Deriche filter
Deriche filter can be used to estimate gradients at bigger spatial scales. This feature is useful when the original image resolution is higher than the spatial scale of the interesting features. For example, very strong gradients around vessels can be mitigated by decreasing the alpha paramater:
The resulting gradient magnitude images are similar to first applying a Gaussian smoothing filter to the image and then computing gradients (lower alpha = smoother gradients). It can be seen that gradient magnitude images of 700 micron data become blurrier faster than 250 micron data. This indicates that the spatial scales of the gradients can be matched by choosing different alphas. When you observe the changes, it is apparent that 700 micron data with alpha=2 provides similar gradient magnitude compared to 250 micron data with alpha=1.
- Use
deriche_alphaparameter together withderichegradient magnitude method:
segmentator /path/to/file.nii.gz --gramag deriche --deriche_alpha 2
- You can also export the gradient magnitude image as a separate nifti file:
segmentator /path/to/file.nii.gz --gramag deriche --deriche_alpha 1.5 --export_gramag
-
Monga, O., Deriche, R., & Rocchisani, J.-M. (1991). 3D edge detection using recursive filtering: Application to scanner images. CVGIP: Image Understanding, 53(1), 76–87. http://doi.org/10.1016/1049-9660(91)90006-B
-
[Data: T1w 700 micron iso.] Gulban O.F., Schneider M., Marquardt I., Haast R.A.M., De Martino F. (2017). Dataset: A scalable method to improve gray matter segmentation at ultra high field MRI. https://doi.org/10.5281/zenodo.1117859
-
[Data: T1w 250 micron iso.] Lüsebrink F., Sciarra A., Mattern H., Yakupov R., Speck O. (2017). Data from: T1-weighted in vivo human whole brain MRI dataset with an ultrahigh isotropic resolution of 250 μm. Dryad Digital Repository. https://doi.org/10.5061/dryad.38s74