Enhancing Breast Cancer Detection through Advanced AI-Driven Ultrasound Technology: A Comprehensive Evaluation of...

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Oncology

Background:

• Breast cancer remains a leading cause of mortality globally.
• Accurate and early detection is crucial for effective treatment and improved patient outcomes.
• Current diagnostic methods, including ultrasound, can benefit from AI-driven enhancements.

Purpose of the Study:

• To evaluate the diagnostic performance of Vis-BUS, an AI-driven ultrasound tool for breast cancer detection.
• To assess the ability of Vis-BUS to differentiate between benign and malignant breast lesions.
• To identify factors influencing the accuracy of AI-based breast lesion classification.

Main Methods:

• Retrospective analysis of 258 breast ultrasound examinations.
• Application of Lesion Detection AI (LD-AI) and Lesion Analysis AI (LA-AI) to b-mode ultrasound images.
• Generation of Cancer Probability Scores (CPS) and assessment of diagnostic accuracy using AUROC and AUPRC.

Main Results:

• Vis-BUS demonstrated high diagnostic accuracy with an AUROC of 0.964 and AUPRC of 0.967.
• Factors like 'Fatty' lesion density (OR 27.7781) and BI-RADS category C5 (OR 10.173) significantly influenced classification accuracy.
• Lesion size and 'Unknown' density were associated with lower classification accuracy, while microcalcifications showed a moderate influence.

Conclusions:

• Vis-BUS significantly enhances diagnostic precision in breast cancer screening.
• The AI tool shows strong potential to support clinical decision-making.
• Further optimization is recommended for lesion density characterization and microcalcification detection to improve overall performance.