2072 - OneIsoMR: A Laser-Free, Direct MR-to-Radiation Alignment Phantom for Precision MR-Guided Radiotherapy

02:30pm - 04:00pm PT

Hall F

Screen: 7

POSTER

Presenter(s)

Clinton Gibson, BS - Stanford Healthcare, Palo Alto, CA

C. T. Gibson¹, D. P. I. Capaldi², Y. Gao³, Z. Yang⁴, S. J. Yu⁵, and L. Skinner⁶; ¹Department of Radiation Oncology, Stanford Health Care, Stanford, CA, ²Department of Radiation Oncology, University of California San Francisco, San Francisco, CA, ³Department of Radiation Oncology, Stanford University, Stanford, CA, ⁴Department of Radiation Oncology, Stanford University School of Medicine, Stanford, CA, ⁵Department of Radiation Oncology, Stanford University, Palo Alto, CA, ⁶Stanford University, Stanford, CA

Purpose/Objective(s): We developed a novel phantom, named OneIso-MR, to precisely measure the offsets between the magnetic resonance (MR) imaging isocenter and the radiation-isocenter of MR-guided radiotherapy systems. Currently an indirect measurement is made through visually aligning to a laser and obtaining both the laser-to-MR and laser-to-radiation offsets individually. The challenge with a combined phantom is that it needs markers that are MR-visible, which are also precisely registered to mega voltage x-ray visible fiducials. This needs to be accurate within a fraction of a mm, and the metal alloys should not include any magnetic elements. By improving MR-to-radiation alignment we can deliver more precise stereotactic radiotherapy.

Materials/Methods: A watertight 3D printed plastic phantom was designed with internal plastic pyramids that indicate the center of the phantom on the MR image. A pocket houses copper-tungsten alloy cylinders that form an “X” pointing to the phantom center. Copper-tungsten alloy was chosen for its density and lack of iron, which is common in many tungsten alloys. Two orthogonal film planes were placed through the phantom center to hold EBT3 radiochromic film. The OneIso-MR phantom was then imaged, aligned and delivered on an MR-guided radiotherapy linac. The four beam angles 45, 135, 225, and 315 each with a 4x4 cm field size, left a projection of the 9.6mm circular tungsten and the field edges on the film planes. The Field edges indicate radiation positioning and the tungsten circles are registered to indicate the MR-center on the film. These were analyzed using Image J software. Uncertainty analysis was performed of both the old and new methods.

Results: Offsets of 0.09, -0.04 and 0.89 mm were found in the lateral, vertical, and longitudinal directions with an uncertainty of +-0.21mm in each direction. This is broadly consistent with the standard method which yielded 0.0, 0.0, 1.0 mm each with an uncertainty of +-0.76 mm. The tables below describe the uncertainty analysis.

Conclusion: MRI to radiation isocenter offsets were measured directly on Film with at least double the precision of the prior manual process. Removal of the visual alignment to lasers also simplifies workflow and reduces the need for tightly maintaining laser position.

Abstract 2072 - Table 1

Standard Method	Uncertainty (mm)	determination
Laser-to-Rad: Manual alignment to laser	0.5	Uncertainty in visual alignment to laser
Laser-to-Rad: detector measurement noise	0.2	Standard error in 10 repeated measurements
Laser-to-MR: Manual alignment to laser	0.5	Uncertainty in visual alignment to laser
Laser to MR: image registration	0.2	Uncertainty in manual MR registration (5 repeats)
Total uncertainty	0.76	Summed in quadrature
OneIso-MR Method
Rad-to-MR: image registration	0.2	Uncertainty in manual MR registration (5 repeats)
Rad-to-MR: Measurement noise on film profiles	0.05	Statistical error in film measurements
Total uncertainty	0.21	Summed in quadrature