Topics covered in this article include:

Introduction

4D CT is crucial for radiotherapy planning, as it captures the impact of respiratory motion on tumor and normal tissue positions. Unlike static 3D CT, 4D CT acquires images across breathing phases, visualizing and quantifying tumor and organ movement, especially in the thoracic and upper abdominal regions.

Several publications and Canadian national guidelines have emphasized the importance of performing QA on 4D CT systems to ensure image quality and motion accuracy in 4D reconstructions.

However, this QA can be time-consuming and involve the purchase of specialized equipment. A recent study (Yang Y, Lee D et al) indicated that analysis time could be shortened fourfold by using automated software that utilizes the ubiquitous Catphan Phantom from The Phantom Laboratory.

Image Owl supported this study by modifying the Catphan analysis within TotalQA and has recently further streamlined the process to automatically perform analysis across multiple acquisitions of the Catphan phantom during the QA procedure.

The study may be downloaded from the JACMP site.

All Catphan models feature a sensitometry module with multiple materials of known characteristics, enabling the analysis of image quality parameters such as HU constancy and linearity. All of the models contain a Teflon TM plug and air voids. These features are cylindrical, with known X, Y, and Z dimensions, and tight manufacturing tolerances. By mounting the Catphan on a motion platform, the reconstructed dimensions of these features can be measured in static image sets and from various phase reconstructions during the 4D CT process to quantify the degree of distortion and motion accuracy.

Measurement Summary

The analysis displays the following test results:

Motion Phantom Parameters
Spatial Integrity - Stationary Series
Image Quality Summary (Stationary Series)
Regular Reconstruction - Spatial Integrity Detail (Teflon)
Regular Reconstruction - Spatial Integrity Detail (Air)
Motion Amplitude reconstruction accuracy (Detail)
Regular Reconstruction - Spatial Integrity Summary (Teflon)
Regular Reconstruction - Spatial Integrity Summary (Air)
Catphan Results
Image Quality Summary
MIP Spatial Integrity (Teflon) w/ Motion
MinIP Spatial Integrity (Air) w/ Motion

Available Tests

There are two options for selecting tests to run on 4D image acquisitions. In the manual setup, specific tests can be added to a Catphan template. Alternatively, the automated setup uses a preset 4D CT Analysis template that is now a standard option in the software. The table below details the available tests in the automated setup.

Template Section	Subsection	Tests
4D CT Analysis	Motion Phantom Parameters	Motion Phantom Parameters
	Stationary Series	Spatial Integrity - Stationary Series Image Quality Summary (Stationary Series)
	Regular Reconstruction Phases Images	Regular Reconstruction - Spatial Integrity Detail (Teflon) Regular Reconstruction - Spatial Integrity Detail (Air) Motion Amplitude reconstruction accuracy (Detail) Regular Reconstruction - Spatial Integrity Summary (Teflon) Regular Reconstruction - Spatial Integrity Summary (Air) Catphan Results Image Quality Summary
	MIP Reconstruction Series	MIP Spatial Integrity (Teflon) w/ Motion
	MinIP Spatial Integrity (Air) w/ Motion	MinIP Spatial Integrity (Air) w/ Motion

Creating a 4D CT Analysis Schedule

To automate the 4D CT image analysis, a dedicated template (4D CT Analysis) is provided to allow automated comparisons between reconstructions of static acquisitions, phased or binned reconstructions, and MIP or MinIP reconstructions.

To set up a 4D CT Analysis schedule, click on Manage > Schedule Configurations and then + Add QA Schedule on the machine that you wish to add the schedule to.

In the Add QA Schedule Page, select the 4D CT Analysis template.

Set up the rest of the schedule settings and save the schedule. For help, see Managing Schedules.

Extracting Phase or Bin Information

There are no specific phase or bin tags in the DICOM headers for the CT images, so they must be extracted from other tags. This is customizable in the software. Click on Manage and navigate to the Configurations settings page, and then click to edit the schedule.

By opening the Image Processing settings, the default settings, shown below, are displayed.

The DICOM Tag Group (default: 0008) and Element (default: 103E) specify the tag in which to find the phase string. The default is the Series Description tag.

The Phase Regular Expression (default: (\d{1,2})%) is a regular expression designed to pull a 1 or 2-digit number followed by a % sign (e.g., 0%, 80%) from the input string. The first match found will be used.

New QA Instructions

Complete the Motion Phantom Parameters Section

Record the parameters used for the motion platform being used for the analysis.

Motion Platform Type: Record the motion platform model.
Surrogate Amplitude: Record the motion amplitude of the motion surrogate on the platform.
Period: Number of seconds for one cycle of the motion platform.
Peak-to-peak S/I Amplitude: The range of Superior/Inferior motion in one cycle of the motion platform.
A/P Amplitude: The range of Anterior/Posterior motion in one cycle of the motion platform.
L/R Amplitude: The range of Left/Right motion in one cycle of the motion platform.

Upload the Stationary Series

Upload the stationary Catphan series. This series will serve as a baseline for subsequent series to be compared against.

Only one stationary series is permitted. If a second stationary series is uploaded, the first one must be removed before the second one can be processed.

Upload the Motion Series

Upload of multiple phased (or binned) sets of images reconstructed at specific points in the motion cycle. The spatial integrity measurements are made for each series and compared against the expected values and the statistics for the stationary series above. The Z-position of the Teflon plug and air void are also reported for each series.

Multiple series may be uploaded simultaneously or sequentially into the upload widget at the top of this section. Series are separated based on the series UID in the DICOM headers, which must be unique.

Uploading of a single MIP and a single MinIP reconstructions are recommended.

Detailed Discussion of Measurements

Spatial Integrity - Stationary Series

This table displays the measured dimensions of the Teflon plug and air void against the expected manufacturing dimensions.

Direction: Direction of the dimension (LAT, AP, SI)
Exp. Width: The specified manufacturing dimensions of the Catphan Teflon plug and air void in mm.
Measured Width (Teflon): The measured dimensions of the Teflon plug from the image set in mm.
Measured Width (Air): The measured dimensions of the air void from the image set in mm.
Diff from Exp. Teflon: Measured-expected difference in mm for the Teflon plug in mm.
Diff from Exp. Air: Measured-expected difference in mm for the Air void in mm.

Image Quality Summary (Stationary Series)

This table displays the summary image quality statistics for the stationary series.

Spatial resolution: Frequency at 10% from MTF in lp/mm
Contrast (detail at 1% contrast): Smallest detected diameter at 1% contrast.
HU Constancy: Max absolute difference from expected CT numbers for air (-1000 HU), LDPE (-100 HU), and acrylic (120 HU).
Uniformity: Max absolute difference from the center ROI of peripheral ROIs.
Noise: Noise in the center region (with a diameter of 40% of the diameter of the module) computed as the standard deviation of measured HU values.

Regular Reconstruction - Spatial Integrity Detail (Teflon)

This test measures the LAT(X), AP(Y), and SI(Z) dimensions of the Teflon plug for each phase and compares those dimensions to the expected values. The maximum absolute deviation from the expected values across all phases is also reported.

Phase: The phase number extracted from the DICOM tags (see previous section).
LAT(X): Width in mm of a horizontal profile through the Teflon plug axial center at the z center of the phantom’s sensitometry module.
AP (Y): Width in mm of a vertical profile through the Teflon plug axial center at the z center of the phantom’s sensitometry module.
SI (Z): Width in mm of a profile through the z slices containing the Teflon plug in the phantom’s sensitometry module. The profile is calculated as the mean of all voxels within 5mm of the plug’s xy center on each slice.
LAT (X) Difference from Expected: LAT (X) measured - LAT (X) expected in mm.
AP (Y) Difference from Expected: AP (Y) measured - AP (Y) expected in mm.
SI (Z) Difference from Expected: SI (Z) measured - SI (Z) expected in mm.
Abs. Max LAT (X) Difference: Absolute maximum of LAT (X) Difference from Expected for all phases.
Abs. Max AP (Y) Difference: Absolute maximum of AP (Y) Difference from Expected for all phases.
Abs. Max SI (Z) Difference: Absolute maximum of SI (Z) Difference from Expected for all phases.

Regular Reconstruction - Spatial Integrity Summary (Teflon)

This table provides a summary of the individual phase measurements for the Teflon plug against the expected dimensions and the stationary values.

Direction: Direction of the dimension (LAT, AP, SI)
Exp. Width: The specified manufacturing dimensions of the Catphan Teflon plug in mm.
Stationary Width: The measured dimensions of the Catphan Teflon plug in the stationary image set in mm.
Mean: Mean dimension of all the phased image sets in mm.
Median: Median dimension of all the phased image sets in mm.
SD: The sample standard deviation of the dimensions of the phased image sets in mm.
Diff from Expected (Recon - exp): The mean dimension - the expected dimension in mm.
Difference from Stationary (Recon - stationary): The mean dimension - the measured stationary dimension in mm.

Regular Reconstruction - Spatial Integrity Detail (Air)

This test measures the LAT(X), AP(Y), and SI(Z) dimensions of the air void for each phase and compares those dimensions to the expected values. The maximum absolute deviation from the expected values across all phases is also reported.

Phase: The phase number extracted from the DICOM tags (see previous section).
LAT(X): Width in mm of a horizontal profile through the air void axial center at the z center of the phantom’s sensitometry module.
AP (Y): Width in mm of a vertical profile through the air void axial center at the z center of the phantom’s sensitometry module.
SI (Z): Width in mm of a profile through the z slices containing the air void in the phantom’s sensitometry module. The profile is calculated as the mean of all voxels within 5mm of the void’s xy center on each slice.
LAT (X) Difference from Expected: LAT (X) measured - LAT (X) expected in mm.
AP (Y) Difference from Expected: AP (Y) measured - AP (Y) expected in mm.
SI (Z) Difference from Expected: SI (Z) measured - SI (Z) expected in mm.
Abs. Max LAT (X) Difference: Absolute maximum of LAT (X) Difference from Expected for all phases.
Abs. Max AP (Y) Difference: Absolute maximum of AP (Y) Difference from Expected for all phases.
Abs. Max SI (Z) Difference: Absolute maximum of SI (Z) Difference from Expected for all phases.

Regular Reconstruction - Spatial Integrity Summary (Air)

This table provides a summary of the individual phase measurements for the air void against the expected dimensions and the stationary values.

Direction: Direction of the dimension (LAT, AP, SI)
Exp. Width: The specified manufacturing dimensions of the Catphan air void in mm.
Stationary Width: The measured dimensions of the air void in the stationary image set in mm.
Mean: Mean dimension of all the phased image sets in mm.
Median: Median dimension of all the phased image sets in mm.
SD: The sample standard deviation of the dimensions of the phased image sets in mm.
Diff from Expected (Recon - exp): The mean dimension - the expected dimension in mm.
Difference from Stationary (Recon - stationary): The mean dimension - the measured stationary dimension in mm.

Catphan Results

This table displays the summary image quality statistics for each phased image series.

Phase: The phase number extracted from the DICOM tags (see previous section).

Spatial resolution: Frequency at 10% from MTF in lp/mm
Contrast (detail at 1% contrast): Smallest detected diameter at 1% contrast.
HU Constancy: Max absolute difference from expected CT numbers for air (-1000 HU), LDPE (-100 HU), and acrylic (120 HU).
Uniformity: Max absolute difference from the center ROI of peripheral ROIs.
Noise: Noise in the center region (with a diameter of 40% of the diameter of the module) computed as the standard deviation of measured HU values.

Image Quality Summary

This table compares the mean of the image quality results for each phase against the stationary measurements for the same tests.

Test: Description of the test
Unit: Test unit
Stationary: The value for the test from the stationary image series.
4D CT Mean: The mean value of the test for all the phased image series.
Difference (4D CT - Stationary): Difference between the mean of the phased series and the stationary series.

MIP Spatial Integrity (Teflon) w/ Motion

This test compares a single image series reconstructed as a MIP-derived image against the stationary series. Only one MIP series is permitted. If a second MIP series is uploaded, the first one must be removed before the second one can be processed.

The MIP image will distort the Teflon plug image so that its expected dimensions will be approximately the expected manufacturing dimensions plus the motion amplitude. This MIP image of the plug is compared against the expected dimensions.

Direction: Direction of the dimension (LAT, AP, SI).
Exp. Width-Teflon: The specified manufacturing dimensions of the Catphan Teflon plug in mm.
Motion Amplitude: The motion amplitude for each direction in mm. These are pulled from the Motion Platform Parameters above.
Measured Width (Teflon Stationary): The measured dimensions of the Catphan Teflon plug in the stationary image set in mm.
Measured Width (Teflon MIP): The measured dimensions of the Catphan Teflon plug in the MIP image set in mm.
Exp. Width + Amplitude: Sum of the expected width and amplitude in mm.
Stationary Width + Amplitude: Sum of the measured stationary width and amplitude in mm.
Measured Width (MIP) - (Exp.+Amplitude): Difference between the measured MIP image dimension and the expected dimension of the MIP Teflon plug in mm.
Measured Width (MIP) - (Stationary+Amplitude): Difference between the measured MIP image dimension and the stationary + amplitude dimension of the MIP Teflon plug in mm.

MinIP Spatial Integrity (Air) w/ Motion

This test compares a single image series reconstructed as a MinIP-derived image against the stationary series. Only one MinIP series is permitted. If a second MinIP series is uploaded, the first one must be removed before the second one can be processed.

The MinIP image will distort the air void image so that its expected dimensions will be approximately the expected manufacturing dimensions plus the motion amplitude. This MinIP image of the void is compared against the expected dimensions.

Direction: Direction of the dimension (LAT, AP, SI).
Exp. Width Air: The specified manufacturing dimensions of the Catphan air void in mm.
Motion Amplitude: The motion amplitude for each direction in mm. These are pulled from the Motion Platform Parameters above.
Measured Width (Air Stationary): The measured dimensions of the Catphan air void in the stationary image set in mm.
Measured Width (Air MinIP): The measured dimensions of the Catphan air void in the MinIP image set in mm.
Exp. Width + Amplitude: Sum of the expected width and amplitude in mm.
Stationary Width + Amplitude: Sum of the measured stationary width and amplitude in mm.
Measured Width (MinIP) - (Exp.+Amplitude): Difference between the measured MinIP image dimension and the expected dimension of the MinIP air void in mm.
Measured Width (MinIP) - (Stationary+Amplitude): Difference between the measured MinIP image dimension and the stationary + amplitude dimension of the MIP air void in mm.