How to execute matRad

To execute the matRad script using MATLAB you need to:

1. Open matRad folder in MATLAB
2. Set patient-specific parameters
3. Execute inverse planning
4. Import additional patient data

If you prefer to only use the GUI to execute matRad, check out the matRadGUI tutorial.

Step 1: Open matRad folder in MATLAB

To use matRad you need to open the matRad folder in MATLAB. Open MATLAB and navigate to the location of the files, if you have cloned the repository it is most likely located in your local Github folder (e.g."C:\Users\username\Documents\GitHub\matRad").

Inside the matRad folder there are several MATLAB-functions used to run matRad, named "matRad_*.m, and *.mat files containing base data and exemplary patient data sets. The main script to run matRad is called "matRad.m". It can be executed section by section.

Step 2: Set patient-specific parameters

In the first section the patient specific parameters have to be set (see image below).:

1. Which patient (-data) should be loaded
2. Which beam angles should be used
3. Which radiation mode should be used

**Selecting a patient**

The lines 24-28 in the screenshot show the patient data sets available by default. Un-comment the data set you wish to use. The dose parameters for the different Volumes (min. dose, max. dose, penalties) are set within the patient data set (cst). If you wish you can adjust these parameters before executing matRad.

**Selecting beam angles**

The lines 32-33 in the screenshot are used to set the gantry and couch angles. Here you can set any angles from 0-359°. Make sure, that you always create pairs of gantry and couch angles, otherwise you won't be able to execute the inverse planning!

**Selecting radiation mode**

The Radiation mode can be set in line 38 in the screenshot. You can choose between photons, protons and carbon.

If you decide to use protons or carbon, it is possible to set the lateral spot spacing (line 31). When using carbon you can also choose between a physical optimization ('none'), an optimization of the biological effect ('effect') or an optimization of the RBE-weighted dose ('RBExD') by adjusting the parameter pln.bioOptimization in line 39.

In case you choose photons it is possible to run an additional MLC sequencing by setting pln.runSequencing and direct aperture optimization is accessible through pln.runDAO.

The desired number of fractions can be set in line 41 in the screenshot.

The other parameters set in this section are generated automatically and should not be changed.

_Screenshot of the parameters section_

Step 3: Execute inverse planning

The matRad.m script can now be executed step by step:

1. load settings
2. initial visualization
3. generate steering file
4. dose calculation
5. inverse planning for IMRT
6. sequencing
7. direct aperture optimization
8. visualization of the resulting treatment plan
9. cShow DVH and quality indicators

To evaluate a single section you have to "activate" it (left mouse-click inside section) and then use "ctrl + enter" or use right mouse-click -> Evaluate Current Section.

**1. load settings**

Now you can execute the first section. You should see the variables cst, ct and pln in your Worspace.

**2. initial visualization**

After the patient data is loaded you can execute the second Section to open the GUI:

In the GUI you can view the patient CT, change the plan parameters and adjust the optimization parameters.

The usage of the GUI is explained to more detail in the [matRadGUI tutorial](How to execute matRadGUI). Here we will focus on the "manual" execution of the matRad script. To "manually" change the optimization parameters you can adjust the cst-cell (see cst-cell documentation for more information).

**3. generate steering file**

In this section the steering file (stf) is created and the matRad steering information is stored as a struct.

The Command Window should show you the progress

**4. dose calculation**

In this section the dose influence matrix (dij-struct) for the defined beam angles is calculated.

Again the progress should be shown in the Command Window.

**5. Fluence optimization**

In this section the fluence is optimized to find the bixel (photons) or spot (protons/carbon) weights minimizing the objective function.

During this process, the current objective function value is displayed:

**6. sequencing**

For photon IMRT the application of a multileaf collimator is necessary. By sequencing the applicable dose distribution can be simulated. The fourth input of matRad_engelLeafSequencing(resultGUI,stf,dij,7) is the number of stratification levels. You can adjust this number to use the number of levels you want.

When the sequencing is finished, the result-struct is updated.

**7. direct aperture optimization**

For photon therapy the multileaf collimator sequencing can be further refined by using an experimental gradient-based direct aperture optimization algorithm, where leaf settings and aperture intensities are optimized simultaneously. Further information inclduing references about the direct aperture optimization algorithm can be found directly in the source code or in the technical documentation about the fluence optimization.

**8. visualization of the resulting treatment plan**

Now you can visualize the resulting treatment plan using the GUI:

In the GUI you can see the resulting dose distribution for the calculated treatment plan. You can choose which plane and slice should be displayed. You can also display a dose profile plot by changing the plot type from intensity to profile.
If you have chosen a biological optimization, then you have several parameters you can choose to be displayed (e.g. RBE-weighted dose, biological effect, α or β values).

**9. Show DVH and quality indicators**

In this section the Dose-Volume-Histograms are calculated and visualized.

The diagram shows the DVH and in the table you will see the mean, maximum and minimum dose for every VOI. Additionally the dose and dose-volume coefficient for several confidence levels are displayed.

Step 4: Import additional patient data

matRad supports the import of patient data stored in the DICOM format. A set of functions designed for this purpose can be found in the subfolder 'dicomImport'. For more information about the usage of the import functions please check out this wiki page.