Premium
SU‐G‐TeP2‐13: Patient‐Specific Reduction of Range Uncertainties in Proton Therapy by Proton Radiography with a Multi‐Layer Ionization Chamber
Author(s) -
Deffet S,
Farace P,
Righetto R,
Macq B,
Vander Stappen F
Publication year - 2016
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 - 10.1118/1.4957048
Subject(s) - imaging phantom , proton therapy , stopping power , hounsfield scale , radiography , proton , ionization chamber , computed radiography , ionization , pencil beam scanning , range (aeronautics) , nuclear medicine , beam (structure) , radiation treatment planning , materials science , optics , physics , detector , nuclear physics , computer science , medicine , computed tomography , artificial intelligence , radiology , ion , radiation therapy , composite material , quantum mechanics , image quality , image (mathematics)
Purpose: The conversion from Hounsfield units (HU) to stopping powers is a major source of range uncertainty in proton therapy (PT). Our contribution shows how proton radiographs (PR) acquired with a multi‐layer ionization chamber in a PT center can be used for accurate patient positioning and subsequently for patient‐specific optimization of the conversion from HU to stopping powers. Methods: A multi‐layer ionization chamber was used to measure the integral depth‐dose (IDD) of 220 MeV pencil beam spots passing through several anthropomorphic phantoms. The whole area of interest was imaged by repositioning the couch and by acquiring a 45×45 mm 2 frame for each position. A rigid registration algorithm was implemented to correct the positioning error between the proton radiographs and the planning CT. After registration, the stopping power map obtained from the planning CT with the calibration curve of the treatment planning system was used together with the water equivalent thickness gained from two proton radiographs to generate a phantom‐specific stopping power map. Results: Our results show that it is possible to make a registration with submillimeter accuracy from proton radiography obtained by sending beamlets separated by more than 1 mm. This was made possible by the complex shape of the IDD due to the presence of lateral heterogeneities along the path of the beam. Submillimeter positioning was still possible with a 5 mm spot spacing. Phantom specific stopping power maps obtained by minimizing the range error were cross‐verified by the acquisition of an additional proton radiography where the phantom was positioned in a random but known manner. Conclusion: Our results indicate that a CT‐PR registration algorithm together with range‐error based optimization can be used to produce a patient‐specific stopping power map. Sylvain Deffet reports financial funding of its PhD thesis by Ion Beam Applications (IBA) during the confines of the study and outside the submitted work. Francois Vander Stappen reports being employed by Ion Beam Applications (IBA) during the confines of the study and outside the submitted work.