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

From Anatomy to Outcome: Linking Glioma Location Patterns to Survival Using Non-Negative Matrix Factorization.

Clinical neuroradiology·2026
Same author

Performance of a GPU- and time-efficient pseudo-3D network for magnetic resonance image super-resolution and motion artifact reduction.

Scientific reports·2026
Same author

Segmentation and structural connectivity of the putamen for targeted convection enhanced drug delivery in Parkinson's disease: a tractography-based approach.

Frontiers in neuroscience·2026
Same author

Repeat resection for recurrent glioblastoma - does timing matter?

Journal of neuro-oncology·2026
Same author

Multimodality mapping of immunotherapy distribution as a predictive marker in glioma.

Neuro-oncology·2025
Same author

Predicting molecular types of adult-type diffuse gliomas based on MRI reports with large language models.

European radiology·2025

Related Experiment Video

Updated: Jul 22, 2025

High-resolution In Vivo Manual Segmentation Protocol for Human Hippocampal Subfields Using 3T Magnetic Resonance Imaging
11:03

High-resolution In Vivo Manual Segmentation Protocol for Human Hippocampal Subfields Using 3T Magnetic Resonance Imaging

Published on: November 10, 2015

9.4K

Automated hippocampal segmentation algorithms evaluated in stroke patients.

Marianne Schell1, Martha Foltyn-Dumitru1, Martin Bendszus1

  • 1Department of Neuroradiology, Heidelberg University Hospital, Im Neuenheimer Feld 400, 69120, Heidelberg, Germany.

Scientific Reports
|July 20, 2023
PubMed
Summary

Six deep learning algorithms for hippocampal segmentation were evaluated on stroke patients. Performance varied significantly, especially with lesions, highlighting the need for careful algorithm selection based on research needs.

More Related Videos

A Comprehensive Protocol for Manual Segmentation of the Medial Temporal Lobe Structures
12:30

A Comprehensive Protocol for Manual Segmentation of the Medial Temporal Lobe Structures

Published on: July 2, 2014

20.4K
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

40.3K

Related Experiment Videos

Last Updated: Jul 22, 2025

High-resolution In Vivo Manual Segmentation Protocol for Human Hippocampal Subfields Using 3T Magnetic Resonance Imaging
11:03

High-resolution In Vivo Manual Segmentation Protocol for Human Hippocampal Subfields Using 3T Magnetic Resonance Imaging

Published on: November 10, 2015

9.4K
A Comprehensive Protocol for Manual Segmentation of the Medial Temporal Lobe Structures
12:30

A Comprehensive Protocol for Manual Segmentation of the Medial Temporal Lobe Structures

Published on: July 2, 2014

20.4K
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

40.3K

Area of Science:

  • Neuroimaging
  • Artificial Intelligence in Medicine
  • Medical Image Analysis

Background:

  • Deep learning (DL) algorithms offer rapid, reproducible segmentation of brain structures like the hippocampus.
  • However, their reliability on complex datasets with structural abnormalities, such as in stroke patients, remains uncertain.
  • Pathology-based domain shifts, particularly lesion presence, can significantly impact segmentation accuracy.

Purpose of the Study:

  • To evaluate and compare the performance of six recent deep learning-based hippocampal segmentation algorithms.
  • To assess algorithm robustness in the presence of stroke-related brain abnormalities and lesions.
  • To determine if any single algorithm consistently outperforms others across various metrics.

Main Methods:

  • Six state-of-the-art DL hippocampal segmentation algorithms were tested.
  • The study utilized the multicentric, open-source ATLAS 2.0 dataset comprising 641 stroke patients.
  • Performance was assessed using volumetric similarity (VS), DICE score, and Hausdorff distance (HD), with concordance correlation coefficients (CCC) used for inter-method comparisons.

Main Results:

  • Algorithms showed variable performance, with CCC ranging from 0.266 to 0.816, indicating they are not interchangeable.
  • Overall good performance was observed (VS: 0.816–0.972, DICE: 0.786–0.921, HD: 2.69–6.34).
  • No single algorithm excelled in all metrics; FastSurfer led in VS, QuickNat in DICE and average HD, and Hippodeep in HD. Segmentation accuracy decreased significantly for ipsilesional (lesion side) segmentation, correlating with lesion size.

Conclusions:

  • Deep learning-based hippocampal segmentation performance is significantly impacted by stroke-related brain abnormalities and lesion presence.
  • Algorithm selection should be tailored to the specific research question and desired evaluation metric, as no single method is universally superior.
  • Further research is crucial to enhance the robustness and accuracy of current hippocampal segmentation methods, particularly for pathological brain conditions.