NeuroShield: AI-Driven Brain Tumor Detection and Risk Assessment

This paper presents a deep learning-based system for automated brain tumor detection and risk assessment using a hybrid model that combines ResNet50 CNN and Swin Transformer V2-B. The system accurately identifies glioma, meningioma, and pituitary tumors from MRI scans, estimates tumor size (area, diameter, perimeter), categorizes tumors by size, and assigns risk levels. It also generates detailed clinical reports. Trained on public datasets, the model achieves high accuracy, sensitivity, and specificity, demonstrating its potential to enhance diagnostic precision, support clinical decision-making, and streamline radiological workflows.