Click here to


Are you sure ?

Yes, do it No, cancel

Low Dose KV-MV Cone Beam CT with a Multi-Layer MV Imager and Edge-Preserving Sinogram Denoising

M Jacobson1*, M Lehmann2, P Huber2, A Wang3, M Myronakis1, M Shi4, D Ferguson1, I Lozano1, Y Hu1, P Baturin3, T Harris1, R Fueglistaller2, J Star-Lack3, C Williams1, D Morf2, R Berbeco1 (1) Brigham and Women's Hospital, Dana-Farber Cancer Institute, and Harvard Medical School, Boston, MA (2) Varian Medical Systems, Baden-Dattwil, Switzerland (3) Varian Medical Systems, Palo Alto, CA (4) University of Massachusetts, Lowell, Lowell MA


(Monday, 7/15/2019) 7:30 AM - 9:30 AM

Room: 221AB

Purpose: Simultaneously acquiring set-up CBCT data using both the kV and MV imagers of an IGRT system reduces scan time, but increases noise in the 3D reconstruction. A kV-MV technique with noise, dose, and spatial resolution comparable to standard kV-only CBCT scans is pursued.

Methods: A scan technique combining two noise-reducing measures is proposed: (1) a prototype high DQE four-layer MV imager. (2) a novel beam hardening correction algorithm with edge-preserving sinogram denoising. Performance was tested in thorax phantom acquisitions taken with the Varian TrueBeam. A 200° kV-MV scan arc was synthesized from sub-arcs of projections collected in separate kV-imager (110°, 125 kVp, 400 mAs) and MV-imager (90°, 2.5 MV FFF, 0.45 MU) acquisitions. The overall kV-MV scan was CTDI-equivalent with a conventional TrueBeam Spotlight mode scan (12.3 mGy). Noise was quantified in terms of the noise percentage (voxel standard deviation as a percentage of mean) in a 3D soft tissue region of interest (ROI). Spatial resolution was quantified in terms of equivalent Gaussian edge blur in an ROI straddling the lung wall. Noise versus resolution performance was compared to a CTDI-equivalent kV-only scan and to a kV-MV reconstruction technique previously published by Yin et al. (2005 Med. Phys. (32) 9).

Results: The proposed technique demonstrated lower noise as a function of spatial resolution than the competing methods, notably half the noise of Yin’s method at 0.3 mm edge blur. The proposed method also exhibited fainter non-uniformity artifacts than Yin’s method and superior bone contrast.

Conclusion: The combination of a multi-layer MV energy and a denoised beam hardening correction algorithm enables noise, resolution, and dose performance comparable to standard kV-imager only set-up CBCT, but with nearly half the gantry rotation time. This is a strong step toward easing breath-hold requirements for lung cancer radiation oncology patients.

Funding Support, Disclosures, and Conflict of Interest: NIH/NCI R01CA188446


Image-guided Therapy, Reconstruction, Cone-beam CT


IM- Cone Beam CT: Development (New Technology and Techniques)

Contact Email