3 Probabilty Maps Normalization (SPM8)

<< Click to Display Table of Contents >>

Navigation:  Processing using Combined Matching Layout > Processing Overview > Initialize/Match page > Automatic Matching > Probability Maps Normalization >

3 Probabilty Maps Normalization (SPM8)

The 3 Probability Maps Normalization (SPM8) is based on PMOD's Java implementation of the Unified Segmentation methodology developed in SPM8 by Ashburner et al [11].

The 3 Probability Maps Normalization (SPM8) button provides access to the different types of probability matching methods. An interface dialog window appears with two tabs. The first one serves to quickly retrieve species sensitive pre-defined or saved configurations of the iterative matching settings, the second one gives access to all parameters.


The following parameters allow fine-tuning the algorithm.


Basic Parameters

Sampling rate

Pixel sampling rate for the calculation.

Input mask

A mask file can be selected which masks the part of the input image which should be disregarded in the normalization.

Advanced Parameters

The Advanced parameters are usually only changed if a normalization fails or if the user aims at a specific effect.


Image denoising prior to the normalization using the fast Non Local Means Analysis method with settings None, Low, Medium, Strong.

Nonlinear Warping

Enable elastic deformation in addition to the affine transformation.

Bias Regularization

Serves for compensating modulations of the image intensity across the field-of-view. Depending on the degree of the modulation, a corresponding setting can be selected from the list: None, Very Light, Light, Medium, Heavy, Very Heavy. The parameter to the right indicates the FWHM [mm] to be applied. The larger the FWHM, the smoother the variation that is assumed.

Affine regularization

Two different initializations of the affine registration are supported, European brains and East Asian brains, as well as No regularization. The setting should correspond to the nature of the subject under study.


Note: The results of the probability normalization for the small animal (RAT) might result inappropriate due to the poor quality of the input image.