| input files | Show / Hide |
Wavelet transform: sub ---> 1 of 64 Wavelet transform: sub ---> 2 of 64 Wavelet transform: sub ---> 3 of 64 Wavelet transform: sub ---> 4 of 64 Wavelet transform: sub ---> 5 of 64 Wavelet transform: sub ---> 6 of 64 Wavelet transform: sub ---> 7 of 64 Wavelet transform: sub ---> 8 of 64 Wavelet transform: sub ---> 9 of 64 Wavelet transform: sub ---> 10 of 64 Wavelet transform: sub ---> 11 of 64 Wavelet transform: sub ---> 12 of 64 Wavelet transform: sub ---> 13 of 64 Wavelet transform: sub ---> 14 of 64 Wavelet transform: sub ---> 15 of 64 Wavelet transform: sub ---> 16 of 64 Wavelet transform: sub ---> 17 of 64 Wavelet transform: sub ---> 18 of 64 Wavelet transform: sub ---> 19 of 64 Wavelet transform: sub ---> 20 of 64 Wavelet transform: sub ---> 21 of 64 Wavelet transform: sub ---> 22 of 64 Wavelet transform: sub ---> 23 of 64 Wavelet transform: sub ---> 24 of 64 Wavelet transform: sub ---> 25 of 64 Wavelet transform: sub ---> 26 of 64 Wavelet transform: sub ---> 27 of 64 Wavelet transform: sub ---> 28 of 64 Wavelet transform: sub ---> 29 of 64 Wavelet transform: sub ---> 30 of 64 Wavelet transform: sub ---> 31 of 64 Wavelet transform: sub ---> 32 of 64 Wavelet transform: sub ---> 33 of 64 Wavelet transform: sub ---> 34 of 64 Wavelet transform: sub ---> 35 of 64 Wavelet transform: sub ---> 36 of 64 Wavelet transform: sub ---> 37 of 64 Wavelet transform: sub ---> 38 of 64 Wavelet transform: sub ---> 39 of 64 Wavelet transform: sub ---> 40 of 64 Wavelet transform: sub ---> 41 of 64 Wavelet transform: sub ---> 42 of 64 Wavelet transform: sub ---> 43 of 64 Wavelet transform: sub ---> 44 of 64 Wavelet transform: sub ---> 45 of 64 Wavelet transform: sub ---> 46 of 64 Wavelet transform: sub ---> 47 of 64 Wavelet transform: sub ---> 48 of 64 Wavelet transform: sub ---> 49 of 64 Wavelet transform: sub ---> 50 of 64 Wavelet transform: sub ---> 51 of 64 Wavelet transform: sub ---> 52 of 64 Wavelet transform: sub ---> 53 of 64 Wavelet transform: sub ---> 54 of 64 Wavelet transform: sub ---> 55 of 64 Wavelet transform: sub ---> 56 of 64 Wavelet transform: sub ---> 57 of 64 Wavelet transform: sub ---> 58 of 64 Wavelet transform: sub ---> 59 of 64 Wavelet transform: sub ---> 60 of 64 Wavelet transform: sub ---> 61 of 64 Wavelet transform: sub ---> 62 of 64 Wavelet transform: sub ---> 63 of 64 Wavelet transform: sub ---> 64 of 64
| matlab report | Show / Hide |
- There are 487941 non-zero coefficients in the mask image (spatial domain).
- There are 11039 non-zero wavelet coefficients in the resulting filtered mean image,
which were selected by the Minimum Description Length (MDL) Method.
- These coefficients are able to recover an energy percentage equal to 99.0128%.
Using the contrasts -> -1 1 0 0
and -> 1 -1 0 0
- The higher p-value considered as significant by the FDR algorithm is -> 0.0049898.
- From the 556 significant coefficients there are 288 positives and 268 negatives.
- The percentage of significant coefficients is -> 5.0367%.
--- significant results ---
1: contrast1_corr_0.05_explained_effect_size.nii -> is the image containing significant positive differences (G1>G2), or positive correlation, at p-value = 0.05.
2: contrast2_corr_0.05_explained_effect_size.nii -> is the image containing significant negative differences (G2>G1), or negative correlation, at p-value = 0.05.
--- additional images -----
3: mean_filtered.nii -> is the filtered between-group mean image in the spatial domain.
4: mean_residual.nii -> is the residual part of the between-group mean image, obtained by the inverse transform
of the wavelet coefficients that were not selected in the MDL data reduction algorithm.
*suggested thresholds to view the image (lower=0.67179, upper=2.0154)
*only are depicted regions where the observed changes in the gray matter tissues are higher in absolute value than 0.67179
| model selection | Show / Hide |
![]() |
See the following reference: Anestis Antoniadis, Irène Gijbels, and Gérard Grégoire.
Model selection using wavelet decomposition and applications. Biometrika, 84:751-763, 1997.
The methodology relies on the minimum description length criterion, which is used
to determine the number of non-zero coefficients in the vector of wavelet coefficients. |
date: 09-May-2012. Hot color (red-yellow) corresponds to significant positive G1>G2 differences (if any), or positive correlations. Winter blue color corresponds to G2>G1 significant negative differences (if any), or negative correlations
