N Bennett¹*, H Zhang², K Mueller³, R Fahrig⁴, A Maier⁵, M Levenston¹, G Gold¹, A Wang¹, (1) Stanford University, Stanford, CA, (2) Memorial Sloan Kettering Cancer Center, New York, NY, (3) Siemens Medical Solutions, San Francisco, CA, (4) Siemens Healthcare GmbH, Forchheim, Germany, (5) FAU University, Erlangen, Germany

• 
  N Bennett

• 
  

TU-F-TRACK 3-6 (Tuesday, 7/27/2021) 4:30 PM - 5:30 PM [Eastern Time (GMT-4)]

Purpose: Cone-beam CT (CBCT) with automatic exposure control saturates the flat-panel detector in peripheral regions, leading to inferior image quality and excess skin dose. For example, in weight-bearing CBCT of knees, a traditional fixed bowtie filter cannot fully address this challenge. Thus, we constructed a prototype dynamic range reducer (DyRaR) which comprises four independently controlled brass wedges on each lateral side. The DyRaR is pre-programmed to conform to the patient-specific knee shape and continuously attenuate the beam beyond the knee outline as the gantry rotates. In this work, we demonstrate the DyRaR for weight-bearing CBCT using a knee phantom.

Methods: A knee phantom was first scanned using the weight-bearing CBCT trajectory to establish baseline image quality and to design phantom-specific DyRaR wedge positions. A second scan of the knee phantom was then acquired with the DyRaR mounted to the collimator of the C-arm. Finally, an air scan with just the DyRaR was acquired to correct the previous scan. Note that direct subtraction of the two scans is not ideal due to differences in beam hardening and scattering. Thus, we segmented the projections of knee+DyRaR scan into knee only, DyRaR only, and overlap regions, and used a 2nd-order polynomial to correct different regions.

Results: The DyRaR substantially attenuated the beam and prevented overexposure in air or near the knee outline by precisely following the designed trajectory. With proper projection correction, we could reconstruct CBCT images of the knee phantom with minor artifacts around the knee boundaries, although further image processing is needed to fully correct beam hardening and residual artifacts.

Conclusion: We have designed a prototype DyRaR and evaluated its application on a C-arm system for weight-bearing CBCT of knees. This study demonstrates the potential to design patient-specific DyRaR trajectories toward achieving optimal CBCT image quality.

Funding Support, Disclosures, and Conflict of Interest: NIH R01 AR065248; NIH S10 RR026714; Siemens Healthineers

Handouts

Keywords

Cone-beam CT, Collimation

Taxonomy

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

Contact Email