Techniques for generating radiotherapy treatment plans and establishing machine learning models for the generation and optimization of radiotherapy dose data are disclosed. An example method for generating a radiotherapy dose distribution using a generative model, trained in a generative adversarial network, includes: receiving anatomical data of a human subject that indicates a mapping of an anatomical area for radiotherapy treatment; generating radiotherapy dose data corresponding to the mapping with use of the trained generative model, as the generative model processes the anatomical data as an input and provides the dose data as output; and identifying the radiotherapy dose distribution for the radiotherapy treatment of the human subject based on the dose data. Another example method for training of the generative model includes establishing values of the generative model and a discriminative model of the generative adversarial network using adversarial training, including in a conditional generative adversarial network arrangement.

In a radiation therapy system, treatment X-rays are delivered to a target volume at the same time that imaging X-rays are also delivered to the target volume for generating image data of the target volume. That is, during an imaging interval in which imaging X-rays are delivered to the target volume, one or more pulses of treatment X-rays are also delivered to the target volume. In each pixel of an X-ray imaging device of the radiation therapy system, image signal is accumulated during portions of the imaging interval in which only imaging X-rays are delivered to the target volume and is prevented from accumulating in each pixel during the pulses of treatment X-rays.

Disclosed herein are systems and methods for identifying radiation therapy treatment data for different patients, such as field geometry. A central server collects patient data, radiation therapy treatment planning data, clinic-specific rules, and other pertinent treatment/medical data associated with a patient. The server then executes one or more machine-learning computer models to predict field geometry variables and weights associated with the patient's treatments. Using the predicted variables and weights, the server execute a clinic-specific set of logic to identify suggested field geometry, such as couch/gantry angles and/or arc attributes. The server then monitors whether end users (e.g., medical professionals) revise the suggested field geometry and trains the model accordingly.

A method of generating a treatment plan for treating a patient with radiotherapy, the method includes obtaining a plurality of sample plans, which are generated by use of a knowledge base comprising historical treatment plans and patient data. The method also includes performing a multi-criteria optimization based on the plurality of sample plans to construct a Pareto frontier, where the plurality of sample plans are evaluated with at least two objectives measuring qualities of the plurality of sample plans such that treatment plans on the constructed Pareto frontier are Pareto optimal with respect to the objectives. The method further includes identifying a treatment plan by use of the constructed Pareto frontier.

Disclosed is a computer-implemented method for determining a position of an anatomical tracking structure in a tracking image usable for controlling a radiation treatment such as at least one of radiotherapy or radio surgery of a patient, a corresponding computer program, a non-transitory program storage medium storing such a program and a computer for executing the program, as well as a system for the position of an anatomical tracking structure in a tracking image usable for controlling a radiation treatment such as at least one of radiotherapy or radio surgery of a patient, a system comprising an electronic data storage device and the aforementioned computer.

It is provided a method for providing a treatment plan for radiotherapy when the delivery is interrupted, the method being performed by a treatment planning system. The method comprises the steps of: detecting that delivery of a first treatment plan has been interrupted; obtaining an indication of delivery of a partial dose, representing the part of the first treatment plan that was delivered prior to the interruption; generating a second treatment plan, wherein the partial dose delivery forms an input to the second treatment plan generation as a background dose; and optimizing the second treatment plan while considering a plurality of scenarios.

Systems and methods are disclosed for performing operations comprising: receiving dose information representing dose delivered during a first radiotherapy treatment fraction; accessing one or more previous dose information representing dose delivered during one or more previous radiotherapy treatment fractions; computing a measure of biologically effective dose (BED) based on a combination of the dose information delivered during a first radiotherapy treatment fraction and the dose delivered during the one or more previous radiotherapy treatment fractions; and performing an isotoxic planning process for delivering a second radiotherapy treatment fraction following the first radiotherapy treatment fraction based on the computed measure of BED.

A method may include obtaining a first image related to one or more target objects generated by a first scan. The method may also include obtaining a first radiation therapy plan for treating the one or more target objects. The method may also include obtaining a second image related to the one or more target objects generated by a second scan. The second scan may be performed later than the first scan. The method may also include determining, based on the first radiation therapy plan, the first image, and the second image, a target radiation therapy plan to treat the one or more target objects. The target radiation therapy plan may be the first radiation therapy plan or a second radiation therapy plan associated with the second image, wherein at least a portion of the determining the target radiation therapy plan may be performed in parallel.

A radiation treatment session is initiated to deliver a therapeutic radiation beam from a therapeutic radiation source to a target. One or more X-ray radiation sources are caused to deliver an imaging radiation beam from the one or more X-ray radiation sources through the target to one or more X-ray detectors to acquire imaging data associated with the target during therapeutic radiation beam delivery. One or more volumetric images are constructed using the acquired imaging data.

To enable an appropriate and quick detection of a change in an internal structure of a patient, a computer program causes a computer to detect a change in an internal structure of a patient. The process includes calculating a second water equivalent thickness obtained from a second three-dimensional image being a three-dimensional image of a patient, which is newly obtained; a process of calculating a change of a first water equivalent thickness from the second water equivalent thickness, the first water equivalent thickness being obtained from a first three-dimensional image being a three-dimensional image of the patient in treatment planning; and a process of calculating a dose volume histogram change for calculating a change in a dose volume histogram from the treatment plan, based on the calculated water equivalent thickness change and correlation information indicating a correlation between a water equivalent thickness change value and dose distribution information.