Funky T2 Acquisition/Reconstruction Artefact?

Hi, all. Hoping someone here have seen similar acquisition issues like this before as our hospital based MRI physicist is still not able to resolve this despite some preliminary debugging.

This is a fairly standard, non-PROPELLER T2 FRFSE sequence on a GE 750W but getting some signal clipping/dropping on certain planes but not others. I have attached an example screenshot as shown below.

, which is similar to what we see in actual research subject.

The issue is, this seems to only affect this sequence in particular, other EPI, T1, scout/Diffusion sequences seems have no such issues. The overall signal reduction seems always on the first few slices and last few slices (where there were less tissues in the slice?).

Any tips or insights greatly appreciated.

More image:

Are we sure this is an acquisition artifact and not an image conversion artifact. If this issue does not appear when looking at the image on the console, but does appear in the NIfTI image, than it might be a conversion error. I think this is unlikely: dcm2niix should detect and correct if the image scaling varies between slices. However, it is worth checking out. Also, can I suggest you use the latest v1.0.20180614 version of dcm2niix? If this is a conversion artifact, please create an issue on the dcm2niix Github page.