Early Brain Tumor Detection Using MRI and Deep Learning: A CNN-Based Classification Approach

Abstract

Early and accurate detection of brain tumors is crucial for improving patient outcomes and enabling effective treatment. Magnetic Resonance Imaging (MRI) is widely used for brain imaging due to its ability to capture detailed anatomical structures. However, manual interpretation of MRI scans is labor-intensive and prone to human error, especially during the early stages of tumor development. This study presents a deep learning-based approach for automating the classification of brain tumors using a custom-designed Convolutional Neural Network (CNN). The proposed CNN architecture was trained and tested on a dataset of 3,000 MRI images comprising four classes: glioma, meningioma, pituitary tumor, and no tumor. To enhance model performance, preprocessing techniques such as image normalization and denoising were applied, followed by data augmentation strategies to address dataset imbalance and improve generalization. The CNN model was implemented in a cloud-based environment (Google Colab Pro+) and evaluated using key performance metrics including accuracy, recall, AUC, and loss. Compared to other deep learning models like ResNet-50, VGG16, and Inception V3, the proposed CNN achieved superior results with 92.29% accuracy, 90.12% recall, 97.42% AUC, and a minimal loss of 0.24. This research highlights the effectiveness of CNNs for early brain tumor detection and demonstrates the potential of deep learning in clinical diagnostics. Future enhancements could focus on explainable AI tools and transfer learning models for improved interpretability and robustness in medical image classification.