CT Calibration Curves for Proton Therapy: A Practical Approach
C Yeager*, C Ainsley, University of Pennsylvania, Philadelphia, PASU-E-T-722 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: To determine a practical approach to CT calibration for proton radiotherapy.
Methods: Measurements of Hounsfield units (HU) of tissue substitutes of known elemental composition were made in a variety of scenarios using two series of plugs manufactured by different vendors. We determined the optimal area within the cross-section of the plugs over which to assess the mean HU, and then studied the mean HU this way for each plug as a function of the slice thickness of the scan, the location of the plane selected for evaluation within the plug, the location of the plug in a 30x40x40 cm3 water-filled tank, the orientation of this phantom, whether the tank was water-filled or air-filled, and whether the plugs were measured individually, one scan at a time, or simultaneously, in a single scan. Coefficients of the commonly-employed stoichiometric calibration of the CT scanner for proton therapy were determined for each aforementioned scenario, and used to calculate CT-number-to-proton-stopping-power-ratio tissue characterization curves.
Results: Different slice thicknesses show no meaningful difference in HU reading. No meaningful difference in HU reading was found when taken at various locations within the plug. CT calibration curve values were found to be within 2% for three different positions of the plug within the phantom, for both orientations of the phantom, whether measured in water or in air, and whether plugs were scanned separately in air or all at once.
Conclusion: We assessed the variation in the CT-number-to-proton-stopping-power-ratio tissue characterization curve to be expected for a variety of different measurement scenarios. We found that such a curve is relatively insensitive to the setup used clinically and we therefore conclude that the necessary measurement data can be obtained quickly and confidently.