A CNN Approach for Brain Tumor Detection Using Diverse Optimizers

The ability to recognize brain tumors via medical imaging is crucial for brain tumor treatment. This proposed work focused on a technique for detecting brain tumors by employing Convolutional Neural Networks (CNNs) to achieve precise and prompt diagnosis of brain tumors. This frame utilizes a variety of optimizers such as Stochastic Gradient Descent, RMSprop, Adam (Adaptive Moment Estimation), Adadelta, Adagrad, Adamax, and Nadam. The model’s performance is evaluated using accuracy, sensitivity, specificity, precision, and recall as the primary metrics. Leveraging a dataset of 6951 MRI images, results reveal varying performance across optimizers, with a peak accuracy of 98.8% emphasizing the need for careful selection based on specific requirements and dataset characteristics. Advancing brain tumor identification, aiming to enhance early intervention and patient outcomes through the integration of cutting-edge technology, particularly Convolutional Neural Networks (CNNs), which offer high precision in analyzing intricate patterns within medical imaging data.