Advanced CNN Architecture for Multiclass Brain Tumor Classification

Brain tumors are a serious health threat, and correct diagnosis is critical to the success of medical intervention. This research suggests a Convolutional Neural Network (CNN)-based approach to automatically classify MRI brain tumor images into four classes: Glioma, Meningioma, No Tumor, and Pituitary. The goal is to maximize diagnostic accuracy, allowing for early detection and better patient outcomes. A dataset of 7,023 MRI images was used, and augmentation techniques were performed to enhance the model generalization. The CNN architecture was fine-tuned with hyperparameter tuning to get consistent classification performance. The end model achieved an average test accuracy of 99.4% and class-wise precision of 99.6% for Glioma, 99.5% for Meningioma, and 99.5% for Pituitary tumors. These findings show the promise of deep learning in medical diagnosis, providing a trustworthy and efficient automatic brain tumor detection and classification tool. This study supports the development of AI-based healthcare solutions, helping radiologists diagnose more quickly and accurately.