Brain Tumor Detection Using Transfer Learning-Based CNN Architectures in IoT Healthcare Industries

AI-driven deep learning has revolutionized brain tumor diagnosis in IoT healthcare by analyzing MRI data with high accuracy. This study classifies glioma, meningioma, and pituitary tumors using Transfer Learning CNN models-Inception-V3, EfficientNetB4, and VGG19. EfficientNetB4 achieved 98.31% accuracy, 98% F-score, 99.33% AUC, 98% recall, and 98% precision, proving effective for early tumor diagnosis. IoT devices continuously collect real-time data, enabling AI to provide immediate insights for personalized treatments. External validation showed VGG19's strong performance with 93.35% accuracy. These findings highlight AI's robustness in brain tumor classification and patient care within IoT healthcare systems.