﻿ Glossary of Terms

# Glossary of Terms

##### Normal Database

In certain situations it is possible to compare patients images against a reference image to localize alterations from the normal uptake pattern. There are two requirements for such methods:

1. Anatomical normalization: To compare the patient images with the reference the patient anatomy must be adjusted to the standard anatomy. There are two domains where this task has proven to be successful. With heart studies the anatomy is usually reduced to a simple geometric model such as an ellipsoid. With brain studies, an elastic transformation is derived which warps the patient images into the stereotactic coordinate system. If functional brain images do not show sufficient anatomical details, the transformation may indirectly be obtained using a set of anatomical brain image of the same patient, such as magnetic resonance images.
2. Standard uptake pattern: Studies must be performed with a large enough set of normal controls, the data anatomically normalized, and the results statistically analyzed. It is important to bear in mind that patient preparation, the acquisition protocol, and reconstruction usually have an impact on the resulting images and therefore study guidelines must be established and followed. As tissue function tends to change with age, the control group should ideally be age-matched to the target patient group. The statistical analysis has to demonstrate a consistent uptake pattern with a sufficiently small standard deviation.

In PMOD a Normal database is also called a Brain Norm.

##### Origin

The origin of an image defines the center of the coordinate system in mm calculated from the front upper left hand corner. The origin is a concept of Analyze data written by SPM. For formats that support images positions (DICOM, ECAT) the origin values are calculated from the position of first slice. For other formats the origin setting is available for change in the loading dialog. If it is not changed, the default values are used.

##### z-score

The z-score defines the deviation of a sample with respect to the mean of a distribution. It is defined by the formula

z = (x-m)/s

where x is the sample value, m the sample mean, and s the standard deviation of the distribution. Therefore z describes the deviation from the mean in number of standard deviations and is positive, when the sample is above the mean, and negative when below.