Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Safety of migraine prophylaxis with calcitonin gene-related peptide monoclonal antibodies in comorbid chronic inflammatory central nervous system disease: A case series.

Headache·2026
Same author

Higher disease reactivation risk in women after fingolimod withdrawal.

Acta neuropathologica communications·2026
Same author

Listening to MS: AI-assisted speech analysis for diagnosis and fatigue prediction (COMMITMENT).

Frontiers in digital health·2026
Same author

Monitoring physical activity by fitness tracking devices : Usage, acceptance and willingness to share data in patients with multiple sclerosis.

Wiener klinische Wochenschrift·2026
Same author

Mitigating the Impact of MR Sequence Parameters: Increasing the Robustness of DL-Based Cortical Thickness Estimates.

Human brain mapping·2026
Same author

Virtual reality-based oculography detects internuclear ophthalmoplegia in multiple sclerosis and other neurological disorders.

Journal of neurology·2026

Related Experiment Video

Updated: Nov 21, 2025

Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
10:25

Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping

Published on: September 25, 2019

48.9K

Simultaneous lesion and brain segmentation in multiple sclerosis using deep neural networks.

Richard McKinley1, Rik Wepfer2, Fabian Aschwanden2

  • 1Support Center for Advanced Neuroimaging, University Institute for Diagnostic and Interventional Neuroradiology, Inselspital, Bern University Hospital, Bern, Switzerland. richard.mckinley@insel.ch.

Scientific Reports
|January 14, 2021
PubMed
Summary
This summary is machine-generated.

Deep learning models, 3D Unet and DeepSCAN, offer fast and reliable segmentation of white matter lesions and grey matter structures in multiple sclerosis MR imaging. DeepSCAN, with anatomical labels, demonstrated superior performance, aligning with human rater variability.

More Related Videos

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
04:25

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

3.3K
Lesion Explorer: A Video-guided, Standardized Protocol for Accurate and Reliable MRI-derived Volumetrics in Alzheimer's Disease and Normal Elderly
12:50

Lesion Explorer: A Video-guided, Standardized Protocol for Accurate and Reliable MRI-derived Volumetrics in Alzheimer's Disease and Normal Elderly

Published on: April 14, 2014

40.6K

Related Experiment Videos

Last Updated: Nov 21, 2025

Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
10:25

Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping

Published on: September 25, 2019

48.9K
Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
04:25

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

3.3K
Lesion Explorer: A Video-guided, Standardized Protocol for Accurate and Reliable MRI-derived Volumetrics in Alzheimer's Disease and Normal Elderly
12:50

Lesion Explorer: A Video-guided, Standardized Protocol for Accurate and Reliable MRI-derived Volumetrics in Alzheimer's Disease and Normal Elderly

Published on: April 14, 2014

40.6K

Area of Science:

  • Neuroimaging
  • Artificial Intelligence
  • Medical Image Analysis

Background:

  • Accurate segmentation of white matter lesions and deep grey matter structures is crucial for multiple sclerosis (MS) quantification using magnetic resonance imaging (MRI).
  • Convolutional neural networks (CNNs) offer potential for automated, efficient, and reliable segmentation in multi-modal MR imaging.

Purpose of the Study:

  • To evaluate the performance of state-of-the-art fully convolutional CNN architectures (3D Unet and DeepSCAN) for segmenting MS lesions and grey matter structures.
  • To assess the impact of dataset shift and the utility of additional anatomical labels on segmentation performance.
  • To compare CNN-based methods against existing reference techniques.

Main Methods:

  • Two CNN architectures, 3D Unet and DeepSCAN, were trained on the 2016 MSSEG dataset, annotated by multiple human raters.
  • Methods were retrained on a larger single-center dataset, with and without additional brain structure labels derived from Freesurfer.
  • Performance was quantified by evaluating dataset shift effects and the impact of anatomical labels, with comparisons to reference methods.

Main Results:

  • Both 3D Unet and DeepSCAN CNNs significantly outperformed other literature methods on the MSSEG dataset, achieving agreement within human inter-rater variability.
  • A performance drop was observed for both architectures when trained on single-center data and tested on the MSSEG dataset.
  • DeepSCAN performance improved with additional anatomical labels, while 3D Unet's performance degraded. DeepSCAN with lesion and anatomical labels yielded the best results.

Conclusions:

  • Fully convolutional CNNs provide robust and accurate segmentation for MS lesion and grey matter quantification.
  • DeepSCAN demonstrates superior adaptability and performance, particularly when augmented with anatomical labels, making it a promising tool for MS neuroimaging analysis.
  • The findings highlight the potential of deep learning for improving the efficiency and reliability of MS quantification in clinical practice.