Proton radiography to improve proton therapy treatment

The quality of cancer treatment with protons critically depends on an accurate prediction of the proton stopping powers for the tissues traversed by the protons. Today, treatment planning in proton radiotherapy is based on stopping power calculations from densities of X-ray Computed Tomography (CT) images. This causes systematic uncertainties in the calculated proton range in a patient of typically 3-4%, but can become even 10% in bone regions [1-8]. This may lead to no dose in parts of the tumor and too high dose in healthy tissues [9]. A direct measurement of proton stopping powers with high-energy protons will allow reducing these uncertainties and will improve the quality of the treatment. Several studies have shown that a sufficiently accurate radiograph can be obtained by tracking individual protons traversing a phantom (patient) [4, 6, 10]. Our studies benefit from the gas-filled time projection chambers based on GridPix technology [11], developed at Nikhef, capable of tracking a single proton. A BaF2 crystal measuring the residual energy of protons was used. Proton radiographs of phantom consisting of different tissue-like materials were measured with a 30 x 30 mm(2) 150 MeV proton beam. Measurements were simulated with the Geant4 toolkit. First experimental and simulated energy radiographs are in very good agreement [12]. In this paper we focus on simulation studies of the proton scattering angle as it affects the position resolution of the proton energy loss radiograph. By selecting protons with a small scattering angle, the image quality can be improved significantly.