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

Fibrinogen in early admission is an independent risk factor for early deep venous thrombosis in elderly patients with hip fracture.

Scientific reports·2026
Same author

Gut microbiome alterations are sex-dependently associated with brain abnormalities in a mouse model of Neurofibromatosis type I.

Molecular psychiatry·2026
Same author

Remote ischaemic conditioning in patients with supratentorial intracerebral haemorrhage (RICH-2): a multicentre, randomised, sham-controlled phase 3 trial in China.

EClinicalMedicine·2026
Same author

Paternal high molecular weight poly I:C administration alters the sperm small non-coding RNA profile and offspring brain development and behavioural phenotype.

Brain, behavior, and immunity·2026
Same author

Finite element analysis of laryngeal mask airway sealing under patient-specific anatomical constraints: an exploratory study.

Computer methods in biomechanics and biomedical engineering·2026
Same author

Inhibition of the NLRP3 Inflammasome With MCC950 Improves Gut Health in Huntington's Disease Mice.

Journal of neurochemistry·2026

Related Experiment Video

Updated: Jul 8, 2025

Individualized rTMS Treatment for Depression using an fMRI-Based Targeting Method
07:12

Individualized rTMS Treatment for Depression using an fMRI-Based Targeting Method

Published on: August 2, 2021

3.6K

Classification of MDD using a Transformer classifier with large-scale multisite resting-state fMRI data.

Peishan Dai1, Ying Zhou1, Yun Shi1

  • 1School of Computer Science and Engineering, Central South University, Changsha, China.

Human Brain Mapping
|December 13, 2023
PubMed
Summary
This summary is machine-generated.

This study introduces a Transformer-Encoder model using resting-state fMRI to identify major depressive disorder (MDD). The model achieved 68.61% accuracy in classifying MDD patients, highlighting frontal lobe abnormalities.

Keywords:
Transformerclassificationfunctional connectivitymajor depressive disorderrecurrence

More Related Videos

MRI-guided dmPFC-rTMS as a Treatment for Treatment-resistant Major Depressive Disorder
08:20

MRI-guided dmPFC-rTMS as a Treatment for Treatment-resistant Major Depressive Disorder

Published on: August 11, 2015

13.9K
Developing Neuroimaging Phenotypes of the Default Mode Network in PTSD: Integrating the Resting State, Working Memory, and Structural Connectivity
10:43

Developing Neuroimaging Phenotypes of the Default Mode Network in PTSD: Integrating the Resting State, Working Memory, and Structural Connectivity

Published on: July 1, 2014

15.1K

Related Experiment Videos

Last Updated: Jul 8, 2025

Individualized rTMS Treatment for Depression using an fMRI-Based Targeting Method
07:12

Individualized rTMS Treatment for Depression using an fMRI-Based Targeting Method

Published on: August 2, 2021

3.6K
MRI-guided dmPFC-rTMS as a Treatment for Treatment-resistant Major Depressive Disorder
08:20

MRI-guided dmPFC-rTMS as a Treatment for Treatment-resistant Major Depressive Disorder

Published on: August 11, 2015

13.9K
Developing Neuroimaging Phenotypes of the Default Mode Network in PTSD: Integrating the Resting State, Working Memory, and Structural Connectivity
10:43

Developing Neuroimaging Phenotypes of the Default Mode Network in PTSD: Integrating the Resting State, Working Memory, and Structural Connectivity

Published on: July 1, 2014

15.1K

Area of Science:

  • Neuroscience
  • Psychiatry
  • Machine Learning

Background:

  • Major depressive disorder (MDD) is a prevalent global psychiatric condition with a significant risk of recurrence.
  • Understanding the neural underpinnings of MDD, especially recurrent forms, is crucial for effective diagnosis and treatment.
  • Resting-state functional magnetic resonance imaging (rs-fMRI) offers a non-invasive method to investigate brain function in MDD patients.

Purpose of the Study:

  • To develop and validate a Transformer-Encoder model for classifying major depressive disorder (MDD) and healthy controls (HC) using rs-fMRI data.
  • To identify brain regions and functional connectivity patterns associated with MDD, particularly recurrent MDD.
  • To evaluate the model's performance on a large-scale, multisite dataset and assess its generalizability across different brain atlases.

Main Methods:

  • Utilized functional connectivity data extracted from large-scale, multisite rs-fMRI datasets.
  • Developed a simplified Transformer-Encoder model, omitting the decoder to reduce complexity and parameters for limited sample sizes.
  • Implemented an optional unsupervised pre-training module to optimize initial parameters and accelerate training.
  • Employed Grad-CAM for visualizing and identifying brain regions contributing to classification accuracy.

Main Results:

  • The Transformer-Encoder model achieved an average classification accuracy of 68.61% on a dataset of 1611 samples (MDD vs. HC).
  • For a specific recurrent MDD dataset, the model reached an average accuracy of 78.11%.
  • Abnormalities were identified in the frontal gyri and cerebral cortex of MDD patients; these regions showed higher contribution in the recurrent MDD group.

Conclusions:

  • The proposed Transformer-Encoder model effectively classifies MDD patients using rs-fMRI functional connectivity.
  • The model demonstrates potential for identifying recurrent MDD and pinpointing affected brain regions, specifically in the frontal cortex.
  • This approach offers an end-to-end classification solution adaptable to multi-site data and various brain atlases.