Hello,

I am concerned that the anatomical image headers for all of the Beijing_Zang data contain incorrect pixel dimensions. In particular, the y and z dimension (ant-pos and sup-inf) axes appear squashed in both the anonymized and skullstripped images. When I normalized either of these anatomicals to MNI space (ICBM-152 in SPM5) the y-z (saggital view) area was expanded by a mean factor of ~2.1 rather than the more reasonable ~1.4 factor that I got with the 19 other fcon_1000 dataset anatomicals that I have processed so far. Perhaps the correct pixel dimensions are approximately 1.33 x 1.2 x 1.2 rather than the stated 1.33 x 1.0 x 1.0? The fit of my normalization seems reasonably good, and the Beijing_Zang anatomicals are of good quality, so I don't think that I have a normalization problem (spm5 normalized all of them without incident). I also obtained an unusually low total brain area for the Beijing_Zang anatomicals when I do a segmentation of the anatomicals and sum the segmentation probabilities for WM and GM.

Thanks,

Timothy J. Herron

Hi Timothy,

I checked again and I am quite sure that the dimentions are 1.0*1.0*1.3mm. "an unusually low total brain area for the Beijing_Zang anatomicals" sounds interesting. How much? If no other errors, I guess the reasons should be considered, 1) east Asian; 2) This datasets have more females than males.

Yufeng

(zangyf@bnu.edu.cn; http://www.restfmri.net)

following up this, please see a recent paper published in NeuroImage, which might give us some about the differences in brains. "The construction of a Chinese MRI brain atlas: A morphometric comparison study between Chinese and Caucasian cohorts. (http://dx.doi.org/10.1016/j.neuroimage.2...)"

Xi-Nian