AI-Powered Synthesis of Structured Multimodal Breast Ultrasound Reports Integrating Radiologist Annotations and Deep Learning Analysis
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.This study presents a semi-automatic method for breast ultrasound (US) reporting, significantly reducing radiologist workload. The AI-powered system streamlines report generation, improving efficiency and patient care.
Area Of Science
- Medical Imaging
- Artificial Intelligence in Healthcare
- Radiology
Background
- Breast cancer is a leading global health concern for women.
- B-mode ultrasound (US) is a crucial, non-invasive tool for early breast cancer detection.
- Manual generation of breast US reports is time-consuming for radiologists.
Purpose Of The Study
- To develop and evaluate a semi-automatic method for streamlining breast US report generation.
- To reduce the reporting burden on radiologists, allowing more time for patient care.
- To integrate deep learning analysis with radiologist annotations for comprehensive reports.
Main Methods
- A semi-automatic method combining radiologist annotations and deep learning analysis of multimodal US images.
- Key modules: Image Classification using Visual Features (ICVF), Type Classification via Deep Learning (TCDL), and Automatic Report Structuring and Compilation (ARSC).
- Utilized DenseNet-121 for classifying tissue stiffness in breast US shear-wave elastography (SWE-mode) images.
Main Results
- Reduced average report generation time to 3.8 minutes.
- Achieved perfect accuracy in matching ground truth reports for suspicious masses.
- Deep learning model achieved 0.865 accuracy, 0.868 precision, 0.847 recall, 0.856 F1-score, and 0.92 AUC for SWE-mode image classification.
Conclusions
- The proposed semi-automatic method significantly enhances the efficiency of breast US report generation.
- The AI-driven approach improves accuracy and reduces radiologist workload, leading to better patient care and satisfaction.
- This technology enables radiologists to focus more on clinical decision-making and patient interaction.
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
