Intensity-modulated radiotherapy plan optimization system, medium and equipment guided by predictive dose distribution

Abstract

The invention provides an intensity modulated radiation therapy plan optimizing method based on predicted dosage distribution guidance and application. The method includes: inputting geometric structure features of a region of interest of a patient into a trained neural network model to obtain organ-endangering three-dimensional dosage distribution prediction; taking the three-dimensional dosage distribution prediction as optimizing guidance and voxel in the region of interest as optimizing constraint to build a radiation field intensity distribution model, wherein an optimized objective function of the radiation field intensity distribution model comprises an objective item based on the three-dimensional dosage distribution prediction and an objective item based on equivalent uniform dosage; on the basis of the optimized objective function, solving an optimizing problem of the radiation field intensity distribution model to acquire an intensity modulated radiation therapy plan. Predicted three-dimensional dosage distribution is utilized to guide plan optimizing of the intensity modulated radiation therapy plan, so that personalized optimizing and voxel-level accurate dosage optimizing can be realized.

Description

technical field [0001] The invention relates to the technical field of radiotherapy planning, in particular to an optimization method and application of an intensity-modulated radiotherapy plan guided by predicted dose distribution. Background technique [0002] Intensity Modulated Radiation Therapy (IMRT) is currently the most widely used tumor radiotherapy technology, which can form a highly conformal and uniform dose distribution inside the target area, while reducing the number of organs at risk and The irradiation dose of normal tissues around the target area can effectively increase the gain ratio of tumor treatment. In the planning design of IMRT, since the ideal dose target or constraint is unknown before the plan design, the plan designer often selects the dose target or constraint based on the current clinical norms based on population statistics, supplemented by his own clinical experience, using artificial The trial and error (Trial and Error) method repeatedly ...

AI Technical Summary

Problems solved by technology

However, this kind of prediction has uncertainty, which will have a great impact on subsequent optimization guidance. How to reasonably and effectively apply the predicted dose distribution information is an important and difficult point

In 2018, Fan et al., in "Automatic treatment planning based on three-dimens ional dose distribution predicted from deep learning technique", took reproducing the predicted dose distribution as the optimization solution strategy, and guided the plan optimization by introducing it into the objective function. Although this method can obtain a feasible plan, the quality of the obtained plan is often only close to the forecast or original plan, thus limiting the space for optimal solution to a certain extent

Images