Modern best practices for image segmentation tasks

I have been working on annotating medical images for an image segmentation task using a UNet. However, I find the entire process painstaking and long. Unfortunately there is also a lot of data that needs to be annotated due to the poor signal to noise in the images. I recently came across monai and concepts like interactive segmentation and segment anything paper. It made me realize that maybe my current approach for annotating is the wrong way to go about …

best practices computervision data however image images medical modern monai noise practices process segmentation signal tasks unet