Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Efficient Brain Extraction of MRI Scans with Mild to Moderate Neuropathology

Hjalti Thrastarson, Lotta M. Ellingsen

TL;DR

The paper tackles robust brain extraction on MRI in the presence of mild-to-moderate neuropathology, where existing methods falter at brain boundary delineation. It introduces a 3D U-net trained with a novel SDT-based loss, using downsampled inputs and silver-standard ground truth to enable fast, data-efficient learning without requiring gold-standard labels. Quantitative evaluations show state-of-the-art or comparable performance with strong outer-brain surface preservation, achieving a mean DSC of $0.964 \pm 0.006$ and ASSD of $1.4 \pm 0.2$ mm on held-out data, and competitive results on external datasets, with inference times under $20$ seconds per image. The approach is practical for preprocessing pipelines and adaptable to single-object segmentation tasks with limited high-quality ground truth, with code publicly available on GitHub.

Abstract

Skull stripping magnetic resonance images (MRI) of the human brain is an important process in many image processing techniques, such as automatic segmentation of brain structures. Numerous methods have been developed to perform this task, however, they often fail in the presence of neuropathology and can be inconsistent in defining the boundary of the brain mask. Here, we propose a novel approach to skull strip T1-weighted images in a robust and efficient manner, aiming to consistently segment the outer surface of the brain, including the sulcal cerebrospinal fluid (CSF), while excluding the full extent of the subarachnoid space and meninges. We train a modified version of the U-net on silver-standard ground truth data using a novel loss function based on the signed-distance transform (SDT). We validate our model both qualitatively and quantitatively using held-out data from the training dataset, as well as an independent external dataset. The brain masks used for evaluation partially or fully include the subarachnoid space, which may introduce bias into the comparison; nonetheless, our model demonstrates strong performance on the held-out test data, achieving a consistent mean Dice similarity coefficient (DSC) of 0.964$\pm$0.006 and an average symmetric surface distance (ASSD) of 1.4mm$\pm$0.2mm. Performance on the external dataset is comparable, with a DSC of 0.958$\pm$0.006 and an ASSD of 1.7$\pm$0.2mm. Our method achieves performance comparable to or better than existing state-of-the-art methods for brain extraction, particularly in its highly consistent preservation of the brain's outer surface. The method is publicly available on GitHub.

Efficient Brain Extraction of MRI Scans with Mild to Moderate Neuropathology

TL;DR

The paper tackles robust brain extraction on MRI in the presence of mild-to-moderate neuropathology, where existing methods falter at brain boundary delineation. It introduces a 3D U-net trained with a novel SDT-based loss, using downsampled inputs and silver-standard ground truth to enable fast, data-efficient learning without requiring gold-standard labels. Quantitative evaluations show state-of-the-art or comparable performance with strong outer-brain surface preservation, achieving a mean DSC of and ASSD of mm on held-out data, and competitive results on external datasets, with inference times under seconds per image. The approach is practical for preprocessing pipelines and adaptable to single-object segmentation tasks with limited high-quality ground truth, with code publicly available on GitHub.

Abstract

Skull stripping magnetic resonance images (MRI) of the human brain is an important process in many image processing techniques, such as automatic segmentation of brain structures. Numerous methods have been developed to perform this task, however, they often fail in the presence of neuropathology and can be inconsistent in defining the boundary of the brain mask. Here, we propose a novel approach to skull strip T1-weighted images in a robust and efficient manner, aiming to consistently segment the outer surface of the brain, including the sulcal cerebrospinal fluid (CSF), while excluding the full extent of the subarachnoid space and meninges. We train a modified version of the U-net on silver-standard ground truth data using a novel loss function based on the signed-distance transform (SDT). We validate our model both qualitatively and quantitatively using held-out data from the training dataset, as well as an independent external dataset. The brain masks used for evaluation partially or fully include the subarachnoid space, which may introduce bias into the comparison; nonetheless, our model demonstrates strong performance on the held-out test data, achieving a consistent mean Dice similarity coefficient (DSC) of 0.9640.006 and an average symmetric surface distance (ASSD) of 1.4mm0.2mm. Performance on the external dataset is comparable, with a DSC of 0.9580.006 and an ASSD of 1.70.2mm. Our method achieves performance comparable to or better than existing state-of-the-art methods for brain extraction, particularly in its highly consistent preservation of the brain's outer surface. The method is publicly available on GitHub.
Paper Structure (6 sections, 2 equations, 4 figures, 3 tables)

This paper contains 6 sections, 2 equations, 4 figures, 3 tables.

Table of Contents

  1. INTRODUCTION
  2. METHOD
  3. Data
  4. Model Architecture and Training
  5. RESULTS
  6. CONCLUSION

Figures (4)

  • Figure 1: The architecture of the proposed model. The numbers in the blocks signify the number of channels it outputs. Each arrow represents an action, with its color indicating the specific action type.
  • Figure 2: Automatic brain segmentation of an image with movement artefacts. Methods are (a) MONSTR, (b) ROBEX, (c) Synthstrip, (d) Synthstrip without CSF, (e) Proposed. Red arrows indicate undersegmentations. The images in the top row display the masks overlaid onto the image, while the bottom row shows the skull stripped brain.
  • Figure 3: Automatic brain segmentation of a standard T1-weighted MRI from a healthy subject. Methods are (a) MONSTR, (b) ROBEX, (c) Synthstrip, (d) Synthstrip without CSF, (e) Proposed. Red arrows indicate oversegmentationCSF. The images in the top row display images with masks overlaid, while the bottom row shows the skull stripped brain.
  • Figure 4: Automatic brain segmentation of a sample from the IXI dataset. Methods are (a) silver-standard ground truth, (b) MONSTR, (c) ROBEX, (d) Synthstrip, (e) Proposed.