XAI-BT-EdgeNet: explainable edge-aware deep learning with squeeze-and-excitation for brain tumor detection and
Deependra Rastogi1, Prashant Johri2, Massimo Donelli3,4
1School of Computer Science and Engineering, IILM University, Greater Noida, India.
Frontiers in Artificial Intelligence
|December 8, 2025
Summary
This study introduces XAI-BT-EdgeNet, an explainable AI framework for accurate brain tumor detection using MRI scans. The model achieves high accuracy, offering transparent and trustworthy AI-driven diagnostic support for clinicians.
Area of Science:
- Artificial Intelligence
- Medical Imaging
- Computer Science
Background:
- Manual analysis of MRI scans for brain tumor detection is time-consuming and prone to errors.
- Developing trustworthy AI for clinical decision support in neuro-oncology is crucial.
Purpose of the Study:
- To present XAI-BT-EdgeNet, an explainable, edge-aware deep learning framework for brain tumor detection using MRI.
- To enhance clinical trust in AI by providing interpretable justifications for diagnostic predictions.
Main Methods:
- A dual-branch CNN architecture fusing semantic (InceptionV3) and edge features.
- Integration of Squeeze-and-Excitation (SE) modules for adaptive feature recalibration.
- Incorporation of four Explainable AI (XAI) techniques (LIME, Grad-CAM, Grad-CAM++, Vanilla Saliency) for transparency.
Main Results:
- Achieved high classification accuracies: 99.58% (training), 99.71% (validation), and 100.00% (testing).
- Demonstrated minimal loss values, indicating model robustness and precision.
- Generated interpretable visual explanations for AI-driven brain tumor detection.
Conclusions:
- XAI-BT-EdgeNet offers a high-performing and interpretable solution for brain tumor classification from MRI scans.
- The framework integrates advanced deep learning with XAI to improve clinical decision-making.
- This approach bridges the gap between AI capabilities and clinical trust in neuro-oncology.
