Why a slice location fallback?

In the Fourier tutorial there are a couple of methods used to try and calculate the distance between slices. The first method is to take the delta between two slice locations. The tutorial mentions that the reason is that some fields are missing in some images due to inconsistencies in MRI equipment. However, I’ve scanned all the SAX images from both the training and validation set and none of them appear to be missing the slice location field. Am I missing something else? Should we expect the final data set to be more incomplete than the two we’ve been given so far? Is this just a general statement that doesn’t really apply to the data for this competition?


Link to Full Article: Why a slice location fallback?