This program is provided to precisely position a medical accelerator’s isocenter relative to a patient prior to radiation therapy treatment. This is accomplished by implanting one or more markers inside the patient, locating those markers relative to an image set and the isocenter of treatment beams, and tracing the images of those markers on radiographs taken prior to each treatment session. Because the markers may be long and flexible, they are located in the image set with this software by tracing their visible path thought the image set, typically a set of stackable CT scans. The markers are then retraced on two or more radiographs taken at the time of treatment. The location of isocenter within the image set for the radiographs is found, and the couch coordinates for each treatment beam is then reported given the couch coordinates for the radiographs and the known isocenter of the beams relative to the image set.

The user of this software must test this patient location system before clinical use. A test procedure is provided by the distributor of this software.

This software is built upon System 2100 from Math Resolutions, LLC, a radiological image display system, and uses the plan and beam functions from Dosimetry Check, also from Math Resolutions, LLC. Reference will be made to specific sections in the manuals of those two integrated products. This program includes all the functions of System 2100 and can optionally include all the functions of Dosimetry Check if turned on with the license key. Otherwise only those Plan and Beam functions necessary for patient positioning will be available from Dosimetry Check. On line manuals or downloadable PDF files are available at www.MathResolutions.com.

This program is available for Microsoft Windows or Linux personal computer systems, or for Silicon Graphics Irix computer systems.

Running the Program

Run program MarkRTTasks, shown below, as a convenience to select and run the programs provided for the multiple functions required. The program interface is shown below. You can simply leave this small window up.

The “Other Utilities” pull down:

Making a selection here runs a separate utility. As a convenience you can pre-select the patient here. But for daily operation of MarkRT while treating patients, make the patient selection from MartRT instead as the program will then directly push the locate patient toolbar. The programs selectable here are described below. For reading in the plan from the planning system or otherwise to enter markers and define an isocenter, you would start with “Read in a Patient Case and Plan”.

Other utilities provides a program to create a CT scan of a water density square, a utility to view the information in a dicom file, and to get a command prompt window.

Program finds small adjustments to the couch position.

The patient should be immobilized during this entire procedure.

You should first closely align the patient with isocenter using the existing tools to accomplish this, such as the laser alignment system on the treatment machine. Here we will solve for small adjustments as indicated by the markers. The final couch movement found here should therefore be small. If a large couch movement is called for, then the patient position must be reviewed. Beam films can be taken to confirm the position of isocenter to compare to the plan. Setting the isocenter to the beam in the “Trace and Solve” popup below should result in exact alignment of the marker images and their projected images with the new images.

Patient Orientation

Program normally assumes that patients are treated head first toward the accelerator gantry unless otherwise indicated. Care must be taken for a feet first orientation. The same precaution should be taken for a prone patient.

You can change the orientation of the patient from the plan toolbar (see the Dosimetry Check manual). The ReadDicomCheck program for the Dicom RT plan download should mark a patient feet first (FF) for each if the Dicom plan file indicates the patient orientation as being feet first. Head first or feet first is a beam attribute, not a plan attribute. Here we will require all beams to have the same orientation.

You cannot change the orientation between supine and prone as you would normally scan the patient supine or prone. Prone patients will be shown prone.

The RTOG protocol is another matter. The protocol is somewhat ambiguous and contradictory. It is stated that the Z axis always points toward the patient’s feet, but the coordinate system is right handed. Then that the X and Y axes refer to the treatment machine and not to the patient, yet scans are to be viewed from the patient’s feet, and that the Z axis depends upon scan order but is to point toward the patient’s feet. Sample RTOG files we have received for a feet first plan were not encoded correctly.

For the rare situation of feet first or prone, you should first test in a phantom and determine a protocol.

Accelerator Couch Coordinate System

The IEC coordinate system is used for the plan tabletop coordinate system of this software. With the gantry pointed toward the floor, the coordinate system is for the X axis to go to your right as you are standing in front of the couch looking into the gantry. The Y axis will be parallel to the axis of rotation of the gantry and points toward the gantry. The Z axis points from isocenter up toward the ceiling.

The direction of the positive couch X, Y, and Z axis in the treatment machine’s coordinate system will be specified in the Geometry file relative to the plan tabletop coordinate system and in the CouchCoord file. X is always associated with the lateral, Y with longitudinal, and Z with height. The Geometry file must specify the directions of the actual coordinate system of the couch relative to the plan tabletop system. The CouchCoord file must also specify the correct directions.

The Geometry file only supports coordinate systems for the couch in which differences in a coordinate will represent differences in distance, and a coordinate defines a unique location of the couch. This coordinate system is used for all couch movement controls for moving isocenter within a plan.

The CouchCoord file is used to provide support for non-coordinate couch systems and is be used to convert delta movements and initial couch coordinates to actual read out numbers. This file is NOT needed if your couch coordinates IS a coordinate system.

In a coordinate system, a point 1 cm negative to the origin will be represented by –1.0 and a point 1 cm to the positive side of the origin will be represented by +1.0. The Geometry file described below will govern the conversion of plan tabletop coordinates to actual couch coordinates for movement of isocenter. The direction of the machine’s couch coordinate system relative to the plan tabletop coordinate system is specified in the file. A constant may be added to the plan tabletop coordinate system so that, for example, the origin might display the number 100.0 if the constant value of 100.0 is specified.

The direction vectors from the Geometry file are used if there is no CouchCoord file in computing the couch coordinates of each beam given the couch coordinates at which the radiographs were taken and the movement needed. Otherwise the CouchCoord file parameters are used for displaying the final couch coordinates.

We have noted unusual systems. Such as on the Elekta machines, the couch lateral will read 1.0 if 1 cm to the right of the origin (to your right while looking into the gantry) and will display 99.0 if 1 cm to the left of the origin. This is not a coordinate system by our definition since the distance from 99.0 to 1.0 is 2.0 cm, not 98.0 cm. Further, in theory, 50 cm could be either 50 cm to the right or to the left. Confusion is only avoided by knowing that the couch will not move that far. If the lateral couch coordinate was reported to be 0.2 cm and the couch must be moved -0.4 cm (to the left in the negative direction), MarkRT (VGRT) will report the lateral couch coordinate for the beam as at –0.2 cm if there is no CouchCoord file. In the room the couch will read 99.8 cm at this location. Since the operator must be familiar with the machine’s readout and has to make such translations under normal use, there should not be any additional confusion here. The same idea applies for couch longitudinal and height movements.

If the CouchCoord file is present, the parameters in it will be used to convert to the new coordinate. In the example above, the coordinate will now change from 0.2 cm to 99.8 cm for a delta of –0.4 cm.

Definition of the Markers and Isocenter

The position of the markers and isocenter of a beam must be known in the same coordinate system. There are two ways to accomplish this. One is to download the treatment plan and CT scans to this program. The markers are then traced through the CT scan set to define them. The isocenter of beams is known from the plan download.

A second method would be to locate specific points using what ever tools are available on the planning system and to reenter those points. The points must include the markers and at least one beam isocenter. This latter method might be more convenient for point markers as opposed to linear markers. Immediately below is the downloading of the treatment plan method.

Reading in of the treatment plan

Because the location of the treatment plan’s isocenter must be known relative to the stackable image set, means is provided to download this information from the radiation therapy treatment planning system, or to otherwise locate isocenter within the image set. Both the RTOG and Dicom RT protocols are supported for reading in the image set and beam geometry.

RTOG, Dicom RT

Two separate utilities are provided to read plans and images from the radiation therapy treatment planning system. See the “RTOG Download” and “Dicom RT Download” sections in the Dosimetry Check manual for details. These two utilities will read the plan and create a patient entry in the patient directory of this system, and store the image set as a stacked image set under the patient entry, and store the plan under the patient entry. See the note above in regards to patient orientation, particularly in regards to the RTOG protocol.

A stacked image set is a group of successive slices that together models the patient anatomy in three dimensions. A patient entry may have more than one stacked image set associated with it, and may have more than one plan. Each plan is linked to a stacked image set and may have more than one treatment beam. After running either utility to create the patient case, run program MarkRT (VGRT) and simply select the patient and plan that was created. For the RTOG or Dicom RT download, the user should select the image set and beam geometry to be downloaded from the planning system. Region of interest (ROI) outlines can be read and optionally displayed by MarkRT (VGRT) in 2D and 3D views, which may prove useful.

Recreation of Plan

If for some reason the plan is not available in RTOG or Dicom RT format, the plan and beam positions can be recreated with MarkRT (VGRT). However, the images used for the plan must be available in Dicom format. Images, such as CT images taken consecutively through the patient are here referred to as a stackable image set, meaning that the images are geometrically related and can be used to define a 3 dimensional patient volume. In MarkRT (VGRT) create a new patient and create a new stacked image set. Read in the images for the stacked image set. Refer to the System 2100 manual sections on Patient and Stacked Image Sets to perform these functions.

In order to position a beam, a skin outline must exist. Go under Contouring and select the auto-body outlining tool. See the System2100 manual.

Next, create a plan and then a beam for that plan. Position the isocenter of that beam in the correct location within the stacked image set (see “Beam” in the Dosimetry Check manual). If there is more than one isocenter in the plan, you may optionally create a beam for each isocenter. The only parameter that matters here is the isocenter of the beam. Other aspects of beam geometry are not used here. You may use the default accelerator provided when creating a beam. It is important that you review the Geometry and CouchCoord files for that accelerator (see below).

Definition of the Markers from the CT Scan Set

Because the markers may be more than points, they are traced here with MarkRT (VGRT) in the stacked image set. Upon running MarkRT (VGRT) select the patient and then select the stacked 1Gimage set.

The markers must be visible in the stacked image set in order for you to trace them. You must assign a name to each maker, if there is more than one, and trace each one with the mouse through the stacked image set. This function is covered in the “Points, Labels, and Markers” section in the System 2100 manual. This tracing will define the marker in the coordinate system of the stacked image set. The markers are then drawn in 2D plane images and in 3D perspective room views. The marker facility is found under the Stacked Image Sets pull down menu on the main toolbar (shown below), to Options, to Markers.

Typing in Marker and Isocenter Coordinates

Coordinates defining the markers and isocenter can be entered directly if they are known in the same coordinate system (IEC or DICOM) relative to the patient. They might be found by locating the markers on the CT scans using the mouse on the planning system, along with the isocenter of a beam. These points would have to be written down and then reentered here. This program’s coordinate system is IEC. The +Y axis points towards the patient’s head, Z axis is up superior, X is the patient’s right to left. In DICOM the +Z axis points towards the patient’s head, X is the same, and +Y axis is down inferior. The program will provide the option of typing in the coordinates in DICOM if your planning system is DICOM. Otherwise you can enter directly in IEC coordinates. You are going to have to know and review your planning system’s coordinate system to know how to enter the coordinates.

Default Coordinate System for Data Entry

There is a file called VGRTTypeIn in the program resources directory that will set the default system used to enter the points here. An example file is shown below:

/* file format version */ 1

// file to specify default coordinate system of the planning system.

// 1 is IEC, -1 is Dicom

-1

Note that there is only one parameter to set, IEC or DICOM. This will preset the choice in the marker and beam isocenter entry popups shown below. Points entered in DICOM will be converted to IEC.

Program MarkerTypeIn

The type in facility is a separate program called MarkerTypeIn. Run this program to enter you data. The main toolbar for this program is shown below:

You will first have to select or create a patient entry. Then you will have to create or select a plan. The current plan name will be shown in the text field on the tool bar. Then you can enter markers and beam isocenters. You can also delete a marker or a beam, or delete an entire plan. When done use the “Show Plan Entries” under the Plan pull down menu to review your entries for correctness.

For each plan you create, a stacked image set will be made with no images. A simple model of the patient will be created to distinguish the superior and inferior directions on the drawing of the accelerator when a solution is found.

Type in a marker

For a new marker you will get the marker type in popup tool shown here:

You must specify a unique name for the marker. The line width selection controls the width of the line in 2D renderings and 3D wire frame renderings. The “Solid Fill/Wire” toggle button controls whether the marker is solid filled on 2D renderings or just a wire outline. The marker diameter controls the diameter of the rendered marker in 2D and 3D renderings. The color controls the color of the marker. See the section “Points, Labels, and Markers” in the System2100 manual for more details.

You must pick whether you are entering the points in DICOM or IEC with the radio box. Then type in each coordinate in centimeters in the provide field, one for X, Y, and Z. Then hit the “Add Point to Sequence List” button. If this is a point marker, you need only enter one point. If a linear marker, then you must enter a sequence of points in order that will define the center line of the marker. When you are done hit the ”Save Marker” button. Beforehand you can delete a specific point.

Type in a beam isocenter

Enter a beam isocenter by selecting “Add Beam” from the beam pull down menu. You will get the below TypeInBeam popup:

You must enter a unique name for each beam if more than one. Then select the accelerator. This selection specifies the directory the program will read for geometric information about the treatment machine. See the below sections on the geometry and couch coordinate files. All beams must be for the same treatment machine and this is strictly enforced in program MarkRT (VGRT).

The proper selection of DICOM or IEC must be made here. A change in selection by the above marker type in tool will cause the same selection here and vice versa, along with a warning message. Type in the coordinates and then hit the “Save Beam” button. Under the beam pull down menu you can delete a prior beam entry. These beams will all be stored with nominal angles which are of no interest or consequence here.

Show Plan Entries

When finished entering the markers and beams, select “Show Plan Entries” on the plan pull down menu for review.

The coordinates for all the markers and beams isocenters will be shown for review for correctness in both IEC and DICOM coordinates, and can be printed.

Locate the Patient

Shown below is the main toolbar of the System 2100 library functions that is the top toolbar for MarkRT (VGRT):

If there is only one plan the program will automatically sequence through the below tool bars after selecting the patient to arrive at the Locate Patient tool bar shown below. If you need to first trace markers on the CT scans, then hit the Return buttons to back up to the main tool bar.

The path you must take will initially be shown in green. You must first select the patient under the Patient pull down. The program will automatically select the next toolbar if there is more than one plan for this patient. Otherwise under the Applications pull down on the main toolbar, select MarkRT (formally VGRT). The first time this tool bar is selected, the Retrieve Plan function will automatically be implemented. If there is only a single plan, it will be selected. Otherwise you will have to select the plan. Otherwise select the plan on the plans toolbar. The plan will be retrieved and the plan toolbar pushed. The stacked image set is also read but not displayed since it is generally not needed and so why wait those few seconds for those images to come up.