Brain Tumor Classification Leveraging SWIN Transformer with Hybrid Whale Remora Optimization Based Hierarchical VGGNet19

Brain tumor classification is a critical aspect of medical diagnostics, demanding accurate and efficient methodologies for improved patient care. With advancements in medical imaging, the need for sophisticated techniques to analyze and classify brain tumors has become increasingly apparent. Traditional methods often fall short in capturing the complexity of tumor characteristics, necessitating the integration of state-of-the-art technologies to enhance diagnostic precision. The proposed methodology integrates four key steps. First, image enhancement is accomplished using an Improved Weiner Filter to refine the input data, enhancing the quality of features for subsequent analysis. Next, the Shifted Window (SWIN) Transformer is employed for segmentation, effectively delineating affected regions and providing a foundation for accurate classification. The third step involves leveraging a Hierarchical VGGNet19 for the classification of brain tumors. This deep learning architecture is processed to discern intricate patterns within medical images, ensuring precise categorization. Finally, hyperparameter tuning is executed using a novel Hybrid Whale Remora Optimization approach. This innovative optimization technique fine-tunes the parameters of the model, enhancing its robustness and accuracy. The experimental results showcase the effectiveness of the proposed model in achieving state-of-the-art performance in brain tumor classification tasks.