Enhancing brain tumor classification with a novel attention based explainable deep learning framework

doi:10.1016/j.bspc.2025.108636

Enhancing brain tumor classification with a novel attention based explainable deep learning framework

2025-09-12

0

OA

AI

Abu Bakar Siddique Mahi

M

Md Palash Uddin *

DOI：10.1016/j.bspc.2025.108636

Original

Save

View PDF

Abstract

En 中文

• SSPANet enhances spatial and contextual learning for MRI brain tumor diagnosis. • SSPANet with ResNet50 achieves 97% accuracy and 95% Kappa, outperforming baselines. • GradCAM++ with SSPANet provides clinically relevant visual explanations. • Theory and experiments confirm SSPANet’s effectiveness in medical imaging.

Keywords:

Brain tumor classification

Attention network

Strip-Style Pooling Attention

Deep learning

Explainability

Enhancing brain tumor classification with a novel attention based explainable deep learning framework

Abstract

AI Summary

Journal

Organization