Low-Dose Megavoltage Cone-Beam CT for Dose-Guided Radiation Therapy

Reviewer: Neha Vapiwala, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: October 21, 2003

Presenter: Jean Pouliot
Presenter's Affiliation: UCSF
Type of Session: Scientific


    Cone-beam CT scanning is an emerging technology that incorporates flat panel detectors for image acquisition rather than the ring-shaped detectors found in standard CT scanners. It offers the potential for 3-D correlation on fluoroscopy units as well as conventional linear accelerators, thus improving the ability to verify and achieve precise treatment delivery. This study is investigating the incorporation of megavoltage cone-beam CT (MVCT) for alignment with kilovoltage planning CT images, creation of 3-D setup from 2-D images, and development of dose-guided IMRT. Prior to its widespread use, however, various challenges exist, including the need to ensure low patient radiation exposure, acceptable quality for CT correlation, and potential for dose reconstruction. This experiment sought to address these challenges.

Materials and Methods

    Listed below are each of the phases of the experiment:
  • Standard linear accelerator (linac): 6 MV dose rate lowered to expose a flat silicone panel using 0.04-0.08 monitor units (MU) for each image obtained.
  • Image acquisition using Rando head phantom or frozen sheep head, with linac turned to deliver 0.01 monitor units (MU) per degree of rotation,
  • MU confirmed by ion chamber measurements
  • Continuous image acquisition, ~90-180 images combined
  • Image reconstruction to visualize 3-D bony anatomy and soft-tissue details
  • Torso phantom used to measure contrast resolution
  • Automatic 3-D image registration using mutual information algorithm
  • Patient alignment and repositioning as indicated based on displacement between image sets
  • Treatment delivery
  • Dose reconstruction using portal images to back-project beams onto the patient volume
  • Evaluation of delivered dose by comparing the registration between the MVCT 3-D image and the reconstructed dose
  • Plan modification as needed, based on the above


  • 3-D MVCT images reconstructed with delivered dose ranging from 5-15 cGy
  • Acquisition time of 30 seconds, time required for half of a gantry rotation
  • Processing time of ~90 seconds for reconstruction of a 256 m3 cube with 1.2 mm voxel size
  • Implanted markers (1 mm x 3 mm) easily visible for all exposure levels

Author's Conclusions

  • A low-dose rate beam and flat silicon imager can be used on a linac to obtain MVCT reconstructed images using as little as 5 MU.
  • The quality of these images, as well as the time needed for acquisition and reconstruction of images, is acceptable for clinical practice.

Clinical/Scientific Implications

    The above data indicates the potential for clinical implementaion of the innovative MVCT for patient alignment and dose verification in radiotherapy treatment. The ability to achieve low patient radiation exposure, together with the quality and rapidity of image acquisition and reconstruction are promising and support further investigation. A prime advantages of this new technology is its lack of dependence on atomic number of exposed elements, thus eliminating artifact of high-Z materials such as radioopaque markers, tooth fillings and hip replacements.

Oncolink's ASTRO Coverage made possible by an unrestricted Educational Grant from Siemens Medical Solutions.