"We think that using PET to monitor ion beam range is a very useful method," said Mitsutaka Yamaguchi, from the Takasaki Advanced Radiation Research Institute at the Japan Atomic Energy Agency (JAEA). "However, the main disadvantage of PET is that some of the main positron emitters have long lifetimes, and the metabolic washout effect occurs." This effect, in which the emitters move within the body prior to decaying, leads to different generation and decay positions and complicates range estimation.
Because bremsstrahlung photons are emitted immediately upon interaction with the incident ion beam, this radiation does not suffer from the washout effect. Furthermore, the photon generation rate is higher than that attained through nuclear reactions. "The main advantage is the lower measurement time, owing to the large generation rate of bremsstrahlung photons," Yamaguchi explained.
By using the dependence of the bremsstrahlung photon energy spectrum on the incident ion energy, it's possible to measure the range position from the spectrum data. The authors note that the method is particularly suited to use with pencil beam scanning techniques, which require high efficiency monitoring of the irradiation of each spot.
Proof of principle
Bremsstrahlung can be classified according to its underlying physical mechanism. For this study, the researchers assumed that the dominant contribution was from secondary electron bremsstrahlung (SEB). To confirm this fact, Yamaguchi and co-workers measured the upper energy limit of the bremsstrahlung spectrum for a low-energy carbon ion beam. The measured upper limit agreed with the theoretical SEB upper limit, indicating that the SEB process was the main component.
To estimate the range position from a bremsstrahlung spectrum, the researchers first calculated the amount of SEB generated in the energy range of 63–68 keV at an emission angle of 90°. The derivative of the bremsstrahlung curve (generated bremsstrahlung versus ion position, with the range position set as the origin) showed a sudden change at –2.7 mm from the origin. This characteristic position indicates the ion beam range.
To validate the monitoring method experimentally, the team checked this characteristic position using a 290 MeV/u carbon ion beam incident upon a water phantom. A CdTe semiconductor detector was placed at the same height as the beam line at a scattering angle of 90°, and a lead collimator used to restrict the direction of the detected photons.
Bremsstrahlung photons with energies of up to around 800 keV were measured, and the overall photon count seen to decrease with increasing energy. The researchers analysed the energy range from 63 to 68 keV – where photon count was high. The photon count decreased as the detector position approached the range, and the derivative of this decrease clearly changed in front of the range position, as predicted by the theoretical analysis.
To calculate the characteristic point of this slope change, the data were fitted with two lines, to fit data from –50 to –8 mm, and from 0 to 50&mm. The intersection point of the two lines was at –6.0 mm. This is slightly smaller than the theoretical prediction of –2.7 mm and indicates that this method can deduce the range position with an accuracy of about 4 mm.
Yamaguchi and colleagues conclude that this study demonstrates the feasibility of accurately determining range position from bremsstrahlung observations. They note that because the energies of bremsstrahlung photons are lower than those of positron annihilation photons, their attenuation in the body is greater (the penetration of 60 keV photons, for example, is about 37% through 5 cm of water), and suggest that this technique is most suited to monitoring irradiation of shallow tumours.
Next, the team plan to optimize the imaging process. "We will develop an imaging apparatus to measure the bremsstrahlung photons efficiently," said Yamaguchi. "We are also trying to develop a physical model to explain the experimental results precisely."
• Related articles in PMB
Beam range estimation by measuring bremsstrahlung
Mitsutaka Yamaguchi et al Phys. Med. Biol. 57 2843
Extension and validation of an analytical model for in vivo PET verification of proton therapy—a phantom and clinical study
F Attanasi et al Phys. Med. Biol. 56 5079
Design study of an in situ PET scanner for use in proton beam therapy S Surti et al Phys. Med. Biol. 56 2667
On the effectiveness of ion range determination from in-beam PET data Fine Fiedler et al Phys. Med. Biol. 55 1989