Related Experiment Video
Updated: Jun 23, 2026
08:05
Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
Published on: December 19, 2020
Can Crowdsourced Annotations Improve AI-based Congestion Scoring For Bedside Lung Ultrasound?
Ameneh Asgari-Targhi1,2, Tamas Ungi3,2, Mike Jin1,4,2
1Brigham and Women's Hospital and Harvard Medical School, Boston, MA, USA.
Medical Image Computing and Computer-Assisted Intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention
|June 22, 2026
Summary
A new lung ultrasound (LUS) method accurately quantifies pulmonary congestion by segmenting B-line artifacts. This approach uses crowdsourced data to standardize interpretation, improving diagnostic accuracy in critical care settings.
Area of Science:
- Medical Imaging
- Ultrasound Technology
- Pulmonary Medicine
Background:
- Lung ultrasound (LUS) is vital for assessing pulmonary congestion in acute care.
- B-line artifacts in LUS are key indicators but suffer from interpretation variability.
- Standardization is needed for reliable LUS analysis in time-sensitive situations.
Purpose of the Study:
- To introduce a novel B-line segmentation method for objective pulmonary congestion scoring.
- To enhance the standardization and accuracy of LUS interpretation.
- To develop a reliable tool for resource-limited settings.
Main Methods:
- Developed a B-line segmentation algorithm for LUS images.
- Utilized a large dataset of 31,000 B-line annotations from over 550,000 crowdsourced opinions.
- Trained and validated the model on LUS images from 299 patients.
Main Results:
- Achieved 94% accuracy in B-line counting (within a margin of 1) on a test set of 100 patients.
- Demonstrated significant improvement in objective B-line quantification.
- The method provides a standardized congestion score.
Conclusions:
- The new B-line segmentation method offers a standardized and accurate approach to assessing pulmonary congestion via LUS.
- Combining extensive crowdsourced data with advanced algorithms enhances LUS diagnostic reliability.
- This tool has the potential to improve clinical decision-making in emergency and acute care settings.