Fig 1.
Fixed horizontal beamline for modulated spot scanning delivery.
(a) Layout of the HIMM; (b) The PB scanning treatment head implemented in the MC code, showing the incident beam direction, scanning magnets (SMX, SMY), ionization chambers (IC); mini-RF, ripple filter; RS, range shifter. Unit: cm.
Table 1.
The initial values for all models.
Fig 2.
Schematic illustrations of concepts underlying the computational method.
Fig 3.
Depth-dose distributions and lateral-dose profiles of carbon ions.
MC calculations of laterally IDDDs in water (solid line) in comparison to measured data (open circles) for carbon ions at 190 MeV/u ((a) w/o ripple filter, (b) with ripple filter) and 260 MeV/u carbon ion ((c) w/o ripple filter, (d) with ripple filter). MC simulations of lateral-dose profiles of carbon ions in water, with 190 MeV/u (e) and 260 MeV/u (f) energy, respectively, sampled in before the Bragg peak, in comparison to experimental measurements taken at HIMM.
Fig 4.
Carbon-ion PB distribution was modeled by Origin software.
3D dose distributions of the total dose for (a)190 MeV/u and (b) 260 MeV/u carbon-ion PB. All data were normalized with the maximum dose value.
Fig 5.
(a) 3D standard Gaussian distribution and (d) the double Gaussian-logistic distribution was plotted in MATLAB. (b) and (c) The projection of Gaussian distribution was drawn in both x- and y-directions. (e) and (f) The projection of double Gaussian-logistic distribution was drawn in both x- and y-directions. (g) The flat plane for standard Gaussian distribution is plotted. (h) The flat plane for double gaussian-logistic distribution is plotted. (i) The calculated flatness and penumbra were compared in the standard Gaussian model and the double Gaussian-logistic model.
Fig 6.
The influence of different spot sizes on flatness and penumbra.
(a-c) The influence of different beam spot (2–6 mm) on dose uniformity and (d-f) penumbra for PB scanning (PBS) CIRT.
Fig 7.
The influence of delivered dose accuracy on flatness and penumbra.
(a-d) 3D Gaussian distribution with delivered dose error was plotted by MATLAB. (e) The influence of different random error (0.001, 0.01 and 0.1) on dose flatness, and (f) penumbra for PB scanning (PBS) CIRT.
Fig 8.
The influence of random loss points on flatness and penumbra.
(a) 3D Gaussian distribution with beam point loss was plotted by MATLAB. (b) and (c) The projection of Gaussian distribution was drawn in both x- and y-directions. (d) The calculated flatness and penumbra were compared in the standard Gaussian model and the double Gaussian-logistic model. p values: **<0.01. Non-significant differences are indicated as n.s.
Fig 9.
The influence of random loss lines on flatness and penumbra.
(a) 3D Gaussian distribution with beam line loss was plotted by MATLAB. (b) and (c) The projection of Gaussian distribution was drawn in both x- and y-directions. (d) The calculated flatness and penumbra were compared in the standard Gaussian model and the double Gaussian-logistic model. p values: **<0.01.
Fig 10.
The influence of spot position error on flatness and penumbra.
(a) 3D Gaussian distribution with beam spot position error was plotted by MATLAB. (b) and (c) The projection of Gaussian distribution was drawn in both x- and y-directions. (d) The calculated flatness and penumbra were compared in the standard Gaussian model and the double Gaussian-logistic model (left and right panel).