Related Experiment Video
Updated: Jan 17, 2026
09:53
Quantifying the Brain Metastatic Tumor Micro-Environment using an Organ-On-A Chip 3D Model, Machine Learning, and Confocal Tomography
Published on: August 16, 2020
7.7K
Dual-feature cross-fusion network for precise brain tumor classification: a neurocomputational approach
Muthalakshmi M1, Surya G2, Mininath Bendre3
1Department of Biomedical Engineering, Vel Tech Rangarajan Dr.Sagunthala R&D Institute of Science and Technology, Chennai, Tamil Nadu, India.
The International Journal of Neuroscience
|September 23, 2025
Summary
A new deep learning model, the Dual-Feature Cross-Fusion Network (DF-CFN), accurately classifies brain tumors from MRI scans. This automated approach enhances diagnostic efficiency and reliability for various tumor types.
Area of Science:
- Neurology
- Medical Imaging
- Artificial Intelligence
Background:
- Brain tumors pose a significant diagnostic challenge, necessitating accurate classification for effective treatment.
- Magnetic Resonance Imaging (MRI) is crucial for non-invasive brain tumor diagnosis, but manual interpretation is subjective and time-consuming.
Purpose of the Study:
- To develop and validate a novel deep learning architecture, the Dual-Feature Cross-Fusion Network (DF-CFN), for automated brain tumor classification using MRI data.
- To improve the accuracy and efficiency of brain tumor diagnosis compared to traditional methods.
Main Methods:
- The DF-CFN model integrates ConvNeXt for global features and a shallow CNN with FcaNet for local features, utilizing a cross-feature fusion mechanism.
- The model was trained and validated on Kaggle and FigShare datasets, including glioma, meningioma, pituitary, and non-tumor classes.
Main Results:
- The DF-CFN achieved high accuracy, reaching 99.33% on the Kaggle dataset and 99.22% on the FigShare dataset.
- Comparative analysis demonstrated the superiority of DF-CFN over baseline and recent models in precision and robustness for brain tumor classification.
Conclusions:
- The proposed DF-CFN model shows significant potential for assisting clinicians in reliable and efficient brain tumor classification from MRI.
- Automated classification using DF-CFN can reduce diagnostic workload and improve patient management strategies.