ESTRO 2024 - Abstract Book

S4322

Physics - Intra-fraction motion management and real-time adaptive radiotherapy

ESTRO 2024

2291

Digital Poster

Evaluation of intrafractional movements for brain and prostate SBRT treatments with CyberKnife

Gábor Stelczer 1,2 , Vékás Márton 1 , Éva Folyovich 1 , Levente Jánváry 1 , Csilla Pesznyák 1,2 , Tibor Major 1,3

1 National Institute of Oncology, Center of Radiotherapy, Budapest, Hungary. 2 Budapest University of Technology and Economics, Institute of Nuclear Techniques, Budapest, Hungary. 3 Semmelweis University, Department of Oncology, Budapest, Hungary

Purpose/Objective:

The safety margins required for intrafractional movements of patients undergoing stereotactic brain or prostate radiotherapy (SRT) with a CyberKnife robotic linear accelerator were analysed.

Material/Methods:

For image-guided radiotherapy on CyberKnife, two perpendicular X-ray images are acquired of the treated region. During treatment, new images are taken in every 60 seconds on average, based on which the tracking algorithm updates the alignment inaccuracy values and corrects them using the robotic arm. The time interval between consecutive images can be set between 15 and 150 seconds, depending on the patient's stability. For intracranial treatments, the 6DSkull tracking algorithm calculates the alignment errors based on the appearance of the cranial bones. For prostate treatments, the Fiducial tracking algorithm determines the target volume misalignment by detecting the position of the four gold seeds (markers) inserted into prostate two weeks before the planning CT. We recorded all the calculated and corrected alignment errors in 50 fractions of 11 prostate treatments and 32 fractions of 17 brain treatments. The inaccuracy of patient’s setup can be characterised by the displacement between two consecutive imaging acquisitions. By interpreting these as separate "mini" fractions, we can apply the van Herk method (M=2.5×Σ+0.7×σ) to calculate the safety margins. In total, the displacements in longitudinal (LONG), lateral (LAT) and vertical (VERT) directions and the rotations (around the three axes) during the treatments were recorded for 82 fractions. In this work, targeting errors due to rotations were not investigated. The standard deviations of the mean displacements per fraction were used to obtain the systematic errors (Σ) for the population. In the LONG, LAT and VERT directions these values for cranial and prostate treatments were 0.06 mm, 0.06 mm and 0.06 mm, and 0.13 mm, 0.07 mm and 0.12 mm, respectively. By averaging the standard deviations of the individual intrafractional displacements, the random errors (σ) for the population were obtained, which were 0.2 mm, 0.17 mm and 0.17 mm in LONG, LAT and VERT directions for cranial, and 0.53 mm, 0.34 mm and 0.58 mm for prostate treatments, respectively. The calculated safety margins (M) required for targeting accuracy in the case of cranial targets in LONG, LAT and VERT directions were 0.3 mm, 0.3 mm and 0.3 mm, respectively. For prostate irradiations, the required safety margins (M) for targeting accuracy were 0.7 mm, 0.4 mm and 0.7 mm in the LONG, LAT and VERT directions, respectively. Results: