Categorization of Brain Tumor Using CNN from 2D MRI Scan Images

In the realm of medical imaging, accurate identification, delineation, and categorization of brain tumors are of paramount significance for prompt intervention and treatment planning. This pioneering research introduces an innovative framework that harnesses Convolutional Neural Networks (CNN), Efficient neural network, Residual Network, and Mobile Network architectures for 2D brain tumor analysis. Deployment of deep learning methodologies to augment precision and furnish valuable insights for medical practitioners. The proposed methodology integrates skipconnection and encoder-decoder architectures, bolstered by data augmentation strategies. Grounded in tumors delineated by experts, particularly within a low-grade glioma (LGG) patient cohort, the framework extrapolates intrinsic affinities to synthesize thousands of representative cases. The efficacy of the proposed approach is substantiated through metrics such as pixel accuracy, F1 score, and MCL. Mobile Network Stood out in our comparative investigation, demonstrating its advantage in classification accuracy. Furthermore, this study sheds light on the consequential impact of brain tumor analysis, offering a promising avenue for advancing diagnostic capabilities and treatment outcomes in the field of neuro-oncology.