MicroPET (Positron Emission Tomography) is an imaging method used for small animals. It is a small scale equivalent to PET applied in medical diagnosis. The major drawback of PET is, that it does not provide precise anatomical information, because it just shows the distribution of the injected radio labeled molecules and no further body structure.
In order to gain quantitative information about the distribution of the tracer in organs, ROIs (Regions of Interest) have to be defined. Due to observing the measured activity in ROIs for a series of time frames, the time dependent distribution of the tracer can be calculated. Normally the ROIs have to be manually defined. Unfortunately the selected regions greatly depend on the user defining them, and selecting the regions may also be quite time consuming.
The software developed during this thesis simplifies and objectivizes the selection of ROIs of reconstructed PET images for mice. First, the user selects one or more clearly visible organs. On the basis of these user defined regions, the software calculates the positions of the remaining organs. This is done by adjusting the PET image to the digital map of an "average mouse" derived from microMRT (Magnetic Resonance Tomography). Therefore the organs provided by this mouse phantom are rescaled and rotated, until the organs reach their best overlap with the actual PET image. Thereby the location of the remaining organs can be estimated.
This document starts with a short introduction into the aim of the work.
Chapter 2 provides a general survey of PET and a short review of tracer production. In Chapter 3 the developed software is introduced, and its functional principles are discussed in detail. In Section 4.1 the microPET-instrument and measurement preparations are presented, and in Section 4.2 software-based results are summarized and compared to those resulting from user-defined selection. In Chapter 5 a concluding discussion is given. Finally, in Chapter 6 possible further use and development of the software is outlined.