NA-MIC Project Weeks
Back to Projects List
SlicerUltrasound Extension development
Key Investigators
- Deepa Krishnaswamy (Brigham and Women's Hospital, USA)
- Tina Kapur (Brigham and Women's Hospital, USA)
- Tamas Ungi (ClaroNav, Canada)
- David Dinh (SlicerUltrasound Team, USA)
- Martin Bellehumeur (Radical Imaging, Germany)
Project Description
SlicerUltrasound is a 3D Slicer extension that currently includes two modules: 1) Anonymize, which removes both metadata-embedded PHI and burned-in text from DICOM images, and 2) Annotate, which supports expert labeling of image findings.
The Anonymize module has been used at multiple hospitals (BWH, Lahey, Indiana Methodist to remove PHI from patient exams. Users import DICOM ultrasound images from their local folder, and apply probe-specific masking templates to remove burned-in identifiers. Users specify the transducer type—curvilinear or phased array—which determines the expected fan shape. By marking three or four points on the image, the module interpolates the imaging sector and masks any visual PHI outside this region while preserving diagnostically relevant content.
The Annotate module allows experts to annotate lung ultrasound video clips, focusing on features such as pleura lines and B-lines. It provides an intuitive interface for frame-by-frame annotation, supports multiple raters, and saves annotation data for future research and machine learning.
Our goals during the project week are to:
- Obtain feedback from developers and clinicians
- Implement changes to the extension
- Connect with and talk to others about the MONAI Ultrasound Working Group - Annotation and Anonymization subgroup.
Objective
- Demonstrate the current extension to Slicer core developers and obtain feedback
- Demonstrate to clinicians and obtain feedback
- Integrate changes to the extension
Approach and Plan
- Describe specific steps of what you plan to do to achieve the above described objectives.
Progress and Next Steps
- Many people provided useful feedback about possible improvements to the modules.
- We also looked into the current OHIF viewer, which already allows for users to annotate pleural lines and B-lines. Martin added functionality needed to save additional annotations.
Illustrations
Anonymize module
Annotate module
Original OHIF app for annotation of pleural lines and B-lines - on our anonymized lung US data:
Enhancements to the OHIF app for annotation of pleural lines and B-lines - on sample cardiac US data. This helps with the clinical workflow, where step 1) the clinician scans through the ultrasound clip and quickly pick the frame that has the maximum number of B-lines, and step 2) the clinician or a fellow can come back later and annotate the pleural lines and B-lines.