Brain Norm Deviation Functionality

The Brain Norm Deviation module of PNEURO supports the creation of normal brain PET databases from a consistent set of normal volunteer images and its application for analyzing the images of test subjects.

The process of constructing the normal pattern - named Normal Brain Database or short Brain Norm - in principle consists of the following steps:

▪The acquisition of images from a set of normal volunteers (controls). Preferably the same acquisition and image reconstruction protocols should be used as in the test subject studies.

▪The stereotactic normalization of the control images, so that the anatomy of the normalized images is comparable across the controls.

▪The scaling of the pixel values in the normalized images relative to an internal reference. The resulting normalized values allow pooling of the data.

▪The analysis of the values across the control collective in each pixel of the stereotactic anatomy. Hereby the normal values and their deviation across the set of normal controls is established in each pixel.

With a database-assisted analysis, the brain PET uptake pattern of a subject can be compared to the normal pattern. To this end, the subject images are normalized and scaled in the same way as the control images, and the resulting pixel values compared with the established normal values. This process results in a map showing the differences between the subject images and the normal pattern, expressed as a z-score value. The z-score map can be investigated in a multitude of ways including fusion with the subject images and 3D rendering (separate option).

Note: The same type of analysis is also possible with SPECT images, and with MR images using the option for voxel-based morphometry (VBM)