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 26, 2026

Manual Segmentation of the Human Choroid Plexus Using Brain MRI
04:25

Manual Segmentation of the Human Choroid Plexus Using Brain MRI

Published on: December 15, 2023

Segmentation-Free Preoperative 3D MRI Classification of Low-Grade Versus High-Grade Glioma Using Task-Oriented Neural

Christos Ch Andrianos1, Spiros A Kostopoulos1, Ioannis K Kalatzis1

  • 1Medical Image and Signal Processing Laboratory, Department of Biomedical Engineering, University of West Attica, 12241 Athens, Greece.

Journal of Imaging
|June 25, 2026
PubMed
Summary

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

Implementation of Instrumental Analytical Methods, Image Analysis and Chemometrics for the Comparative Evaluation of Citrus Fruit Peels.

Foods (Basel, Switzerland)·2025
Same author

Development of a Fully Optimized Convolutional Neural Network for Astrocytoma Classification in MRI Using Explainable Artificial Intelligence.

Journal of imaging·2025
Same author

Prediction of remaining surgery duration based on machine learning methods and laparoscopic annotation data.

Biomedizinische Technik. Biomedical engineering·2025
Same author

Over 30 Years of DiI Use for Human Neuroanatomical Tract Tracing: A Scoping Review.

Biomolecules·2024
Same author

Single-cell and spatial transcriptomics analysis of non-small cell lung cancer.

Nature communications·2024
Same author

Acquisition of epithelial plasticity in human chronic liver disease.

Nature·2024

This study introduces a novel segmentation-independent framework for classifying brain gliomas using a specialized Convolutional Neural Network (CNN). The approach accurately distinguishes low-grade from high-grade gliomas (LGG/HGG) on MRI scans, improving diagnostic capabilities.

Area of Science:

  • Neuroimaging
  • Artificial Intelligence in Medicine
  • Oncology

Background:

  • Gliomas are the most common primary brain tumors, necessitating accurate Magnetic Resonance Imaging (MRI) for effective patient management.
  • Current computer-aided diagnosis methods often depend on complex tumor segmentation, posing a challenge for clinical application.
  • Developing segmentation-independent diagnostic tools can streamline the analysis of brain tumors.

Purpose of the Study:

  • To propose and evaluate a novel segmentation-independent framework for classifying volumetric low-grade versus high-grade gliomas (LGG/HGG).
  • To leverage task-oriented Neural Architecture Search (NAS) for designing an optimized Convolutional Neural Network (CNN) for glioma classification.
  • To assess the model's performance on multi-center datasets and compare it against existing benchmarks.
Keywords:
3D convolutional neural networkexplainable artificial intelligencegliomaneural architecture searchsegmentation-free classification

Related Experiment Videos

Last Updated: Jun 26, 2026

Manual Segmentation of the Human Choroid Plexus Using Brain MRI
04:25

Manual Segmentation of the Human Choroid Plexus Using Brain MRI

Published on: December 15, 2023

Main Methods:

  • A segmentation-independent framework utilizing a 3D U-Net-based backbone optimized via Neural Architecture Search (NAS) with Tree-structured Parzen Estimator (TPE) and Hyperband pruning.
  • Incorporation of residual connections and Squeeze-and-Excitation (SE) attention mechanisms to enhance feature representation and model stability.
  • Validation using internal 5-fold cross-validation on a multi-center dataset (1194 patients, T1-CE and FLAIR MRI sequences) and external testing on the REMBRANDT cohort.

Main Results:

  • The proposed model achieved 88.25% internal accuracy and 75.51% external accuracy, with macro-F1 scores of 87.37% (internal) and 73.77% (external).
  • The model outperformed benchmark 3D Convolutional Neural Networks (CNNs) in glioma classification.
  • Explainable Artificial Intelligence (XAI) analysis confirmed robust tumor localization without segmentation supervision, and additional experiments showed strong performance in IDH mutation prediction (89.51% accuracy) and multi-grade classification (78.74% accuracy).

Conclusions:

  • The developed segmentation-independent framework offers a promising approach for accurate glioma classification using MRI.
  • The task-oriented NAS efficiently designed a robust CNN architecture, demonstrating superior performance compared to traditional methods.
  • The model's ability to generalize and predict other glioma characteristics highlights its potential clinical utility in neuro-oncology.