⚡ Quick Summary
A recent study introduced the whole-lesion-aware network based on freehand ultrasound video (WAUVE) for breast cancer assessment, demonstrating a remarkable AUC of 0.8998. This innovative approach shows comparable diagnostic performance to experienced radiologists, enhancing the potential for real-time clinical applications.
🔍 Key Details
- 📊 Dataset: 2912 videos from 2771 patients, plus 190 videos from 175 patients for external validation
- 🧩 Features used: Freehand ultrasound video
- ⚙️ Technology: WAUVE, compared with 2D-ResNet50 and TimeSformer models
- 🏆 Performance: WAUVE achieved an AUC of 0.8998 in external validation
🔑 Key Takeaways
- 📊 WAUVE outperformed the 2D-ResNet50 model and matched the TimeSformer model in diagnostic accuracy.
- 💡 The model showed a sensitivity of 97.39% and specificity of 49.33%.
- 👩🔬 Radiologists using WAUVE improved their average specificity by 6.67%.
- 🏆 Consistency among radiologists increased from 0.807 to 0.838 with WAUVE assistance.
- 🌍 Study conducted across multiple centers, enhancing the robustness of findings.
- 🔬 Clinical application of WAUVE is promising for real-time breast cancer risk assessment.
📚 Background
The assessment of breast cancer through traditional ultrasound imaging has faced challenges due to the operator-dependence of standardized image acquisition and the limitations of static images in capturing the full extent of lesions. The introduction of artificial intelligence (AI) in this domain aims to enhance diagnostic accuracy and streamline workflows, making it essential to explore innovative approaches like WAUVE.
🗒️ Study
This prospective multicenter study developed the WAUVE model using a comprehensive dataset of 2912 freehand ultrasound videos collected from May 2020 to August 2022. The model’s performance was validated against static and dynamic models, with an additional independent validation set of 190 videos from December 2022 to April 2023, ensuring a robust evaluation of its diagnostic capabilities.
📈 Results
The WAUVE model achieved an impressive AUC of 0.8998 in the external validation set, indicating its strong predictive capabilities. When compared to four experienced radiologists, WAUVE’s sensitivity and specificity were comparable, with only minor differences in performance metrics, suggesting that the model can serve as a valuable tool in clinical settings.
🌍 Impact and Implications
The findings from this study highlight the potential of WAUVE to transform breast cancer assessment by providing a real-time, accurate risk score. This could lead to improved patient outcomes through timely interventions and more informed clinical decisions. The integration of such AI-driven technologies into routine practice may enhance the overall efficiency and effectiveness of breast cancer diagnostics.
🔮 Conclusion
The WAUVE model represents a significant advancement in breast cancer assessment, demonstrating that non-standardized ultrasound scanning can yield results comparable to those of experienced radiologists. This study underscores the promising future of AI in clinical applications, paving the way for more precise and efficient diagnostic tools in healthcare. Continued research and development in this area are essential to fully realize the benefits of such technologies.
💬 Your comments
What are your thoughts on the integration of AI in breast cancer assessment? We would love to hear your insights! 💬 Leave your comments below or connect with us on social media:
Whole-lesion-aware network based on freehand ultrasound video for breast cancer assessment: a prospective multicenter study.
Abstract
BACKGROUND: The clinical application of artificial intelligence (AI) models based on breast ultrasound static images has been hindered in real-world workflows due to operator-dependence of standardized image acquisition and incomplete view of breast lesions on static images. To better exploit the real-time advantages of ultrasound and more conducive to clinical application, we proposed a whole-lesion-aware network based on freehand ultrasound video (WAUVE) scanning in an arbitrary direction for predicting overall breast cancer risk score.
METHODS: The WAUVE was developed using 2912 videos (2912 lesions) of 2771 patients retrospectively collected from May 2020 to August 2022 in two hospitals. We compared the diagnostic performance of WAUVE with static 2D-ResNet50 and dynamic TimeSformer models in the internal validation set. Subsequently, a dataset comprising 190 videos (190 lesions) from 175 patients prospectively collected from December 2022 to April 2023 in two other hospitals, was used as an independent external validation set. A reader study was conducted by four experienced radiologists on the external validation set. We compared the diagnostic performance of WAUVE with the four experienced radiologists and evaluated the auxiliary value of model for radiologists.
RESULTS: The WAUVE demonstrated superior performance compared to the 2D-ResNet50 model, while similar to the TimeSformer model. In the external validation set, WAUVE achieved an area under the receiver operating characteristic curve (AUC) of 0.8998 (95% CI = 0.8529-0.9439), and showed a comparable diagnostic performance to that of four experienced radiologists in terms of sensitivity (97.39% vs. 98.48%, p = 0.36), specificity (49.33% vs. 50.00%, p = 0.92), and accuracy (78.42% vs.79.34%, p = 0.60). With the WAUVE model assistance, the average specificity of four experienced radiologists was improved by 6.67%, and higher consistency was achieved (from 0.807 to 0.838).
CONCLUSION: The WAUVE based on non-standardized ultrasound scanning demonstrated excellent performance in breast cancer assessment which yielded outcomes similar to those of experienced radiologists, indicating the clinical application of the WAUVE model promising.
Author: [‘Han J’, ‘Gao Y’, ‘Huo L’, ‘Wang D’, ‘Xie X’, ‘Zhang R’, ‘Xiao M’, ‘Zhang N’, ‘Lei M’, ‘Wu Q’, ‘Ma L’, ‘Sun C’, ‘Wang X’, ‘Liu L’, ‘Cheng S’, ‘Tang B’, ‘Wang L’, ‘Zhu Q’, ‘Wang Y’]
Journal: Cancer Imaging
Citation: Han J, et al. Whole-lesion-aware network based on freehand ultrasound video for breast cancer assessment: a prospective multicenter study. Whole-lesion-aware network based on freehand ultrasound video for breast cancer assessment: a prospective multicenter study. 2025; 25:75. doi: 10.1186/s40644-025-00892-y