Room: AAPM ePoster Library
Purpose: To develop an analytic statistical model for analyzing CTV random motion and provide a quantitative estimate of CTV coverage loss when PTV margin is zero.
Methods: Both CTV and PTV were assumed to be spherical with an equal radius (i.e., PTV margin=0). A previously developed statistical model was used to analyze the translational and rotational random motions of CTV. The combined random motion followed a 3D normal distribution with zero mean and standard deviation ?e. Percent CTV volume still covered by PTV caused by random motion P in the radial direction was calculated using the relation of sphere-sphere intersection and was also a random variable noted as Q. Cumulative distribution function (CDF) and reliability function of Q were derived from the CDF of P that follows the chi-square distribution. All distances were normalized to s?.
Results: Reliability function for Q was plotted as a function of various CTV sizes. Assuming s?=0.5mm, there is a significant chance (~20%) that the treatment might miss the target completely if the CTV is small (1mm in diameter or smaller). “95% of the times, at least 95% of the CTV is covered” is not achievable for smaller targets with CTV size<42mm in diameter. For common CTV sizes (20mm in diameter or less) encountered in single-fraction SRS, only 60%-90% of CTV could be covered by the prescription dose 95% of the time.
Conclusion: This work provides a quantitative estimate of CTV coverage loss for SRS/SBRT that was traditionally ignored due to the lack of proper modeling and/or blind belief of the accuracy of immobilization devices and delivery systems. The model predicted that for single-isocenter for multiple-targets technique, CTV %coverage probability will decrease with the increased distance between the CTV and iso-center because the larger distance increases rotational error and therefore ?e.