The detection of tumors from MRI images remains a crucial aspect of healthcare diagnostics, necessitating accurate segmentation for effective diagnosis and treatment planning. However, conventional segmentation methods face challenges such as image noise, tumor characteristic variations, and limited dataset sizes. To address these challenges, we propose a novel approach that combines the U-Net architecture with advanced image augmentation techniques. U-Net, a convolutional neural network (CNN) architecture, has demonstrated remarkable performance in biomedical image segmentation tasks by capturing both local and global features. By leveraging this capability, we aim to enhance the accuracy and robustness of tumor segmentation in MRI images. Additionally, to mitigate the limitations of small datasets and improve generalization, we employ sophisticated image augmentation strategies such as rotation, scaling, and elastic deformations. These techniques allow us to artificially expand the dataset and introduce variations resembling real-world scenarios. We adhere to strict privacy protocols to ensure the confidentiality and integrity of patient data, with all MRI images anonymized and compliant with medical data regulations. Experimental results on an MRI image dataset show promising improvements in tumor segmentation accuracy compared to conventional methods. Through rigorous evaluation and comparison with existing approaches, we validate the effectiveness of our proposed framework in enhancing tumor detection capabilities. In conclusion, the integration of U-Net architecture with advanced image augmentation techniques provides a robust solution for tumor detection in MRI images, contributing to the advancement of computer-aided diagnosis systems in healthcare and potentially leading to more accurate and timely diagnosis, thus improving patient outcomes.