Hi @hcp4715,
sorry for the late reply.
Yeah, keeping on is one of the most crucial parts!
Ok, glad it solved the first issue.
Now to the rest. First of all, could you provide us with further information on your setup and the corresponding sequences? Looks like a Siemens scanner, no?
The validator is right and gives you a hint that the two DTI sequences you have apparently differ in some sort, which is also visible in the dcm_dir_name and the dim3. Could you let us know what’s the difference between both? Maybe a diverging phase encoding direction (as e.g. evident in the tutorial you linked)? However, checking your dicom.tsv that might not be all, which transitions into your second question.
The relationship between number of dcm and nii also depends on your setup and the data acquisition protocol parameters. For example using a Siemens scanner there’s something that is called mosaic format, that is a DICOM format in which a 3D image is stored on a 2D grid instead of a series of 2D slices. For example, if you check your resting scans: they have dim3=60 and dim4=480, meaning that the 60 slices you acquire within each TR are on one 2D grid, aka the mosaic and therefore you have 480 dcm. If that wouldn’t be the case your number of dcm would be n slice X n measured volumes, that is 28800 dcm. Now within the t1w, t2w and dwi you see that dim3 matches your number of dcm, as you sample the measurement volume once and not multiple times, that is one dcm for each slice.
Wrt your dwi scans: if we divide dim3 of the first dwi sequence by dim3 of the second, that is 4680/72, we get 65, which might hint to dim4 of the first dwi sequence being stacked in dim3. I’m by far no DTI person, but that should be the number of volumes, aka your gradient directions. So, instead of having one 3D mosaic for each gradient direction, you have n slices X n gradient directions. As this furthermore stresses an existent difference wrt your second dwi scan, I would suggest you define two separate dwi identifiers, for example indicating the phase encoding direction (for an example check this neurostars post).
I’m sorry for this confusing answer, I’m super bad at explained these things.
Let’s hope other, more expert-mode folks will intervene if I’m completely off…
Cheers, Peer