Dear Zhao,
usually if you see the same GM effect with the inverse contrast in WM this is due to a lack of an absolute threshold in the statistical model. You have to prevent that WM locations are analyzed in the GM analysis. Therefore, an initial absolute threshold of 0.1 is recommended that can be finally increased up to 0.25 (check that no important effects are too much cut).
Best,
Christian
On Thu, 12 May 2016 16:02:09 -0400, Wang Zhao <[log in to unmask]> wrote:
>Hello everyone,
> I have some questions for my VBM analysis result. I compared the
>brain structure of Chinese dyslexia children (DD), age-matched
>controls(AC) and reading-matched controls (RC), and found that one of
>overlapped grey matter clusters of DD>AC and DD>RC (30, 0, -2) had
>almost the same location as the overlapped white matter cluster of
>AC>DD and RC>DD (30, -6,-5). I?m not sure why it happened, and I also
>don?t know how to report this result, because I can?t decide whether
>this cluster belongs to grey or white matter. I list all of steps I
>did during the analysis, could you help me to check it?
>
>1.Using TOM8 to create TPM template based on the gender and age for my
>subjects. They are all Chinese children from 8-11 years old.
>
>2.The VBM8 workflow:
>1)Normalized and Segmented:
>VBM8 ? Estimate and Write
>Parameters:
>Volumes<-X?Select Files?[select raw data]?Done
>Estimation options:
>Tissue Probability Map?[use the template made by TOM8 ]
>Gaussians per class?[use default]
>Bias regularisation?[use default]
>Bias FWHM?[use default]
>Affine Regularisation?[*ICBM space template-Asian brains]
>Warping Regularisation?[use default]
>Sample distance?[use default]
>Extended options:
>Dartel Template?[use default]
>Use SANLM de-noising filter?[use default]
>MRF Weighting?[use default]
>Clean up any partitions?[Light Cleanup]
>Display and print results?[use default]
>Writing options?[use default]
>File?Save Batch
>File?Run Batch
>
>2)Check data
>VBM8 ?Check data quality?Display one slice for all images
>Parameters:
>Volumes<-X?Select Files?[select the new files]?Done
>Proportional scaling?[yes]
>Show slice in mm?[use default]
>File?Save Batch
>File?Run Batch
>
>VBM8 ?Check data quality?Check Sample Homogeneity Using Covariance
>Parameters:
>Volumes<-X?Select Files?[select the ?m0wrp1? files]?Done
>Proportional scaling?[yes]
>Show slice in mm?[0]
>Gap to skip slices?[1]
>Nuisance?[Enter nuisance variables]
>File?Save Batch
>File?Run Batch
>
>3)Smooth
>SPM menu?Smooth
>Parameters:
>Images to smooth<-X?Select Files?[select the ?m0wrp1?/?m0wrp2? files]?Done
>FWHM?[8,8,8]
>Data type?[use defaults]
>Filename Prefix?[use defaults]
>File?Save Batch
>File?Run Batch
>
>3.Group comparison
>
> I don?t know whether the pre-processing causes the mixed result of
>grey matter and white matter analysis. Should I change the TPM
>template and ICBM template to the default choice? Or should I just
>treat it as some characteristics of Chinese dyslexia children? But how
>could I report this result in the paper?
>
>I?m looking forward for your reply. Thank you for your help!
>
>Yours,
>Zhao Wang
|