A Hybrid Approach for Customized Skull-Stripping MR Brain Images

In high resolution MRI brain images, skullstripping procedures are typically considered to be crucial for removing non-brain tissues that provide a significant obstacle for computerized brain image segmentation and analysis methods using convolutional neural networks. Still, existing applications disregard image freedom and their rigidity makes it difficult to adapt to outside needs. In this study we propose a hybrid method which allows users to interact with the image through Matlab graphical user interface (GUI) to generate customized skullstripping (CSS) images. This application is valid for T1-weighted and T2-weighted MRI brain images.