|
|
|
|
|
Input Data
Whenever possible, a brain PET study should be complemented by an anatomical T1-weighted MR study with isotropic high resolution in the order of 1 mm covering the entire brain. This will allow the accurate adjustment of the brain structures to the patient anatomy and conveniently support any interactive fine adjustments.
Evaluation Space for PET Statistics
The calculation of the PET VOI statistics can be performed in different spaces, depending on the available input data: the MNI template space, the MR space and the original PET space. If the user would like to avoid any interpolation of the original PET values, he should evaluate the VOIs in the PET space. In this case he is strongly encouraged to reconstruct the PET images with a pixel size of about 1 mm. Otherwise, the brain VOIs will be truncated and become coarse.
The resolution effect on the VOIs is illustrated below. The left image shows the brain contours with the resolution of an MR image (0.8mm in-plane), the right with the resolution of a PET image (2.4mm). Note that the correspondence of the slices is only approximate.
Image Cropping
The algorithms in the PNEURO modules work best if the images do not contain too much information from outside the brain. Therefore, after loading the MR and PET images, PNEURO offers a cropping facility as illustrated below.
To enable the cropping option, please check the Crop box. Consequently, yellow rectangles are overlaid on the images indicating the extent of the cropping box. To move the center of the box, please click into the image. To change the box size in the three directions, use the corresponding selections below MR Image. Once the cropping box is properly placed, activate the MR Image (or PET image) button to perform the (irreversible) image cropping. If the cropping is not done, a confirmation message will appear when progressing to the next stage. After performing a cropping operation, the function is blocked. In order to apply a different crop box the data has to be loaded again.