US-12616851-B2 - Two-step beam geometry optimization and beam entry angles without isocenter

US12616851B2US 12616851 B2US12616851 B2US 12616851B2US-12616851-B2

Abstract

Systems and methods are disclosed for optimizing a treatment plan using all degrees of freedom including those related to beam geometry parameters, the optimization including a step for limiting the search space for the beam geometry parameters using a trained machine learning model, and systems and methods are disclosed for obtaining beam geometry parameters for treatment planning that do not require knowledge of the beam delivery device isocenter.

Inventors

Mikko Hakala
Shahab BASIRI
Kellee Donnelly
Elena CZEIZLER
Esa KUUSELA

Assignees

SIEMENS HEALTHINEERS INTERNATIONAL AG

Dates

Publication Date: 20260505
Application Date: 20230913

Claims (20)

1 . A method for optimizing a treatment plan for a radiation beam delivery system, the method comprising: a first step configured to limit a search space for an optimization objective of an optimization algorithm, the optimization objective being a beam geometry (BG) parameter of the radiation beam delivery system; and a second step configured to optimize the beam geometry (BG) parameter based on the limited search space obtained in the first step without explicit determination of the radiation beam delivery system isocenter.
2 . The method of claim 1 , wherein the first step includes one of: predicting an initial choice for the beam geometry (BG) parameter, inputting, by a user, the initial choice for the beam geometry (BG) parameter, or using a template to enter the initial choice for the beam geometry (BG) parameter, wherein the predicting of the initial choice for the beam geometry (BG) parameter includes executing a machine learning model that is trained to predict the initial choice for the beam geometry (BG) parameter based on patient geometry.
3 . The method of claim 2 , wherein the first step further includes one of: predicting a feasible range for the beam geometry (BG) parameter and/or a largest acceptable deviation from the initial choice for the beam geometry (BG) parameter, inputting, by the user, a feasible range for the beam geometry (BG) parameter and/or a largest acceptable deviation from the initial choice for the beam geometry (BG) parameter, or templating a feasible range for the beam geometry (BG) parameter and/or a largest acceptable deviation from the initial choice for the beam geometry (BG) parameter.
4 . The method of claim 3 , wherein the beam geometry (BG) parameter includes one of a gantry angle, a collimator angle, and a radiation field direction of the radiation beam delivery system.
5 . The method of claim 3 , wherein the second step includes: receiving, as input to the treatment plan optimization algorithm, information regarding desired dose distribution within a treatment volume of a patient, and treatment parameters; calculating dose distribution within the treatment volume by executing the treatment plan optimization algorithm; determining whether the calculated dose distribution is within an acceptable threshold of the desired dose distribution; and iteratively modifying one or more of the treatment parameters including the beam geometry (BG) parameter until the calculated dose distribution is within the acceptable threshold or an endpoint has been reached, wherein the beam geometry (BG) parameter is modified based on the search space obtained in the first step.
6 . The method of claim 5 , wherein the beam geometry (BG) parameter is modified at each iteration of the optimization process using one of a trained reinforcement learning agent, a simplex method, a simulated annealing method, or any other deterministic, stochastic or heuristic optimization method.
7 . The method of claim 2 , wherein the machine learning model is a knowledge-based supervised machine learning model trained using historical treatment plans.
8 . The method of claim 1 , wherein the optimization step includes executing an automated process to predict an initial choice for the beam geometry (BG) parameter without explicit determination of the radiation beam delivery system isocenter.
9 . The method of claim 8 , wherein the automated process includes: executing a trained machine learning model to predict, based on patient geometry as input, a beam entry angle relative to a reference point in the patient; and determining the initial beam geometry (BG) parameter based on the predicted beam entry angle.
10 . The method of claim 9 , wherein the determining of the initial beam geometry (BG) parameter includes: displaying the predicted beam entry angle superimposed on the patient geometry on a display for evaluation by a user; allowing the user to accept the predicted beam entry angle or interactively modify the predicted beam entry angle on the display until an acceptable beam entry angle is obtained; and allowing the user to accept the reference point as the isocenter or manually enter an isocenter, wherein the accepted beam entry angle is chosen as the initial beam geometry (BG) parameter when the reference point is accepted as the isocenter, and the initial beam geometry (BG) parameter is calculated from the accepted beam entry angle when the manually entered isocenter is in close proximity to the reference point.
11 . The method of claim 10 , wherein the patient geometry includes a body contour and a target structure, the reference point is the center of mass of the target structure, and the beam geometry (BG) parameter is a gantry angle.
12 . The method of claim 11 , further comprising: shifting the body contour and the target structure relative to the determined isocenter; and further adjusting the gantry angle based on the shift.
13 . An automated method for obtaining beam geometry (BG) parameters for treatment planning without explicit determination of an isocenter of a radiation beam delivery system delivering the treatment plan to a patient, the method comprising: determining a reference point within the target; obtaining a predicted beam entry angle relative to the reference point by executing a machine learning model trained to predict, based on patient geometry, the beam entry angle relative to the reference point; displaying the predicted beam entry angle superimposed on the patient geometry for evaluation by a user; allowing the user to accept the predicted beam entry angle or to interactively modify the predicted beam entry angle until an acceptable beam entry angle is obtained; and allowing the user to accept the reference point as the isocenter or manually enter an isocenter, wherein the accepted beam entry angle is chosen as the beam geometry (BG) parameter when the reference point is accepted as the isocenter, and the beam geometry (BG) parameter is calculated from the accepted beam entry angle when the manually entered isocenter is in close proximity to the reference point.
14 . The method of claim 13 , wherein the patient geometry includes a body contour and a target structure, the reference point is a center of mass of the target structure, and the beam geometry (BG) parameter is a gantry angle.
15 . The method of claim 14 , further comprising: shifting the body contour and the target structure relative to the determined isocenter; and further adjusting the gantry angle based on the shift.
16 . A system for developing a treatment plan for the delivery of a prescribed radiation dose to a treatment volume within a patient, comprising: a processor; and a memory coupled to the processor, the memory including instructions that when executed by the processor cause the processor to: receive information related to the prescribed radiation dose, the treatment volume, and a plurality of parameters associated with the radiation beam delivery system, the plurality of parameters including a beam geometry (BG) parameter; develop a treatment plan optimization model based on the received information, the treatment plan optimization model being configured to find an optimal dose distribution within the treatment volume; and generate an optimal treatment plan based on the treatment plan optimization model, wherein the generating of the optimal treatment plan includes: a first step configured to limit a search range for the beam geometry (BG) parameter; and a second step configured to optimize the treatment parameters, wherein the beam geometry (BG) parameter is optimized based on the limited search field obtained in the first step and without explicit determination of an isocenter of a system used for the radiation dose delivery.
17 . The system of claim 16 , wherein the first step includes one of: predicting, inputting by a user, or using a template to determine an initial choice for the beam geometry (BG) parameter and a feasible range for the beam geometry (BG) parameter and/or a largest acceptable deviation from the initial choice for the beam geometry (BG) parameter, wherein the predicting includes executing a trained machine learning model to predict, based on patient geometry as input, the initial choice for the beam geometry (BG) parameter.
18 . The system of claim 17 , wherein the second step includes: calculating dose distribution within the treatment volume; determining whether the calculated dose distribution is within an acceptable threshold of the desired dose distribution; and iteratively modifying one or more of the treatment parameters including the beam geometry (BG) parameter until the calculated dose distribution is within the acceptable threshold or an endpoint has been reached, wherein the beam geometry (BG) parameter is modified based on the search range obtained in the first step.
19 . The system of claim 18 , wherein the beam geometry (BG) parameter is modified using one of a trained reinforcement learning agent, a simplex method, a simulated annealing method, or any other deterministic, stochastic or heuristic optimization method.
20 . A system for obtaining beam geometry (BG) parameters for treatment planning without explicit determination of an isocenter of a radiation beam delivery system delivering the treatment plan to a patient, the system comprising: a processor; and a memory coupled to the processor, the memory including instructions that when executed by the processor cause the processor to: determine a reference point within the target; obtain a predicted beam entry angle relative to the reference point by executing a machine learning model trained to predict, based on patient geometry, the beam entry angle relative to the reference point; display the predicted beam entry angle superimposed on the patient geometry for evaluation by a user; allow the user to accept the predicted beam entry angle or to interactively modify the predicted beam entry angle until an acceptable beam entry angle is obtained; and allow the user to accept the reference point as the isocenter or manually enter an isocenter, wherein the accepted beam entry angle is chosen as the beam geometry (BG) parameter when the reference point is accepted as the isocenter, and the beam geometry (BG) parameter is calculated from the accepted beam entry angle when the manually entered isocenter is in close proximity to the reference point, the processor being further configured to: shift the patient geometry relative to the determined isocenter; and further adjust the gantry angle based on the shift.

Description

FIELD The present disclosure relates generally to treatment plan optimization processes and generating beam geometry (BG) solutions for treatment plans, and more particularly, to systems, methods, and devices for automated treatment plan optimization processes having all degrees of freedom including those related to beam geometry (BG) parameters, and to systems, methods, and devices for generating beam geometry (BG) solutions without having to define the isocenter of the external radiation beam delivery system. BACKGROUND Radiation therapy involves medical procedures that use external radiation beams to treat pathological anatomies (tumors, lesions, vascular malformations, nerve disorders, etc.) by delivering prescribed doses of radiation (X-rays, gamma rays, electrons, protons, and/or ions) to the pathological anatomy, while minimizing radiation exposure to the surrounding tissue and critical anatomical structures. In general, a full radiotherapy planning and treatment workflow includes several phases: a treatment planning phase, a treatment delivery phase, and a monitoring and evaluating phase in which the progress of the treatment, e.g., the dose accumulation is monitored. The treatment delivery phase generally overlaps with the monitoring and evaluation phase. Treatment is generally delivered in many sessions, which span several weeks. The patient is monitored throughout the duration of the treatment to evaluate the progress of treatment and whether there is any need to re-plan or adapt the treatment. In the treatment planning phase, first a precise three-dimensional (3D) map of the anatomical structures in the area of interest (head, body, etc.) is constructed using any one of (or combinations thereof) a computed tomography (CT), cone-beam computed tomography (CBCT), magnetic resonance imaging (MRI), positron emission tomography (PET), 3D rotational angiography (3DRA), or ultrasound techniques. This determines the exact coordinates of the target within the anatomical structure, namely, locates the tumor or abnormality within the body and defines its exact shape and size. On these images, organs at risk (OARs) in the region of interest are also delineated. This is followed by a prescription step, where the level of radiation which should be delivered to the target (tumors) and the level of protection of the OARs which needs to be achieved to avoid side-effects for the patient is specified. In the prescription step, a motion path for the radiation beam is also computed to deliver a dose distribution to the target within a treatment volume that the radiation oncologist finds acceptable, considering a variety of medical constraints. Then, a team of specialists develop a treatment plan using special computer software to optimally irradiate the tumor and minimize dose to the surrounding normal tissue by designing beams of radiation to converge on the target area from different angles and planes. In the treatment delivery phase, the radiation treatment plan is executed. During this phase, the radiation dose is delivered to the patient according to the prescribed treatment plan. Generally, a treatment plan is delivered to the patient over a series of radiation treatments referred to as fractions. There are many factors, however, such as, differences in a patient's setup position, changes that might occur if a patient's tumor regresses or if the patient loses weight during therapy, and uncertainties introduced by motion, for example, that can contribute to differences between the prescribed radiation dose distribution and the actual dose delivered (i.e., the actual dose delivered to the target during the radiation treatment). These anatomical and physiological changes can cause the target volumes and surrounding anatomical structures and organs to move and change in size and shape during the therapy. As such, executing or continuing to execute the initial treatment plan may result in an actual received dose distribution that differs from the planned distribution, and thus reduced doses to target volumes and/or increased doses to organs at risk (OARs). During the treatment delivery phase, therefore, the treatment plan may be adapted to the image of the day to better reflect the current situation. This involves making modifications to the initial treatment plan to match the new location and shape of the target volume and surrounding anatomical structures based on subsequently acquired image data. Generating an optimal treatment plan, whether it is the initial treatment plan generated during the treatment planning phase, or the adapted plan generated during the treatment delivery phase of an adaptive treatment workflow, can be time consuming and tedious, especially in the field of intensity modulated radiation therapy (IMRT) and volumetric-modulated arc therapy (VMAT) where complex matrix manipulations are required for generating and optimizing treatment plans. The complexity and duration of the treatment planning