Intrafraction target motion during radiation therapy decreases the conformality and quality of the treatment. Methods to reduce the impact of intrafraction motion include passive techniques such as encompassing the motion and active techniques such as gating or real-time tracking. The Radixact linear accelerator is a helical tomotherapy (HT) device that contains an intrafraction motion management system, called Synchrony, which uses kilovoltage (kV) images to track the target motion, and jaw and multi-leaf collimator (MLC) movements to compensate for motion during delivery. At the start of this project, there were limited publications on this technology. The goal of this work was to investigate the ability of Radixact Synchrony to correct for intrafraction motion. This included analyzing the basic functions of Synchrony such as tracking and dosimetric delivery, and other aspects of the system such as dose deviations to organs risk (OARs) and patient dose from kV radiographs.Measurements and simulations were performed to investigate how beam characteristics change as the jaws sway and the MLC leaves shift. Due to the unflattened beam, jaw sway may decrease output by up to 3.6% and 1.7% for the 1.0 cm and 2.5 cm jaw settings, respectively. In addition, output can change by up to 2.6% for a shift of one MLC leaf and 5.1% for a shift of two MLC leaves, which correspond to shift magnitudes of 6.25 mm and 12.50 mm, respectively. However, unlike jaw sway which always decreases output, leaf shifts can increase or decrease output depending on the planned target location in the bore. Next, delivery quality assurance (DQA) measurements were performed using a Delta4 Phantom+ on a Hexamotion stage to investigate how synchronized plans with target motion compared to the static planned dose. Among 13 clinical HT treatment plans analyzed, all plans had gamma pass rates above the generally accepted action limit of 90% for standard IMRT QA criteria (3%, 2 mm, 10% threshold) and 11 of the 13 had pass rates greater than 99%. The effect of adding motion synchronization, as opposed to leaving the motion un-compensated, increased pass rates by 16.5% on average. This indicated that the motion synchronized plans deliver dose that is very similar to the planned dose.The ability of Synchrony to track target motion was investigated using measurements comparing known Delta4 phantom motion to Synchrony-modeled motion. Synchrony relies on a correlation between the patient's chest and the internal target motion. When the target/chest correlation was strong, root-mean square (RMS) tracking errors were less than 1.5 mm regardless of respiratory pattern, amplitude, or kV imaging period. However, larger errors (6 mm) were observed when changes to the target/chest correlation were introduced. The ability of the system to track these changes increased with more frequent imaging, suggesting that gantry periods should be kept short (20 s) for synchronized treatments. Next, a deformable dose accumulation framework was developed and used to calculate dose deviations to OARs from motion-synchronized treatments. Analysis was performed for five lung and five liver subjects by comparing the static planned dose to the doses calculated using the framework that included target motion, deformation, and motion of the jaws and MLC. Dose deviations from the static plan away from the target reached up to 19 Gy for one subject, and deviations of at least 6 Gy were observed for all 10 subjects. Several examples were presented in which the dose deviations may change the deliverability of a plan with respect to critical OAR dose constraints and ultimately sway clinical decisions. These calculations are important for tracking systems like Synchrony that only consider the location changes of the target and not OARs. Lastly, the patient dose and fiducial visibility on kV radiographs was explored. Patient dose was calculated by first measuring point doses in water and then creating a
展开▼
