Unencrypted login | home

Program Information

Respiratory Signal Triggered 4D Cone-Beam Computed Tomography On a Linear Accelerator

no image available
B Cooper

B Cooper1,2*, R O'Brien1, P Keall1, (1) The University of Sydney, Sydney, NSW, (2) ACT Health, Canberra, ACT

SU-C-213CD-5 Sunday 1:30:00 PM - 2:15:00 PM Room: 213CD

Purpose: To compare imaging dose and image quality for a commercially available 4D CBCT system and a proposed respiratory signal triggered 4D CBCT system which uses fewer projections and thus can reduce imaging dose burden.

Methods: A large original set of full-fan kilovoltage projection images of a stationary thorax phantom was acquired through 200 degrees of gantry rotation. This dataset was reconstructed to produce a 'ground truth' CBCT dataset (GT). Using acquisition parameters from a commercially available 4D CBCT system, the original projections were resampled to produce a 4D CBCT dataset (A). A sine wave respiratory signal, with period 4 seconds, is assumed and used to split projections into phase bins. Respiratory signal triggered 4D CBCT uses the same acquisition parameters, but the respiratory signal is used to trigger projection acquisition so that only one projection per phase bin is acquired (dataset B). For the commercial 4D CBCT system (A) and the respiratory signal triggered 4D CBCT system (B), image quality was quantified using the root mean squared deviation (RMSD) of the pixel values from the ground truth (GT) dataset. Imaging dose was assessed assuming one projection equals one dose unit.

Results: The imaging dose for 4D CBCT system (A) was 118 dose units compared to the proposed 4D CBCT system (B) 58 dose units. The RMSD was calculated as 105.4, std. dev. 5.4 and 107.5, std. dev. 4.0 for the 4D CBCT system (A) and the proposed 4D CBCT system (B) respectively.

Conclusions: The respiratory signal triggered 4D CBCT system exhibits a significantly reduced dose compared to the commercial 4D CBCT system. This comes at a small reduction in image quality. Future work will be to investigate the performance of the proposed 4D CBCT system with real respiratory signals.

Contact Email