Dosimetric impact on fiducial array distortion for tracking in robotic radiosurgery
Carregando...
Data
Data de publicação
Autores IPEN
Orientador
Título da Revista
ISSN da Revista
Título do Volume
É parte de
É parte de
É parte de
Radiotherapy and Oncology
Exportar
Resumo
Purpose/Objective: Target tracking through online image guidance became an inherent part of the diffused usage of new
hypofractionated regimens. Tissue implanted fiducial markers is one of the available options widely used as a
surrogate for intrafraction target motion. The Cyberknife system uses a fiducial extraction algorithm capable of
tracking the fiducial array center of mass (COM) allowing a thorough evaluation of fiducial positioning. In order to flag
any array distortions, caused by organ deformation or fiducial migration, a parameter known as rigid body (RB),
determines differences in the distances between fiducial pairs as measured during the treatment and the planning.
If RB thresholds are out of tolerance, treatment interruptions are triggered and according to the manufacturer,
tracking accuracy might be affected. Despite manufacturer recommendations, dose differences caused by rigid body
errors (RBE), and whether the distance of targets from fiducial COM maximize this effect are not fully known.
Therefore, the objective of this work is to create a new 3D printed phantom capable of inducing fiducial errors,
mimicking fiducial array distortions, and measuring the dose differences caused by RBEs.
Material/Methods: A Cyberknife model M6 equipped with Incise2 MLC was used in this study. The new phantom contained three
embedded fiducials, 2 fixed (F2, F3) and one moveable (F1) in the longitudinal direction. Two plans were built in
Accuray Precision (V3.1.0) using two targets, T1 and T2 located at 2,5 cm and 6,5 cm respectively from fiducial COM.
Both plans were delivered firstly without induced errors, mimicking the ideal scenario. Subsenquencially, errors of 2
mm, 4 mm and 6 mm were applied in F1 using a caliper. Two semiflex A1SL ionization chambers, C1 and C2 were
used to measure the dose at targets T1 and T2 respectively. Additionally, an ETB3 film was placed in the phantom
coronal mid plane to access dose distribution. RBEs values and dose differences were registered and reported
accordingly. Figure 1 illustrate the axial (left) and coronal (right) slices for the plan created for T1. The purple and blue
lines represent isodoses of 8Gy and 2Gy respectively.
Results: Dose differences were seen to be greater as RBE gets wider and for targets farther from fiducial CDM. Displacing F1
by only 2mm, corresponding to an RBE of 1,8 mm in this study, led to an absolute dose difference of 7% for T2. When
displacing F1 by 6mm, yielding an RBE of 4,6 mm, the measured absolute dose differences reached up to 23% in the
high gradient region for T2. Conversely for T1 the impact was maintained bellow 5%, mitigated by the proximity with
COM. Furthermore, wider values of RBE led the fiducial extraction algorithm to trigger array rotations that were not
totally representative. The impact on dose distribution in the coronal mid plane of the phantom is illustrated in Figure
2 for T1 (lower) and T2 (upper).
Conclusion: RBE has been shown to be a powerful indicator of fiducial array deformation, albeit users must be aware that larger
RBE might induce array rotations compromising dosimetric accuracy. This is particularly important for targets farther
from fiducial COM. Performing these analyses is paramount for a better understanding of fiducial tracking accuracy
in Cyberknife and confirms manufacturer recommendations.
Como referenciar
PICIOLI, MARCELO R.; ZEITUNI, CARLOS A.; PLATA, ALVARO R.; MONGUA, JOSE L.R.; VILLAMIZAR, RIXY P.; ARCILA, JHONALBERT A.; MARANGONNI, FILIPPO; PINO, MATIAS. Dosimetric impact on fiducial array distortion for tracking in robotic radiosurgery: a phantom study. Radiotherapy and Oncology, v. 194, p. S4860-S4863, 2024. Supplement 1. DOI: 10.1016/S0167-8140(24)02384-3. Disponível em: https://repositorio.ipen.br/handle/123456789/48747. Acesso em: 31 Dec 2025.
Esta referência é gerada automaticamente de acordo com as normas do estilo IPEN/SP (ABNT NBR 6023) e recomenda-se uma verificação final e ajustes caso necessário.