This research proposes a deep learning-based fusion model integrating MobileNetV2 and EfficientNetB0 for multi-class brain tumor diagnosis using MRI images. The model leverages MobileNetV2's lightweight efficiency and EfficientNetB0's representational strength to achieve robust feature extraction while maintaining computational feasibility. The fusion approach achieved 95.33% validation accuracy with precision, recall, and F1-score of 96%, demonstrating strong performance for real-world clinical applications.

• Deep Learning
• Transfer Learning
• MobileNetV2
• EfficientNetB0
• Medical Imaging
• Brain Tumor Classification