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 Experiment Video

Updated: Jun 29, 2026

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

MTA-Swin: A Multi-Token Attention Swin Transformer for Brain Tumor Classification with Leakage-Free MRI Benchmarking.

Dong Lu1, Yu Zhang1, Divya Chaudhary2

  • 1College of Engineering, Northeastern University, 401 Terry Ave N, Seattle, WA, 98109, United States.

Journal of Medical Systems
|June 21, 2026
PubMed
Summary
This summary is machine-generated.

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

Corrigendum to "Heavy ion beam irradiation-induced mutational profiles in Saccharomyces cerevisiae and their dependencies on dose and intracellular state inform an enhanced microbial breeding strategy" [J. Biotechnol. 410 (2026) 236-257].

Journal of biotechnology·2026
Same author

Dynamic changes in In vitro rumen fermentation, nutrient degradation, and microbial communities of fermentation inoculant-treated licorice stem and leaf silage under different dry matter contents.

PloS one·2026
Same author

Comparison of Child-Pugh and MELD scores in predicting survival of hepatocellular carcinoma patients with splenomegaly undergoing TACE: a real-world study.

Frontiers in medicine·2026
Same author

Effects of Immunonutrition on Intestinal Microecology in Patients With Gastrointestinal Cancers: Mechanisms and Clinical Applications.

Nutrition reviews·2026
Same author

Synthesis of 3-Difluoromethyl-1,3,4-oxadiazole-thiones from Aroylhydrazides, TMSCF<sub>3</sub>, and Elemental Sulfur.

The Journal of organic chemistry·2026
Same author

From technology to interaction: How ritual interaction drives participation intention in immersive digital environments.

iScience·2026

This study introduces a method to remove duplicate data in brain tumor MRI datasets, creating a reliable benchmark. A new model, MTA-Swin, achieves 98.57% accuracy in brain tumor classification, offering a practical diagnostic tool.

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Computer Vision

Background:

  • Accurate brain tumor classification is crucial for patient diagnosis and treatment.
  • Magnetic resonance imaging (MRI) is vital for early detection.
  • Existing MRI datasets often contain duplicates, leading to unreliable deep learning model evaluations.

Purpose of the Study:

  • To address data leakage in brain tumor MRI datasets caused by duplicates.
  • To develop an automated pipeline for identifying and removing duplicate scans.
  • To propose and evaluate an enhanced deep learning model for improved brain tumor classification.

Main Methods:

  • Systematic analysis of duplicate-induced data leakage in MRI datasets.
  • Development of an automated data cleaning pipeline.
Keywords:
Brain tumor classificationComputer visionData leakageDeep learningMedical imagingMulti-Token attentionSwin transformer

More Related Videos

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

Related Experiment Videos

Last Updated: Jun 29, 2026

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

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

  • Creation of a leakage-free benchmark dataset (3,522 unique scans).
  • Proposal of MTA-Swin, a Swin Transformer with Multi-Token Attention.
  • Pre-training MTA-Swin on ImageNet-1K and fine-tuning on the cleaned dataset.

    • Main Results:

    • A leakage-free benchmark dataset was established.
    • MTA-Swin achieved 98.57% overall accuracy across three random seeds.
    • The model outperformed thirteen baseline methods.
    • Stratified cross-validation and Grad-CAM analyses confirmed robustness and interpretability.

    Conclusions:

    • Automated data cleaning is essential for reliable brain tumor MRI dataset evaluation.
    • MTA-Swin demonstrates superior performance in brain tumor classification.
    • The proposed model shows potential as a practical computer-aided diagnostic support system.