An automated system for quantification/analysis of mouse PET images

Learning Objectives ROI values, TACs and biological parameters derived automatically from the system.

Summary: We have developed an integrated software system to automate the image analysis of mouse FDG PET images, which includes extracting regions of interest (ROIs), time activity curves (TACs) and biological parameters such as standard uptake value (SUV), transport and metabolic rates and receptor/enzyme density. The software uses small animal PET and CT data together with other pre-scan data (mouse weight, input function, etc.). Whole body CT data was spatially normalized and registered to the Digimouse atlas CT template using a Matlab-based SPM compatible software. Using this information, the ROI data was extracted from the PET images for six major organs; heart, liver, kidney, brain, lung and bladder. SUVs, rate constants and distribution volume can be estimated using the image data together with separately obtained input function data and an in-house developed program. The new generated data is also available online for users to retrieve. Modular design is incorporated in the system to allow easy expansion to accommodate future development and improvement of analysis techniques and for other tracers. The system is also constructed to be easy to adapt to various scanners and platforms. All ROI values, TACs and biological parameters derived automatically from the system were comparable to those by manual processing that is time consuming and labor intensive. We have developed an automated system for image analysis for routine mouse PET studies that is expected to shorten the image analysis time of mouse PET study by more than an order of magnitude.