Multi-Class Brain Tumor Image Classification Utilizing ResNet50: A Transfer Learning Approach for Medical Imaging

There is a lot of hope that medical diagnostics could be improved by using deep learning intended for brain tumor categorization. In order to categorize brain malignancies, there are 4 distinct type’s glioma, pituitary tumor, meningioma and no tumor. This work presents an improved convolutional neural network (CNN) model based on ResNet50. To enhance model convergence, a systematic technique was used to alter the learning rate. An outstanding test accuracy of 98.54% and a classification accuracy of 98.48% were attained by the experimental results, which were validated using a benchmark MRI dataset. This study optimizes ResNet50 with a refined learning rate strategy, achieving high accuracy (98.54%) in brain tumor classification. The integration of explainable AI enhances model transparency, making it more reliable for clinical decision support. To make the model's decision-making process clear, our methodology uses explainable AI techniques like visual analytics and confusion matrices. Clinical decision support is improved by results that validate our improved ResNet50 reliability with further consistency in tumor categorization of humanbrain.