This manual explains the capabilities of QModeling, and the way to use it. It is divided in four main sections: Graphical User Interface (GUI), Implemented models, Command line functions and Include your own model.
The Graphical User Interface (GUI) section, which is organized in a questions and answers format, has four different subsections according to the four main steps to be followed when using the toolbox. Each of them is carried out by a specific submodule: Data loading and TAC generation; model selection and preprocessing; generation of parametric images and display of parametric images. Thus, QModeling allows generating parametric images for a particular dynamic PET study, by fitting one of the implemented models (SRTM, SRTM2, Patlak or Logan) to the data. Brief information and references about the four compartmental models are included at the Implemented models section.
You can also fit each of the four compartmental models to multiple studies using the Command line functions. Three command line functions (to load studies, preprocess them and generate whole brain images) are implemented for that purpose.
At the Include your own model section, it is described how to include in the QModeling software the implementation of a new kinetic model.
Graphical User Interface (GUI)
The main window is divided into six different parts or subframes. The four of them placed at the right and the bottom of the window represent the four submodules of the toolbox. The other two show the graphical and numerical results, at the right of the main window. When the program is started, all the subframes are inactive except for the first one. After introducing the relevant input data, the following frames will be activated.
In this flowchart you can see some pictures of the different steps.
Command line functions
In order to fit one of the four kinetic models included in QModeling to multiple studies, three main command line functions have been implemented:
QModeling_generateTACs.m: To load multiple PET studies (each one with a reference and a target mask) and to obtain the TACs from all of them.
QModeling_preprocessTACs.m: From the reference and target TACs of multiple studies, this function allows the user to fit one of the four kinetic models.
QModeling_parametric_images.m: To obtain the parametric images from the fitting of one of the kinetic models implemented to multiple PET studies (pixel-wise calculation).
To obtain more information about one of these functions, type help and the name of the function at the Matlab command window. Furthermore, a demo script is included at the command line folder of the toolbox to show the user how the command line could be run. It is called demoQModeling.m and typing help demoQModeling.m at the command window, more information will be displayed. Also, there are many comments on the script to help the user to understand all the steps.
Include your own model
The aim of this section is to provide users with the necessary information to modify the QModeling software, in order to include in it additional kinetic models. These instructions only refer to the implementation of this new function via command line. Users are welcome to contact the authors for any further advice, of if they are interested in including a model in the GUI, after developing it for the command line. This information has not been included in this manual for the moment, as adding a new model to the GUI requires modifying some more routines in the package.
A new Matlab function has to be developed, under the name QModeling_newModel_Name. Here, newModel_Name is the name that you want to give to the model. The three already implemented functions (QModeling_SRTM.m, QModeling_SRTM2.m and QModeling_PatlakRef.m) can be taken as examples. The new function should be run from QModeling_preprocessTACs.m and from QModeling_parametric_images.m. With the first one, the fit will be done at a ROI level. The second one will yield parametric images, after a voxel level fit.