CNN for Small Bowel Segmentation

I’m working on a small project focused more on the logic behind a CNN model rather than the final result. I have 3D medical images (TC and RM) of one patient, along with ground truth data for some visible parts of the small bowel. I’d like to use these images to create and train a CNN model, applying data augmentation techniques to enhance the dataset. I found an article that outlines the process of building it.