Abnormal Brain Tumors Classification Using ResNet50 and Its Comprehensive Evaluation

Abstract

Brain tumors present significant health risks due to abnormal cell growth, potentially leading to organ dysfunction and mortality in adults. Magnetic resonance imaging (MRI) is crucial for tumor classification, but limited expertise in this area necessitates advanced methods for accurate diagnosis. Deep Learning has emerged as a pivotal tool, yet gaps remain in achieving optimal accuracy. This study addresses these gaps by proposing an enhanced model for classifying meningioma, glioma, and pituitary gland tumors, thereby improving precision in brain tumor detection. Trained on a dataset of 5712 images, the model achieves exceptional accuracy (99%) in both training and validation datasets, with a focus on precision. Leveraging techniques such as data augmentation, transfer learning with ResNet50, and regularization ensures stability and generalizability. Evaluation on a 1311-image test set reveals outstanding class-specific accuracies (glioma: 98.33%, meningioma: 94.44%, no tumor: 100.00%, pituitary: 100.00%). Comprehensive metrics including precision (0.983559), recall (0.983219), F1 score (0.983140), and AUC (ROC) (0.999038) underscore the model’s efficacy. This study demonstrates the potential of deep learning in early brain tumor diagnosis, surpassing conventional methods and laying a robust foundation for future research in neural network-based classification algorithms for brain tumors.

Keywords

• Abnormal brain tumor comprehensive evaluation
• MRI DL
• segmentation
• types of brain tumor
• MRI