ADAPTIVE SEA LION OPTIMIZED RESNET50 (SLIA-RN50) FOR ENHANCED MRI TUMOR CLASSIFICATION

MRI-based tumor classification requires high precision, adaptability, and computational efficiency to ensure reliable diagnosis. Conventional deep learning models struggle with feature extraction, misclassification, and optimization inefficiencies, limiting their effectiveness in medical imaging. The purpose of this study is to enhance classification performance by integrating bio-inspired optimization with deep learning. A Sea Lion Inspired Adaptive ResNet50 (SLIA-RN50) framework is introduced to optimize hierarchical feature extraction, improve adaptability, and enhance decision-making in MRI tumor classification. The methodology involves leveraging sea lion-inspired strategies to refine multi-scale integration, collaborative filtering, and energy-efficient processing, ensuring balanced precision and recall. The proposed SLIA-RN50 model outperforms existing architectures in classification accuracy, precision, and F-measure, significantly reducing false positives and false negatives. Experimental results validate the effectiveness of the optimized framework, confirming its potential for automated MRI-based tumor detection. The bio-inspired model presents an efficient and scalable solution for improving computational techniques in medical image classification.