Click here to


Are you sure ?

Yes, do it No, cancel

4DCT and 4DCBCT Under-Predict Intrafraction Lung Target Motion During Radiotherapy

E Steiner1*, C Shieh1 , V Caillet1,2 , R O'Brien1 , J Booth1,2 , A Briggs2 , N Hardcastle2 , K Szymura2 , D Jayamanne2 , T Eade2 , P Keall1 , (1) The University of Sydney, The University Of Sydney, NSW, (2) Royal North Shore Hospital, St Leonards, NSW


(Wednesday, 8/1/2018) 1:45 PM - 3:45 PM

Room: Karl Dean Ballroom B1

Purpose: 4DCBCT facilitates verification of lung tumor motion ranges before each treatment fraction as the motion ranges can exceed the observed motion from planning 4DCT. This work aims to quantify whether 4DCT or 4DCBCT represent the actual motion range during treatment, using Calypso motion signals as the ground truth.

Methods: Eight patients receiving lung SABR (stereotactic ablative radiotherapy) were implanted with three Calypso beacons close to their tumor. 4DCBCT was reconstructed for 23 fractions (1-4 fx/patient) from a 1min CBCT scan. The real-time intrafraction beacon centroid (“target�) motion range during treatment (95% of the motion = 2SDs to limit the effect of extreme values) for each fraction was compared with the beacon centroid motion segmented for all 10 bins of the 4DCT and 4DCBCT sets. The correlations and linear regressions between the real-time target motion and the motion ranges in 4DCT and 4DCBCT were investigated for the LR, SI, AP and 3D directions, respectively.

Results: 4DCT under-predicts intrafraction target motion range by 68, 83, 69 and 36% in the SI, LR, AP and 3D directions, respectively. 4DCBCT under-predicts intrafraction target motion range by 50, 53, 64 and 24% in the SI, LR, AP and 3D directions, respectively. 4DCBCT showed better correlations (Pearson r SI:0.98, LR:0.77, AP:0.91, 3D:0.94) with the intrafraction target motion than 4DCT (SI:0.95, LR:0.64, AP:0.73, 3D:0.90). For LR (mean range: 3.6mm, max range: 9.1mm) and AP (mean range: 4.9mm, max range: 10.2mm) motion during treatment the correlation to the motion ranges in 4DCT and 4DCBT was low, however, the motion was also smaller in these directions than the SI motion during treatment (mean range: 9.4mm, max range: 24.1mm).

Conclusion: 4DCT and 4DCBCT failed to represent the full target motion range. SI motion during treatment was shown to be reliably predicted from 4DCBCT using a linear scaling factor.

Funding Support, Disclosures, and Conflict of Interest: Varian Medical Systems have partially funded the patient study that enabled this analysis.


Lung, Respiration, Target Localization


IM/TH- RT X-ray Imaging: 4D CBCT imaging/therapy implementation

Contact Email