Related Experiment Video
Updated: Aug 10, 2025
06:44
Modeling Brain Metastases Through Intracranial Injection and Magnetic Resonance Imaging
Published on: June 7, 2020
7.4K
Role of Ensemble Deep Learning for Brain Tumor Classification in Multiple Magnetic Resonance Imaging Sequence Data
Gopal S Tandel1, Ashish Tiwari2, Omprakash G Kakde3
1School of Computer Science and Engineering, VIT Bhopal University, Sehore 466114, India.
Diagnostics (Basel, Switzerland)
|February 11, 2023
Summary
This study introduces a non-invasive computer-aided diagnosis tool for brain tumor grading, utilizing magnetic resonance imaging (MRI) sequences. The ensemble deep learning model achieved high accuracy, identifying FLAIR-MRI as the most significant sequence for glioma classification.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Oncology
Background:
- Biopsy, the gold standard for tumor grading, is invasive and can be fatal for brain tumor patients.
- A non-invasive computer-aided diagnosis (CAD) tool is crucial for accurate and safe brain tumor grading.
- Magnetic resonance imaging (MRI) offers various sequences for visualizing tumor structures, but the optimal sequence for classification remains unclear.
Purpose of the Study:
- To develop and evaluate a non-invasive CAD tool for differentiating low-grade versus high-grade gliomas.
- To determine the most effective MRI sequence for brain tumor classification using deep learning models.
- To improve the accuracy and consistency of brain tumor grading through an ensemble deep learning approach.
Main Methods:
- Three MRI datasets were created using T1-Weighted (T1W), T2-weighted (T2W), and fluid-attenuated inversion recovery (FLAIR) sequences.
- Five established convolutional neural networks (AlexNet, VGG16, ResNet18, GoogleNet, ResNet50) were employed for tumor classification.
- An ensemble algorithm (MajVot) using majority voting was proposed, combined with a five-fold cross-validation (K5-CV) protocol.
Main Results:
- The ensembled classifier achieved high test accuracies: 98.88% (FLAIR), 97.98% (T2W), and 94.75% (T1W).
- FLAIR-MRI data demonstrated superior performance, improving accuracy by 4.17% over T1W-MRI and 0.91% over T2W-MRI.
- The proposed MajVot ensemble algorithm significantly improved average accuracy across all datasets compared to individual deep learning models.
Conclusions:
- FLAIR-MRI sequences are highly significant for accurate brain tumor classification, outperforming T1W and T2W sequences.
- The proposed ensemble deep learning model (MajVot) offers a robust and accurate non-invasive method for glioma grading.
- This CAD tool has the potential to enhance patient outcomes by providing a safer alternative to invasive biopsy procedures.
Related Concept Videos
01:24Magnetic Resonance Imaging
5.4K
Magnetic resonance imaging (MRI) is a noninvasive medical imaging technique based on a phenomenon of nuclear physics discovered in the 1930s, in which matter exposed to magnetic fields and radio waves was found to emit radio signals. In 1970, a physician and researcher named Raymond Damadian noticed that malignant (cancerous) tissue gave off different signals than normal body tissue. He applied for a patent for the first MRI scanning device in clinical use by the early 1980s. The early MRI...
5.4K
01:27Imaging Studies IV: Magnetic Resonance Imaging
42
Introduction:Magnetic Resonance Imaging, or MRI, can include a specialized imaging technique of the urinary system known as Magnetic Resonance Urography (MRU). This radiation-free technique uses strong magnetic fields and radio waves to produce detailed images with the help of a computer. MRU is particularly effective for visualizing fluid-filled structures like the kidneys, ureters, and bladder.Applications of MRI in the Genitourinary SystemKidneys and Ureters: MRI detects tumors, cysts,...
42