Question Posted 09/23/16:
Dear MRI-Experts,
We are wondering why the ADNI T1 sequences resulted in non-isotropic voxels. In most cases isotropic voxels were preferred, especially when the images are analyzed by automatic (group or population-based) image analysis procedures. In these studies, usually the images will be aligned to a common coordinate frame and this requires rotations of the voxel array that will misalign the original acquisition matrices. Thus, in many voxel-based morphometry studies 1x1x1 mm resolutions were preferred.

Another point is that the field of view differs between the MR sequences and seems to have a relatively big FOV in RL direction.

Will the ADNI3 protocols use the same basic sequences or are you changing the ADNI standard in respect to these points.

In our neurological hospital (at University Hospital Schleswig-Holstein (UKSH), Kiel, Germany) we are currently planning to establish new standard MRI protocols. These protocols will be applied to several patient groups (mainly dementia & movement disorders).

We are thinking about the possibilities to join and support the ADNI project in the future. Our idea is to make most of our sequences compatible with the standardized ADNI MRI protocols.

Thank you very much in advance.

Best regards
Oliver Granert
Response posted 10/11/16 by ADNI MRI Core:
Thank you for your question to the ADNI MRI Core.

The reason for the non-isotropic voxels is really just a carry forward from what was possible when ADNI1 started nearly 12 years ago. We wanted to try to keep the T1 data as consistent as possible throughout the study.

Thank you