A Hybrid Approach for Accurate Brain Tumor Detection Using Deep Learning Techniques

Brain tumors are a major global health concern, contributing to many deaths each year. Early and accurate diagnosis is essential for improving patient outcomes. In this paper, we introduce a new hybrid approach for detecting brain tumors. We started by preprocessing MRI images—gray-scaling, enhancing contrast, and applying masking—to highlight important features. We then classified the images using five deep-learning techniques, and the custom VGG16 and VGG19 models performed better than the other models. Our study utilized a dataset of 7,023 images across four categories: Glioma, Meningioma, No-Tumor, and Pituitary tumors. The results from using VGG16 and VGG19 were impressive, achieving 97% accuracy, 97% precision, and a 97% F1 score. These findings underscore the potential of our approach to enhance diagnostic accuracy and ultimately improve patient care.