Related Experiment Video
Updated: Jul 25, 2026
Comprehensive Autopsy Program for Individuals with Multiple Sclerosis
Published on: July 19, 2019
White matter hyperintensity segmentation of multiple sclerosis and neuromyelitis optical spectrum disorders using
Li Zhang1, Kai Niu2, Yinglu Sun3
1Department of Radiology, the First Hospital of Jilin University, Changchun, China.
Background:
Assessing white matter hyperintensity (WMH) is essential for the diagnosis, treatment, and prognosis of multiple sclerosis (MS) and neuromyelitis optical spectrum disorder (NMOSD). MS and NMOSD present dispersed small lesions alongside larger aggregated lesions that are irregularly shaped, posing challenges for the automatic segmentation of WMH on magnetic resonance images. Furthermore, research on NMOSD brain WMH segmentation is limited due to the rare nature of the disease. This study aims to propose a deep learning method for MS and NMOSD brain WMH segmentation.
Methods:
In this study, we propose a 2.5D Fourier Convolutional ResUnet (FrC-ResUnet). It utilizes a spectral encoder to extract global information, enabling accurate segmentation of scattered lesions. Additionally, the model incorporates the selective features module (SFM) and the convolutional block attention module (CBAM) to enhance lesion-background differentiation and outline the lesions distinctly. We evaluated our approach on the MS public and local datasets of MS and NMOSD.
Results:
Compared to U-Net, ResUNet, FC-DenseNet, AttentionUNet, lesion prediction algorithm (LPA) and Sequence Adaptive Multimodal SEGmentation (SAMSEG), the 2.5D FrC-ResUnet achieved the highest Dice similarity coefficient (DSC) on three different datasets, with values of 0.710, 0.667, and 0.822, respectively.
Conclusions:
The 2.5D FrC-ResUnet demonstrates accurate and robust segmentation of NMOSD brain WMH. Meanwhile, the model excels in segmenting MS brain WMH, particularly when confronted with irregularly shaped and dispersed lesions.
Insights
This study introduces a deep learning model, 2.5D FrC-ResUnet, for segmenting white matter hyperintensities (WMH) in multiple sclerosis (MS) and neuromyelitis optical spectrum disorder (NMOSD) brain scans, achieving superior accuracy.
Area of Science:
- Neuroimaging
- Medical Image Analysis
- Artificial Intelligence
Background:
- Accurate segmentation of white matter hyperintensities (WMH) is crucial for diagnosing and managing multiple sclerosis (MS) and neuromyelitis optical spectrum disorder (NMOSD).
- Irregularly shaped and dispersed WMH in MS and NMOSD present significant challenges for automated segmentation on MRI.
- Limited research exists for NMOSD brain WMH segmentation due to disease rarity.
Purpose of the Study:
- To develop an advanced deep learning method for precise brain WMH segmentation in both MS and NMOSD.
- To address the challenges posed by complex lesion morphologies and improve segmentation accuracy.
Main Methods:
- Proposed a novel 2.5D Fourier Convolutional ResUnet (FrC-ResUnet) model incorporating a spectral encoder for global information extraction.
- Integrated selective features module (SFM) and convolutional block attention module (CBAM) to enhance lesion differentiation and boundary definition.
- Evaluated the model on public and local datasets encompassing MS and NMOSD brain MRI scans.
Main Results:
- The 2.5D FrC-ResUnet achieved superior performance compared to U-Net, ResUNet, FC-DenseNet, AttentionUNet, LPA, and SAMSEG.
- Attained the highest Dice Similarity Coefficients (DSC) of 0.710, 0.667, and 0.822 across three distinct datasets.
- Demonstrated robust segmentation capabilities for both MS and NMOSD brain WMH.
Conclusions:
- The 2.5D FrC-ResUnet offers accurate and robust segmentation of brain WMH in NMOSD.
- The model effectively segments MS brain WMH, especially irregularly shaped and dispersed lesions.
- This deep learning approach advances automated WMH segmentation for neurological disorders.
