Premium
SU‐E‐T‐747: Uncertainty Reduction in Proton Prostate Planning
Author(s) -
Flampouri S,
Slopsema R,
Li Z
Publication year - 2011
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.3612711
Subject(s) - hounsfield scale , prostate , nuclear medicine , proton therapy , calibration , cone beam computed tomography , calibration curve , reduction (mathematics) , range (aeronautics) , mathematics , medicine , radiology , statistics , geometry , materials science , computed tomography , radiation therapy , cancer , detection limit , composite material
Purpose: To identify the sources of uncertainty in proton dose calculation for prostate treatment, quantify and minimize their effect. Methods: 125 prostate patients were used to measure the distribution of patient axial area at the mid‐level of the prostate. The CT images of a patient‐subgroup representative of various patient size and age, were segmented into thirteen different tissues. The tissue volume in the pelvis and within the dose receiving space (two lateral fields) were respectively recorded. The Hounsfield number (HU) of the tissues was also recorded. Three calibration curves from HU to proton relative stopping power (RSP) were created. The curves corresponding to the average patient size and extremes, were based on the stoichiometric method but they were specific to the prostate since only pelvic tissues were used. Additionally when constructing the curves the position of tissues inside the body was taken into account. For the creation of the final curve, tissue‐wise linear fit was used with the tissues weighted by their volume inside the beam space. The reduction on the uncertainty due to the prostate customized calibration curve was measured. Results: The prostate‐curve presented some differences with our current fit‐all calibration curve, mainly in the high density area. However the most significant effect was on the uncertainty reduction due to: prostate relevant patient size, position of tissues in the body, and appropriate tissue weighting. The uncertainty in the proton range was reduced for the average patient by 1% while for the bigger patients up to 1.5%. Conclusions: A customized prostate calibration curve was created taking into account the size of the patient and considering only pelvic tissues and their volume in the beam path of two lateral beams. This is a simple solution, easily implemented the reduces the proton range uncertainty and consequently the proximal and distal margins.