Question
Question Posted 01/12/25:
Hi. How can I apply gradient unwarp to the dMRI imaging. Do you have the coefficient file of the scanners? Besides, could you tell me how can I acquire the metadata of dMRI imaging? Thanks in advance.
Hi. How can I apply gradient unwarp to the dMRI imaging. Do you have the coefficient file of the scanners? Besides, could you tell me how can I acquire the metadata of dMRI imaging? Thanks in advance.
Response posted 01/12/25 by Rob Reid:
Hello, scanners have (usually) been automatically applying gradient unwarping as part of reconstruction for years, and since ADNI1 did not have dMRI, most ADNI dMRIs have already had gradient unwarping.
But I suspect what you really want is to correct the diffusion gradients for spatial variation; neither the scanners nor (I believe) ADNI do that. b is proportional to g^2, so it definitely has an effect in the outer parts of the head when gradient nonlinearity is an issue. However, ADNI always scans with the head at the isocenter of symmetric full-body scanners, so the gradient nonlinearity effect is small* within the brain. Logistically, it has become more difficult to get the gradient nonlinearity coefficients as scanner manufacturers have moved to keep the spatial unwarping step internal. Please let us know if you manage to get any.
* smaller than the effect of imaging gradients on the scanners I have checked.
The DICOM headers are the best way to get the scan parameter metadata.
But I suspect what you really want is to correct the diffusion gradients for spatial variation; neither the scanners nor (I believe) ADNI do that. b is proportional to g^2, so it definitely has an effect in the outer parts of the head when gradient nonlinearity is an issue. However, ADNI always scans with the head at the isocenter of symmetric full-body scanners, so the gradient nonlinearity effect is small* within the brain. Logistically, it has become more difficult to get the gradient nonlinearity coefficients as scanner manufacturers have moved to keep the spatial unwarping step internal. Please let us know if you manage to get any.
* smaller than the effect of imaging gradients on the scanners I have checked.
The DICOM headers are the best way to get the scan parameter metadata.
