A comprehensive quality assurance procedure for 4D CT commissioning and periodic QA

Abstract

Purpose: The 4D computed tomography (CT) simulation is an essential procedure for tumors exhibiting breathing-induced motion. However, to date there are no established guidelines to assess the characteristics of existing systems and to describe meaningful performance. We propose a commissioning quality assurance (QA) protocol consisting of measurements and acquisitions that assess the mechanical and computational operation for 4D CT with both phase and amplitude-based reconstructions, for regular and irregular respiratory patterns.

Methods: The 4D CT scans of a QUASAR motion phantom were acquired for both regular and irregular breathing patterns. The hardware consisted of the Canon Aquilion Exceed LB CT scanner used in conjunction with the Anzai laser motion monitoring system. The nominal machine performance and reconstruction were demonstrated with measurements using regular breathing patterns. For irregular breathing patterns the performance was quantified through the analysis of the target motion in the superior and inferior directions, and the volume of the internal target volume (ITV). Acquisitions were performed using multiple pitches and the reconstructions were performed using both phase and amplitude-based binning.

Results: The target was accurately captured during regular breathing. For the irregular breathing, the measured ITV exceeded the nominal ITV parameters in all scenarios, but all deviations were less than the reconstructed slice thickness. The mismatch between the nominal pitch and the actual breathing rate did not affect markedly the size of the ITV. Phase and normalized amplitude binning performed similarly.

Conclusions: We demonstrated a framework for measuring and quantifying the initial performance of 4D CT simulation scans that can also be applied during periodic QAs. The regular breathing provided confidence that the hardware and the software between the systems performs adequately. The irregular breathing data suggest that the system may be expected to capture in excess the target motion and geometry, but the deviation is expected to be within the slice thickness.

Keywords: 4D CT; CT simulation; quality assurance; radiotherapy; respiratory motion-management.

FIGURE 1

Example of measurements for shape preservation (left) and ITV elongation (right). The shape preservation is the measurement of the inferior and superior border between the inhale (50%) and exhale (0%) breathing cycle. The ITV elongation is the total extent of the target during the full breathing cycle. The margin for the ITV is shown with the upper and lower volume tolerance as a function of the reconstructed slice thickness

FIGURE 2

(a) The average motion amplitude of the target was calculated from the breathing pattern recorded ±2 cm above the stationary sphere. (b) The acquired Anzai breathing trace with the different scan acquisition periods along the breathing trace demonstrated for various pitch settings

FIGURE 3

The irregular breathing trace between the QUASAR motion phantom software and the Anzai laser gating system overlaid

FIGURE 4

Target volume of the regular breathing for (a) phase and (b) n‐amplitude based binning for various pitch settings. (a) Target volume for phase binning for each pitch setting and a pitch setting reconstructed at 1 mm slice thickness (dark grey). The tolerances based on slice thickness are shown, with an additional set of tolerances (dashed lines) for the 1 mm reconstruction. (b) Target volume for each amplitude binning for each pitch setting. Half slice thickness‐based tolerances are shown

FIGURE 5

ITVs of the irregular breathing pattern for both phase (a) and n‐amplitude (b) based binned for various pitch settings. The variable theoretical ITV and the resulting tolerances are shown for each pitch setting

FIGURE 6

Shape preservation of the regular breathing for (a) phase and (b) n‐amplitude based binning for various pitch settings. Edge distance between breathing phases for 1 mm slice thickness reconstruction is shown (dark grey). Half and full slice (3 mm) thickness‐based tolerances are shown. ITV elongation of the regular breathing for phase (c) and amplitude (d) based binning for various pitch settings. The variable theoretical ITV elongation and the resulting tolerances are shown for each pitch setting

FIGURE 7

ITV elongation of the irregular breathing for (a) phase and (b) n‐amplitude based binning for various pitch settings. The variable theoretical ITV elongation and the resulting tolerances are shown for each pitch setting

FIGURE 8

Center of mass (COM) displacement of the regular breathing for (a) phase and (b) n‐amplitude based binning for various pitch settings. COM distance between breathing phases for 1 mm slice thickness reconstruction is shown (dark grey). Half and full slice (3 mm) thickness‐based tolerances are shown

FIGURE 9

Center of mass (COM) displacement of the irregular breathing for (a) phase and (b) n‐amplitude based binning for various pitch settings. The average motion amplitude for each pitch setting acquisition and the resulting tolerances are shown