Related Experiment Video
Updated: Aug 3, 2025
06:48
Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
Published on: January 7, 2019
9.0K
Weighted Average Ensemble Deep Learning Model for Stratification of Brain Tumor in MRI Images
Vatsala Anand1, Sheifali Gupta1, Deepali Gupta1
1Chitkara University Institute of Engineering and Technology, Chitkara University, Rajpura 140401, Punjab, India.
Diagnostics (Basel, Switzerland)
|April 13, 2023
Summary
This study introduces a deep learning ensemble model for brain tumor classification using MRI images. The model enhances diagnostic accuracy, aiding radiologists in early detection and improving patient outcomes.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Oncology
Background:
- Early brain tumor diagnosis is crucial for effective treatment and improved patient outcomes.
- Non-invasive methods like Magnetic Resonance Imaging (MRI) are vital for tumor detection.
- Deep learning offers rapid analysis of medical images, potentially accelerating diagnosis.
Purpose of the Study:
- To propose a weighted average ensemble deep learning model for accurate brain tumor classification.
- To enhance diagnostic capabilities by combining multiple deep learning models.
- To reduce diagnostic time and improve the reliability of brain tumor detection from MRI scans.
Main Methods:
- A weighted average ensemble deep learning model was developed.
- Feature spaces were extracted from VGG19, CNN without augmentation, and CNN with augmentation models.
- Grid search was employed to determine optimal weights for ensembling.
- The Cancer Genome Atlas (TCGA) dataset, comprising 3929 lower-grade glioma MRI images, was utilized.
Main Results:
- The proposed ensemble model demonstrated superior performance compared to individual models.
- The model achieved higher accuracy, precision, and F1-score in brain tumor classification.
- Ensembling effectively reduced overfitting by integrating diverse model strengths.
Conclusions:
- The weighted average ensemble deep learning model shows significant potential for accurate brain tumor classification.
- This model can serve as a valuable second opinion tool for radiologists interpreting brain MRI scans.
- The findings suggest a promising advancement in AI-assisted oncological diagnostics.
