Hybrid CNN-ViT model achieves 97.6% accuracy in brain tumor MRI classification

Researchers have developed a novel hybrid deep learning model that merges Convolutional Neural Networks (CNNs) with Vision Transformers (ViTs) for improved brain tumor classification from MRI scans. This new architecture utilizes an Adaptive Attention Gate to dynamically weigh the contributions of local features captured by CNNs and global dependencies identified by ViTs. The model achieved a high test accuracy of 97.60% on the Brain Tumor MRI Dataset, outperforming existing single-branch models and fusion methods. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces a hybrid deep learning architecture that could improve diagnostic accuracy in medical imaging.

RANK_REASON This is a research paper detailing a novel hybrid deep learning model for medical image classification.

Read on arXiv cs.CV →

• CNN
• ViT
• MRI
• Brain Tumor MRI Dataset
• Kaggle
• Syed Ibad Hasnain

• paper
• other