AI-assisted Method for Efficiently Generating Breast Ultrasound Screening Reports.

Area of Science:

• Medical Imaging
• Artificial Intelligence in Healthcare
• Radiology

Background:

• Dense breast cancer screening commonly utilizes ultrasound.
• Manual report generation is time-consuming, labor-intensive, and prone to errors.
• Need for efficient and accurate reporting in breast ultrasound screening.

Purpose of the Study:

• To develop an AI pipeline for automated generation of breast ultrasound screening reports.
• To enhance clinical screening efficiency and minimize repetitive report writing.
• To assist physicians in the early detection of dense breast cancer.

Main Methods:

• A novel AI pipeline was developed to generate reports from ultrasound images.
• The system was trained and validated on a dataset of 4809 breast tumor instances.
• AI generated personalized preliminary reports, requiring physician review and adjustment.

Main Results:

• The AI pipeline significantly improved physician work efficiency by up to 90%.
• Reduced repetitive tasks associated with manual report writing.
• Demonstrated effectiveness in generating preliminary reports, particularly for common benign and normal cases.

Conclusions:

• AI-powered personalized report generation is clinically valuable.
• Physician acceptance is higher for AI-generated reports compared to template-based systems.
• The proposed pipeline offers a promising solution for optimizing breast ultrasound screening workflows.