Hi SCT Team, I’m working with two T2-weighted spinal cord datasets:
GSP acquired: 0.8×0.8×0.8 mm resolution
and Clinical Protocol: 2×1×1 mm resolution
I am able to extract intramedullary lesion estimates and morphometrics, but the differing voxel sizes and slice counts make it challenging to compare the results accurately. Is there a recommended interpolation method or resampling technique to harmonize the clinical protocol data with the GSP data? Additionally, if you have any existing datasets or published studies using non-generic protocols that could serve as benchmarks for these comparisons, I would appreciate it. I have a sample dataset registered to PAM50 that I can share if anyone is interested in taking a closer look.
Thanks in advance for your help!
Fauziyya
Hi @Mfauziyya,
I’m missing some context, but if you wish to work with the same resolution, then you can indeed resample your clinical dataset to 0.8mm isotropic.
Thank you! To give you more context, I have two T2-weighted datasets; one acquired with a GSP at 0.8×0.8×0.8 mm resolution and another using a different protocol ‘clinical’ at 2×1×1 mm resolution. My goal is to compare intramedullary lesion estimates and morphometrics between these datasets.
I am not entirely sure what additional considerations you would recommend to ensure that the resampled clinical dataset remains valid for quantitative comparisons. for the isotropic, I registered the image directly to the template space using sct_register_to_template. However, for the clinical scan, do I need to register the template to the image first, or should I apply additional steps to the T2 image registration itself? How do I set up the different registration steps to achieve proper alignment for comparing these images?
I have shared a sample of the images with you via email. My goal is to be able to compare the estimates from these two images despite their different resolutions.
Thank you again for your assistance.
Fauziyya