Study On Dose Uncertainty Associated With Tissue Inhomogeneity Using The Cyberknife Synchrony Respiratory Tracking System

?Objective?This study used the Cyber Knife? Synchrony Respiratory Tracking System(SRTS)to treat a target with variable motion ranges inside a QUASARTMphantom with variable tissue thicknesses,then compared the measured point doses of the target with3D/4D accumulated doses.The effect of density heterogeneous on dose uncertainty has been investigated and evaluated.?Method?The Cyber Knife?,combined with the Synchrony Respiratory Tracking System(SRTS),is a robotic treatment platform designed to track the tumor during radiotherapy.Due to the complexity of respiration-induced real-time tracking of the tumor,certain variations of the patient model(e.g.,tissue density)cannot be compensated by the SRTS.The QUASARTMphantom has been divided into 8 groups by the bones thickness and motion state,separately imitating different thickness of bones(0mm,10 mm,20mm,40mm)and different motion state(30mm,12 BPM or static).Then compared the measured point doses of the target with 3D/4D accumulated doses.The dose has been calculated by the two algorithms(Ray-tracing and Monte Carlo).A right lung cancer case was also chosen to compare 3D/4D dose distributions by the two algorithms of the target and organs at risk.?Results? When the phantom was static,the dose deviations between 3D dose and measured point dose on 10 mm,20 mm and 40 mm of additional tissue thickness was3.48%?2.90%?3.39%?3.31%(Ray-tracing);0.02%?-0.81%?1.01%?-1.04%(Monte Carlo).When the phantom was dynamic,the dose deviations between 4D accumulated dose and measured point dose on 10 mm,20 mm and 40 mm of additional tissue thickness was5.53%?5.81%?6.15%?5.13%(Ray-tracing);2.06%?1.66%?2.91%?1.27%(Monte Carlo).In the lung cancer case,the 4D accumulated doses by Monte Carlo algorithm can make a better simulation of dose distributions in the targets and OARs.?Conclusion? 1.Excluding the influence of tracking errors,when the phantom was dynamic,the dose deviations between 3D dose and measured point dose were mainly influenced by patient model changes.The 3D/4D accumulated doses by Monte Carlo algorithm showed better agreement with the measured point dose,all deviations being below 3%.2.A right lung cancer case showed that 3D doses and 4D accumulated doses vary in the target edge and the surrounding OARs.The 4D accumulated doses and 4D treatment planning by Monte Carlo algorithm can make a better simulation of dose distributions in the targets and OARs than 3D doses and static treatment planning,especialy in the case of dynamic and tissue heterogeneous.