Efficient Brain Tumor Segmentation in MRI Images using U-Net

The Brain tumor segmentation from MRI images is a crucial task in medical imaging, aiding in diagnosis, treatment planning, and monitoring patient progress. This project presents an efficient solution for automatic brain tumor segmentation using the U-Net deep learning architecture. U-Net, with its encoder-decoder structure, is specifically designed for biomedical image segmentation tasks, enabling precise localization of tumor regions in MRI scans. The model was implemented using Python, leveraging TensorFlow and Keras for building and training the network, while NumPy was used for data manipulation. A user-friendly interface was created using Streamlit, allowing for real-time image uploads and visualization of segmentation results. The project employs MRI images from a publicly available dataset, where the images are preprocessed to standard dimensions and normalized for consistent performance. The model was trained using binary cross-entropy as the loss function and optimized using the Adam optimizer, achieving high accuracy in segmenting tumor regions.