>>>>> On Wed, 09 Feb 2005 17:01:49 +0100, <[log in to unmask]> said:
Karsten> As one rule of thumb, you should not scale images, which are
Karsten> already scaled or where the voxel values are having a
Karsten> specific meaning or, more obvious, a unit; for example,
Karsten> quantified rCBF data from PET. So, in your case, it depends
Karsten> on the pre-processing of your DTI data, but, I guess, that
Karsten> the voxel values in your anisotropy maps are reflecting the
Karsten> absolute degree of anisotropy, i.e. values between 0 and 1,
Karsten> with 0 = isotropy. If that is the case, you should not scale
Karsten> individually.
Dear Karsten,
Thank you for your reply.
As you said, FA map has absolute value on each voxel. It is well
known, however, there are systemic errors inherent in quantitative
measurements and a wide variation across the subjects, especially on
CBF study. These factors could cause some misinterpretation.
Therefore, data normalization referred to some region or global mean
would be useful to enhance regional specificity. I'm not sure if this
idea also fits fractional anisotropy maps.
Any help would be really appreciated in advance.
--
Kohkichi Hosoda M.D.
Department of Neurosurgery,
Hyogo Emergency Medical Center/Kobe Red Cross Hospital,
1-3-1 Wakinohama-Kaigan-Dori, Chuo-ku, Kobe, 651-0073, JAPAN
Telephone 078-241-3131
Fax 078-241-2772
e-mail address : [log in to unmask]
|