Room: Track 4
Purpose: To investigate the performance of a novel, commercially available sequential 4DCT (‘seq4D’) scan mode and compare motion artifacts in reconstructed 4D images against a conventional, helical 4DCT (‘hel4D’) scan mode and reconstruction algorithm.
Methods: The seq4D technique consists of two main steps: 1. Initial learning phase of the breathing pattern using a respiratory correlate signal to set optimal baseline scan parameters. 2. Sequential scanning with automatic adaption of scan parameters to the current respiratory pattern. The test object was a dynamic thorax phantom with a 3 cm simulated lesion that was moved in a sinusoidal pattern along the caudocranial direction with a 2 cm magnitude and 12 bpm frequency. Three scans with different motion patterns were acquired for each scan type: a) a regular sinusoidal pattern, b) a sinusoidal pattern with a 6 s pause in the middle of the scan, c) a sinusoidal pattern with the frequency doubled for 6 seconds in the middle of the scan. The size of the lesion and the magnitude of the motion in the reconstructed 4D images were measured and compared to the true values.
Results: Both the seq4D and hel4D method reproduced the size and motion of the lesion accurately for the regular breathing pattern a). For pattern b) the differences to the true values were <1mm (motion) and 1 mm (size) for seq4D compared to 1.2 mm (motion) and 7 mm (size) for hel4D. For pattern c) the differences were <1mm (motion) and 2.5 mm (size) for seq4D compared to <1mm (motion) and 18 mm (size) for hel4D.
Conclusion: In scans with irregular breathing patterns, the seq4D method resulted in significantly smaller motion artifacts compared to hel4D that were most apparent in the reconstructed size of the lesion.