Brain Tumor Detection Using Modified VGG-19 Model

Brain tumors are life-threatening diseases, and their early detection is crucial but difficult through manual analysis. They can be identified by magnetic resonance imaging (MRI). This study presents a deep learning approach using 3D convolutional neural networks (CNNs) for brain tumor detection. The CNN architecture was based on a modified VGG19 model, which was trained on a large MRI dataset to learn about different types of tumors and identify difficult tumor patterns. Compared to existing models like VGG16 and ResNet50, which attained accuracies of 77% and 94% respectively, our proposed model achieved an accuracy of 95%. This deep learning methodology demonstrates the potential to enhance brain tumor detection, optimize diagnostic procedures, reduce human effort, and improve patient outcomes.