Purpose: Cardiac motion is not currently accounted for during the treatment of central lung and mediastinal tumors. The aim of this study is to adapt the gantry rotation speed and projection rate in accordance with patient cardiac and respiratory rates to eliminate both cardiac- and respiratory-induced motion blurring during image acquisition. Here, we simulate this adaption on a standard linear accelerator.
Methods: The 4D-XCAT phantom was used to simulate the acquisition of dual cardiac- and respiratory-gated CBCT thoracic images from 17 patient-measured ECG and respiratory traces. Our adaptive protocol acquired evenly angular spaced projections during 60-80% of the cardiac cycle and during 20% displacement at peak exhale of the respiratory cycle. A respiratory-only gated CBCT acquisition with constant gantry speed and projection rate was used as a comparator. Image metrics of edge-response-width (ERW) and contrast-to-noise-ratio (CNR) were used to compare and characterise the image sharpness and contrast. The effect of the total number of projections acquired and magnitude of cardiac motion (0.5 cm or 1.0 cm) on scan time and image quality were also investigated.
Results: Our adaptive protocol provided a decrease between 8-64% and 22-35% in the median ERW (indicating improved image sharpness) over conventional acquisition in both the AP and SI directions respectively. Our adaptive protocol also enabled more consistent, albeit lower median CNR values over conventional acquisition. Notably, the magnitude of cardiac motion had no observable effect on the median CNR. The median total scan times across the 17 patients with our adaptive protocol ranged from 117s (40 projections) to 296s (100 projections), compared with 240s for the conventional protocol (1320 projections).
Conclusion: Dual cardiac- and respiratory-gated thoracic imaging is feasible on a standard linear accelerator. Compared to conventional acquisition, our protocol provides increased flexibility by enabling a trade-off between image quality and total scan time.
Funding Support, Disclosures, and Conflict of Interest: This research was supported by grant #1123068 which was awarded through the Priority-driven Collaborative Cancer Research Scheme and funded by Cancer Australia. Ricky O Brien acknowledges the support of a Cancer Institute NSW Career Development fellowship. Paul Keall acknowledges the support of an NHMRC Senior Principal Research Fellowship.