Proton computed tomography (pCT) has been proposed as an alternative to X-ray computed tomography (CT) for acquiring relative to water stopping power (RSP) maps used for proton treatment planning dose calculations. In parallel, it has been shown that dual energy X-ray CT (DECT) improves RSP accuracy when compared to conventional single energy X-ray CT. This study aimed at directly comparing the RSP accuracy of both modalities using phantoms scanned at an advanced prototype pCT scanner and a state-of-the-art DECT scanner. Two phantoms containing 13 tissue-mimicking inserts of known RSP were scanned at the pCT phase II prototype and a latest generation dual-source DECT scanner (Siemens SOMATOM Definition FORCE). RSP accuracy was compared by mean absolute percent error (MAPE) over all inserts. A highly realistic Monte Carlo (MC) simulation was used to gain insight on pCT image artifacts which degraded MAPE. MAPE was 0.55% for pCT and 0.67% for DECT. The realistic MC simulation agreed well with pCT measurements (MAPE = 0.69%).

We present a method to accurately predict image noise in proton computed tomography (pCT) using data generated from a Monte Carlo simulation and a patient or object model that may be generated from a prior x-ray CT image. This enables noise prediction for arbitrary beam fluence settings and, therefore, the application of fluence-modulated pCT (FMpCT), which can achieve prescribed noise targets and may significantly reduce the integral patient dose. We extended an existing Monte Carlo simulation of a prototype pCT scanner to include effects of quenching in the energy detector scintillators and constructed a beam model from experimental tracking data. Simulated noise predictions were compared to experimental data both in the projection domain and in the reconstructed image. Noise prediction agreement between simulated and experimental data in terms of the root-mean-square (RMS) error was better than 7% for a homogeneous water phantom and a sensitometry phantom with tubular inserts. For an anthropomorphic head phantom, modeling the anatomy of a five-year-old child, the RMS error was better than 9% in three evaluated slices. We were able to reproduce subtle noise features near heterogeneities. To demonstrate the feasibility of Monte Carlo simulated noise maps for fluence modulation, we calculated a fluence profile that yields a homogeneous noise level in the image. Unlike for bow-tie filters in x-ray CT this does not require constant fluence at the detector and the shape of the fluence profile is fundamentally different. Using an improved Monte Carlo simulation, we demonstrated the feasibility of using simulated data for accurate image noise prediction for pCT. We believe that the agreement with experimental data is sufficient to enable the future optimization of FMpCT fluence plans to achieve prescribed noise targets in a fluence-modulated acquisition.

No abstract

Monte Carlo (MC) simulation is considered as the most accurate method for calculation of absorbed dose and fundamental physics quantities related to biological effects in carbon ion therapy. To improve its computational efficiency, we have developed a GPU-oriented fast MC package named goCMC, for carbon therapy. goCMC simulates particle transport in voxelized geometry with kinetic energy up to 450 MeV/u. Class II condensed history simulation scheme with a continuous slowing down approximation was employed. Energy straggling and multiple scattering were modeled. δ-electrons were terminated with their energy locally deposited. Four types of nuclear interactions were implemented in goCMC, i.e., carbon-hydrogen, carbon-carbon, carbon-oxygen and carbon-calcium inelastic collisions. Total cross section data from Geant4 were used. Secondary particles produced in these interactions were sampled according to particle yield with energy and directional distribution data derived from Geant4 simulation results. Secondary charged particles were transported following the condensed history scheme, whereas secondary neutral particles were ignored. goCMC was developed under OpenCL framework and is executable on different platforms, e.g. GPU and multi-core CPU. We have validated goCMC with Geant4 in cases with different beam energy and phantoms including four homogeneous phantoms, one heterogeneous half-slab phantom, and one patient case. For each case 3 × 107 carbon ions were simulated, such that in the region with dose greater than 10% of maximum dose, the mean relative statistical uncertainty was less than 1%. Good agreements for dose distributions and range estimations between goCMC and Geant4 were observed. 3D gamma passing rates with 1%/1 mm criterion were over 90% within 10%) isodose line except in two extreme cases, and those with 2%/1 mm criterion were all over 96%. Efficiency and code portability were tested with different GPUs and CPUs. Depending on the beam energy and voxel size, the computation time to simulate 107 carbons was 9.9–125 sec, 2.5–50 sec and 60–612 sec on an AMD Radeon GPU card, an NVidia GeForce GTX 1080 GPU card and an Intel Xeon E5-2640 CPU, respectively. The combined accuracy, efficiency and portability make goCMC attractive for research and clinical applications in carbon ion therapy.

Proton computed tomography (CT) has been described as a solution for imaging the proton stopping power of patient tissues, therefore reducing the uncertainty of the conversion of x-ray CT images to relative stopping power (RSP) maps and its associated margins. This study aimed to investigate this assertion under the assumption of ideal detection systems. We have developed a Monte Carlo framework to assess proton CT performances for the main steps of a proton therapy treatment planning, i.e. proton or x-ray CT imaging, conversion to RSP maps based on the calibration of a tissue phantom, and proton dose simulations. Irradiations of a computational phantom with pencil beams were simulated on various anatomical sites and the proton range was assessed on the reference, the proton CT-based and the x-ray CT-based material maps. Errors on the tissue's RSP reconstructed from proton CT were found to be significantly smaller and less dependent on the tissue distribution. The imaging dose was also found to be much more uniform and conformal to the primary beam. The mean absolute deviation for range calculations based on x-ray CT varies from 0.18 to 2.01 mm depending on the localization, while it is smaller than 0.1 mm for proton CT. Under the assumption of a perfect detection system, proton range predictions based on proton CT are therefore both more accurate and more uniform than those based on x-ray CT.