The first operation in our pipeline is the Input ROI. This operation is included by default in every pipeline and allows us to limit the portion of the dataset to be analysed. The main effect of restricting the Input ROI is to speed up pipeline execution. This is particularly useful during the pipeline building process as we might go back and forth through the pipeline trying to optimise the parameters, but can also be used to facilitate the segmentation of very large objects (objects typically larger then 1000 pixels across), and to avoid wasting time processing portions of the image that are not of interest without the need to crop the data.