Support:Documents:Manual:COMKAT utilities

From COMKAT wiki
Jump to navigation Jump to search

Input function estimation

This tool is designed to estimate the input function for small animal (rats and mice) without blood samples. One blood sample may be included for better accuracy but is not required. The manuscript of this method has been accepted for publication on Journal of Nuclear Medicine. If you would like to read the article please contact Dean Fang [1].

To use this function, run command 'inputest' or execute it under COMKATROOR/comkat_GUI.

A window will pop up. Click 'Load TAC curves'. Load a file in excel or csv that has time-activity curves for ventricular cavity and surrounding myocardium. After successful data loading, it should look like:

InputEstScreen 0.JPG

Then click on the popup menu of 'load the pre-set parameters'. Specify whether it's a rat or mouse. Adjust the parameters if desired. Then click 'Estimate the input'. Results should look like:

InputEstScreen 1.JPG

The estimated parameters of tau, A1~A3, L1~L3 can then be applied to the Feng input as the input function to a compartment model.

Diagnostic functions

Validation suite

The validation suite is aimed for users to validate a new COMKAT installation. It generates model output calculated by COMKAT and analytic solution. The user may then compare output from his/her computer and the developer's computer. The difference should be very little if not any if the installation is correct.

To run validation, run command 'validate' under Matlab. A window should pop up as:

Validate screen.JPG

Click on the model that you would like to compare your results to the developer's. Then wait without doing anything. (If you click on any of the figure during execution the display of results might be wrong.) Wait until a report is generated in your browser. Go to the results section and compare the figures of results from your computer and the developer's.

Graphical analysis functions

The graphical analysis may be called from the COMKAT GUI or with command line functions.

  • From the GUI

You must first have both the input function and experimental data defined. Click 'Tools -> Graphical methods'. On the GUI then click 'Apply and run analysis'. Results should look like this:

  • Command line functions

You may find these functions under COMKATROOT/graphicalMethods.

Image reading functions


We have a compiled version of DICOM loading function, called 'dcmreadfile' that can be found under COMKATROOT/ComkatImageTool.
imagedata = dcmreadfile(filename)
The output imagedata is a structure that contains information of all the fields stored as DICOM header. It also contains the pixel data as in imagedata.PixelData. In other words, it reads both the metadata (dicominfo function) and the pixel data (dicomread function). Because we used an advanced DICOM open-source library and wrote dcmreadfile as a mex file, the speed of loading DICOM files of dcmreadfile is about 10 times faster than dicominfo.


You can find the functions to load image files under 'ImageReadingFunctions'. The constructor for the imageVolumeData object can be found under the @imageVolumeData folder.

Blood activity monitor functions

The blood activity monitor (BAM) is a device to measure the activity in blood with continuous counting. You may find a GUI, BAM_converter_GUI, for processing these files including deconvolution and truncation under COMKATROOT/comkat_GUI.

The screen capture of the BAM converter