Related Experiment Video
Updated: Jun 11, 2025
Patient-Specific Polyvinyl Alcohol Phantom Fabrication with Ultrasound and X-Ray Contrast for Brain Tumor Surgery Planning
Published on: July 14, 2020
BOOSTING SKULL-STRIPPING PERFORMANCE FOR PEDIATRIC BRAIN IMAGES
William Kelley1,2, Nathan Ngo1,2, Adrian V Dalca1,2,3,4
1Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA 02129, USA.
Insights
We developed developmental SynthStrip (d-SynthStrip), a new pediatric skull-stripping tool. It accurately removes non-brain tissue from developing brains, outperforming existing methods for diverse pediatric neuroimaging datasets.
Area of Science:
- Neuroimaging
- Medical Image Analysis
- Pediatric Neuroscience
Background:
- Skull-stripping is crucial for analyzing brain images.
- Existing tools are often inadequate for pediatric neuroimaging due to developmental variations and image artifacts.
- There is a growing need for specialized pediatric skull-stripping methods to support large-scale developmental studies.
Purpose of the Study:
- To introduce developmental SynthStrip (d-SynthStrip), a novel skull-stripping tool specifically designed for pediatric brain images.
- To address the limitations of adult-centric tools in processing diverse pediatric neuroimaging data.
- To provide a robust and efficient solution for pediatric neuroimaging data processing.
Main Methods:
- Developed d-SynthStrip, a skull-stripping model based on an existing robust framework.
- Utilized a synthesis approach, exposing neural networks to highly variable images generated from label maps.
- Trained and validated the model on diverse pediatric brain image datasets, considering various scan types and age groups.
Main Results:
- d-SynthStrip demonstrated superior performance compared to existing pediatric skull-stripping baselines.
- The model achieved high accuracy across different scan types and pediatric age cohorts.
- d-SynthStrip offers a fast runtime, completing processing in under one minute, which is competitive with the fastest existing methods.
Conclusions:
- d-SynthStrip is a highly effective and efficient tool for pediatric brain image skull-stripping.
- The model's tailored approach addresses the unique challenges of pediatric neuroimaging.
- This tool will facilitate large-scale multi-institutional pediatric research by enabling reliable data processing.
Abstract:
Skull-stripping is the removal of background and non-brain anatomical features from brain images. While many skull-stripping tools exist, few target pediatric populations. With the emergence of multi-institutional pediatric data acquisition efforts to broaden the understanding of perinatal brain development, it is essential to develop robust and well-tested tools ready for the relevant data processing. However, the broad range of neuroanatomical variation in the developing brain, combined with additional challenges such as high motion levels, as well as shoulder and chest signal in the images, leaves many adult-specific tools ill-suited for pediatric skull-stripping. Building on an existing framework for robust and accurate skull-stripping, we propose developmental SynthStrip (d-SynthStrip), a skull-stripping model tailored to pediatric images. This framework exposes networks to highly variable images synthesized from label maps. Our model substantially outperforms pediatric baselines across scan types and age cohorts. In addition, the <1-minute runtime of our tool compares favorably to the fastest baselines. We distribute our model at https://w3id.org/synthstrip.
