Automated Brain Tumor Detection in MRI Scans through Deep Learning Techniques
Main Article Content
Abstract
Brain tumors represent a significant challenge in modern medicine, making accurate detection crucial for effective treatment. Manual interpretation of Magnetic Resonance Imaging (MRI) scans can be time-consuming and prone to human error. This study utilizes advancements in machine learning, particularly Convolutional Neural Networks (CNNs), to automate the detection and classification of brain tumors in MRI scans. We developed a comprehensive dataset comprising images from four categories: glioma, meningioma, pituitary tumors, and non-tumor cases. Rigorous preprocessing techniques, including data augmentation and normalization, ensured the dataset's quality and balance. Our CNN model achieved an impressive accuracy rate of 96%, utilizing hypercolumn extraction and multi-scale convolutional layers. The architecture includes convolutional layers for feature extraction, pooling layers for dimensionality reduction, and fully connected layers for classification. Advanced techniques like dropout were implemented to mitigate overfitting. This research emphasizes the role of automated diagnostic systems in improving tumor detection, especially in areas with limited access to skilled radiologists. Ultimately, our findings contribute to the integration of AI-enabled tools in clinical workflows, facilitating effective diagnostics and personalized patient care