Dynamic optimization system based on respiratory movement

Abstract

The invention discloses a dynamic optimization system based on respiratory movement, and belongs to a radiotherapy plan optimization system. The dynamic optimization system specifically comprises a time phase data processing module which obtains a CT image and acquisition time of each breathing phase by using 4D CT, a dose calculation module used for calculating a dose contribution matrix of each respiratory phase, and calculating the dose of an organ sampling point with the time probability according to a dynamic model formula, an optimization judgment module used for verifying the organ position and the irradiation dose, conducting plan optimization if deformation of the organ position or an error of the irradiation dose exists, and otherwise, adopting original plan parameters, and a sub-field optimization module used for adopting a sub-field optimization method for plan optimization, obtaining the dose of each organ point according to a dynamic model, calculating the shape and weight of a sub-field meeting the expected dose when the minimum value of the target function is calculated, and outputing a result. According to the method, plan optimization is carried out by considering a dynamic model, and the optimal plan parameters are automatically optimized.

Description

technical field [0001] The invention relates to a radiotherapy plan optimization system, in particular to a dynamic optimization system based on breathing motion. Background technique [0002] The planning optimization system calculates planning parameters through computer software. The optimized planning parameters include subfield weights and subfield shapes. Traditional calculations include "two-step method" and direct subfield optimization. The "two-step method" is divided into two steps: intensity optimization and subfield segmentation. Intensity optimization is to calculate the intensity distribution corresponding to each field according to the optimization algorithm. Subfield segmentation is to convert the intensity distribution into the subfield shape of each MLC. Direct subfield optimization is the subfield shape and subfield weight that can be directly implemented through direct optimization calculation. [0003] At present, the optimization of planning parameters...

AI Technical Summary

Problems solved by technology

[0003] At present, the optimization of planning parameters is "two-step method" or direct subfield optimization. In practice, image images and dose images can be used to adjust the plan according to the feedback of image changes to make the plan more accurate. At present, the method of adjusting the plan is mostly through Manually adjust the plan parameters on the computer for optimization. The manual adjustment method takes a lot of time and finds a plan that meets expectations through trial and error.

At the same time, these optimizations do not take into account the impact of breathing timing. In the existing technology, reducing the impact of breathing movement is mainly through machine-assisted control of breathing, such as respiratory gating technology. Instead of establishing a dynamic model for optimization, and manually adjusting the plan It is not guaranteed to be the optimal solution

Images