Evaluation of a head rest prototype for rotational corrections in three degrees of freedom.

Cranial stereotactic irradiations require accurate reproduction of the planning CT patient position at the time of treatment, including removal of rotational offsets. A device prototype was evaluated for potential clinical use to correct rotational positional offsets in image-guided radiotherapy workflow. Analysis was carried out with a prototype device "RPS head" by gKteso GmbH, rotatable up to 4° in three dimensions by hand wheels. A software tool accounts for the nonrectangular rotation axes and also indicates translational motions to be performed with the standard couch to correct the initial offset and translational shifts introduced by the rotational motion. The accuracy of angular corrections and positioning of an Alderson RANDO head phantom using the prototype device was evaluated for nine treatment plans for cranial targets. Corrections were obtained from cone beam computed tomography (CBCT) imaging. The phantom position was adjusted and the final position was then verified by another CBCT. The long-term stability of the prototype device was evaluated. Attenuation by the device along its three main axes was assessed. A planning study was performed to evaluate if regions of high-density material can be avoided during plan generation. The device enabled the accurate correction of rotational offsets in a clinical setup with a mean residual angular difference of (0.0 ± 0.1)° and a maximum deviation of 0.2°. Translational offsets were less than 1 mm. The device was stable over a period of 20 min, not changing the head support plate position by more than (0.7 ± 0.6) mm. The device contains high-density material in the adjustment mechanism and slightly higher density in the support structures. These can be avoided during planning generation maintaining comparable plan quality. The head positioning device can be used to correct rotational offsets in a clinical setting.

Technical note: Determining equivalent squares of high-energetic photon fields.

PURPOSE: We developed a method based on a physical pencil beam model for accurate equivalent square calculations for rectangular and irregular fields, for different definitions of equivalent squares, for beams with and without flattening filter, different photon energies, and depths in water. METHODS: We considered two equivalent square definitions: equal dose at a point on the beam axis and equal depth dose, measured as tissue phantom ratio at 20 and 10 cm depth ( TPR 20 , 10 $\text{TPR}_{20,10}$ ). As dose engine, we used an analytical pencil beam model. By integrating the pencil beam kernels, we assigned square fields to rectangular fields minimizing the dose, respectively, the TPR 20 , 10 $\text{TPR}_{20,10}$ difference. The results were compared with measurements at 100 mm depth for nominal beam energies of 6 and 18 MV, the Sterling equation, the geometric mean, and data from BJR Suppl 25 (British Institute of Radiology, 1996). RESULTS: Pencil beam results were closest to the measurements. An energy dependence of several millimeters for small field dimensions and depth dependencies for very elongated fields were observed. For the assignment of WFF square to FFF rectangular fields, using the equal- TPR 20 , 10 $\text{TPR}_{20,10}$ definition, our method agrees with previously published results. For circular fields approximated by leaves, we found deviations to the data from BJR Suppl. 25 below 1 mm for diameters smaller than 200 mm. CONCLUSIONS: Our study shows that the validity range for geometric mean and Sterling equation is limited. Ergo, instead of specifying specific validity ranges, we suggest using the pencil beam method, valid for all aspect ratios, including elongated fields in the primary dose dominated regime. We published our method as python library and graphical user interface on GitHub. Users can choose between two definitions of equivalent square and between WFF and FFF fields. The implemented pencil beam method for irregular fields is also usable for quality assurance such as monitor unit checks.

Small field output correction factors at 18 MV.

BACKGROUND: The response of various detectors in the radiotherapy energy range has been investigated, especially for 6 and 10 MV energies for small fields, and is summarized in TRS-483. However, data for accelerator energies above 10 MV are sparse or unavailable for many detectors, especially for the energy of 18 MV. Small variations in field output factors for the commissioning of a treatment planning system can have a high impact on calculation of dose distributions. PURPOSE: Many studies describe an energy dependence of the response for a large number of detectors. We wanted to close the gap for the 18 MV energy regime and determined field output correction factors for different detectors at 18 MV. METHODS: An ELEKTA Versa HD accelerator at 18 MV was used together with a PTW MP3 water phantom at an SSD of 90 cm. The following detectors were examined: PTW Semiflex 31021, PinPoint 3D 31022, diode 60012, diode 60008 and microDiamond 60019, Sun Nuclear EDGE detector, IBA PFD, SFD, Razor Chamber, Razor Nano Chamber and Razor Diode, Standard Imaging Scintillator Exradin W2 1x3, W2 1x1 and Gafchromic EBT3 film. The dose response was determined at a depth of 10 cm for square fields between 0.5 and 10 cm side length. As reference data a composure of radiochromic film data for small fields ( s ≤ 3 $s\le 3$  cm) and data of all compatible chambers for larger fields ( s ≥ 3 $s\ge 3$  cm) was used. The effective field sizes of small fields were determined from profiles obtained on radiochromic film. The obtained field output correction factors obey the rules of the TRS-483 protocol. RESULTS: The W2 1x1 scintillator and the Razor Chamber showed the smallest deviations from the reference curve. The shielded diodes (diode 60008, EDGE detector) showed the highest over-response at small fields, followed by PFD, microDiamond and the unshielded diodes (diode 60012, SFD). The ionization chambers exhibited the well-known volume effect, that is, strong under-response at small fields of up to 9% for the PinPoint 3D, 7% for the Razor Chamber and up to 30% for the Semiflex detector for the smallest studied field size. The small chambers showed a polarity effect in axial orientation, especially the Razor Nano Chamber. Corrections at 18 MV are generally larger than those provided by TRS-483, continuing the trend of increasing corrections between 6 and 10 MV also at a higher accelerator energy. Only the PinPoint 3D Chamber showed a slightly smaller correction. CONCLUSIONS: Field output correction factors were determined for square field sizes between 0.5 and 10 cm at 18 MV. Most detectors needed a larger correction than at 6 and 10 MV. Thus, the use of correction factors will improve beam data for 18 MV.