Deep CNN Optimization Method for MRI Image-Based Brain Tumor Identification

We use magnetic resonance imaging (MRI) scans to find brain tumors, which can harm the brain and appear suddenly. These methods can be untrustworthy and time-consuming. Deep learning systems analyze MRI images to predict brain tumors. This work used the BraTS 2019 and 2020 datasets to create a deep CNN model. To make the model more accurate and efficient, we used white-headed wood optimization as a baseline. This advanced optimization method greatly increased the number of brain tumors that researchers could find. This helps the tool be more dependable for diagnosing early and planning treatment. White-headed timber-based optimization enhances the model following preprocessing, segmentation, and feature extraction. Accuracy (Acc), sensitivity (Sen), and specificity (Spe) are the three terms. We achieved similar results for d1 (98.38%, 98.32%, and 98.89%) and d2 (97.75%, 98.54%, and 99.16%) in a 10-fold test. The work aims to develop a deep CNN using WHTO to predict brain cancer based on MRI data.