is a prevalent disease globally, with China being no exception to its impact. With breast cancer ranking as the second most common cancer worldwide, early detection is crucial for improving prognosis. The Breast Imaging Reporting and Data System (BI-RADS) categorizes breast lesions based on malignancy likelihood, emphasizing the significance of early diagnosis and treatment. In China, breast ultrasound examination is widely accepted due to its non-invasive nature, affordability, and real-time visualization capabilities.

However, the exponential growth in medical imaging data poses challenges for sonographers, whose workload has significantly increased over the years. The scarcity of skilled sonographers and the overwhelming workload have led to errors in manual analysis, highlighting the need for more efficient and accurate diagnostic methods. Artificial intelligence (AI) has emerged as a promising solution in medical imaging research, excelling at identifying complex patterns and providing quantitative assessments of lesions.

AI has been successfully applied in various disease diagnoses, including breast cancer. Studies have shown that AI can enhance diagnostic accuracy and reduce workload in breast cancer screening. AI software can automate feature analysis in breast ultrasound images, reducing the time needed for interpretation compared to sonographers. By assisting sonographers in analyzing breast nodule features, AI software can significantly improve healthcare delivery, particularly in underserved areas with limited access to skilled sonographers.

In a recent study evaluating the accuracy of breast ultrasound image analysis software, the agreement between the software and sonographers in assessing features was high. The software demonstrated high accuracy in analyzing features such as echotexture, echo pattern, orientation, shape, margin, calcification, and posterior echo attenuation. The study also found that the software significantly reduced analysis time compared to sonographers, showcasing its potential to enhance diagnostic accuracy and reduce sonographers’ workloads.

Future research in breast imaging software could focus on refining algorithms to analyze dynamic video and expanding training datasets to improve applicability. Additionally, exploring AI assistance for less experienced sonographers and evaluating its impact on diagnostic accuracy could further advance the field. Overall, AI software presents a promising solution to streamline breast cancer screening processes, improve diagnostic accuracy, and ensure high-quality care for underserved populations.