Multiple Sclerosis Lesion Segmentation Using Statistical and Topological Atlases

This paper presents a new fully automatic method for segmentation of brain images that possess multiple sclerosis (MS) lesions. Multichannel magnetic resonance images are used to delineate multiple sclerosis lesions while segmenting the brain into its major structures. The method is an atlas based segmentation technique employing a topological atlas as well as a statistical atlas. An advantage of this approach is that all segmented structures are topologically constrained, thereby allowing subsequent processing with cortical unfolding or diffeomorphic shape analysis techniques. Validation on data from two studies demonstrates that the method has an accuracy comparable with other MS lesion segmentation methods, while simultaneously segmenting the whole brain.

Download Full-text

Review of Deep Learning Approaches for the Segmentation of Multiple Sclerosis Lesions on Brain MRI

Frontiers in Neuroinformatics ◽

10.3389/fninf.2020.610967 ◽

2020 ◽

Vol 14 ◽

Author(s):

Chenyi Zeng ◽

Lin Gu ◽

Zhenzhong Liu ◽

Shen Zhao

Keyword(s):

Multiple Sclerosis ◽

Deep Learning ◽

Brain Mri ◽

Multiple Sclerosis Lesion ◽

Dice Similarity Coefficient ◽

Learning Approaches ◽

Review Of The Literature ◽

Lesion Segmentation ◽

Segmentation Methods ◽

Future Direction

In recent years, there have been multiple works of literature reviewing methods for automatically segmenting multiple sclerosis (MS) lesions. However, there is no literature systematically and individually review deep learning-based MS lesion segmentation methods. Although the previous review also included methods based on deep learning, there are some methods based on deep learning that they did not review. In addition, their review of deep learning methods did not go deep into the specific categories of Convolutional Neural Network (CNN). They only reviewed these methods in a generalized form, such as supervision strategy, input data handling strategy, etc. This paper presents a systematic review of the literature in automated multiple sclerosis lesion segmentation based on deep learning. Algorithms based on deep learning reviewed are classified into two categories through their CNN style, and their strengths and weaknesses will also be given through our investigation and analysis. We give a quantitative comparison of the methods reviewed through two metrics: Dice Similarity Coefficient (DSC) and Positive Predictive Value (PPV). Finally, the future direction of the application of deep learning in MS lesion segmentation will be discussed.

Download Full-text

Comparative multiple sclerosis lesion segmentation in magnetic resonance images

2017 Electric Electronics, Computer Science, Biomedical Engineerings' Meeting (EBBT) ◽

10.1109/ebbt.2017.7956784 ◽

2017 ◽

Author(s):

Selin Isoglu ◽

Elif Isikci Koca ◽

Dilek Goksel Duru

Keyword(s):

Multiple Sclerosis ◽

Magnetic Resonance ◽

Magnetic Resonance Images ◽

Multiple Sclerosis Lesion ◽

Lesion Segmentation

Download Full-text

Convolutional Neural Networks Applied to Multiple Sclerosis Lesion Segmentation on 3D Brain Magnetic Resonance Images

9th International Conference on Pattern Recognition Systems (ICPRS 2018) ◽

10.1049/cp.2018.1278 ◽

2018 ◽

Author(s):

R. Naranjo ◽

G. Ulloa ◽

H. Allende-Cid ◽

H. Allende

Keyword(s):

Multiple Sclerosis ◽

Neural Networks ◽

Magnetic Resonance ◽

Convolutional Neural Networks ◽

Magnetic Resonance Images ◽

Multiple Sclerosis Lesion ◽

Lesion Segmentation

Download Full-text

A robust Expectation-Maximization algorithm for Multiple Sclerosis lesion segmentation

10.54294/1pbpd4 ◽

2008 ◽

Author(s):

Daniel Garcia-Lorenzo ◽

Sylvain Prima ◽

Sean P. Morrissey ◽

Christian Barillot

Keyword(s):

Multiple Sclerosis ◽

Expectation Maximization ◽

Expectation Maximization Algorithm ◽

User Interaction ◽

Intensity Inhomogeneity ◽

Multiple Sclerosis Lesion ◽

Lesion Segmentation ◽

Skull Stripping ◽

Fully Automatic

A fully automatic workflow for Multiple Sclerosis (MS) lesion segmentation is described. Fully automatic means that no user interaction is performed in any of the steps and that all parameters are fixed for all the images processed in beforehand. Our workflow is composed of three steps: an intensity inhomogeneity (IIH) correction, skull-stripping and MS lesions segmentation. A validation comparing our results with two experts is done on MS MRI datasets of 24 MS patients from two different sites.

Download Full-text