A system for estimating a dose from a proton therapy plan includes a memory that stores machine instructions and a processor coupled to the memory that executes the machine instructions to subdivide a representation of a volume of interest in a patient anatomy traversed by a planned proton field into a plurality of voxels. The processor further executes the machine instructions to determine the distance from the source of the planned proton beam to one of the voxels. The processor also executes the machine instructions to compute the discrete contribution at the voxel to an estimated dose received by the volume of interest from the planned proton beam based on the distance between the source and the volume of interest.

Systems and methods can include obtaining computerized physician intent data representing an initial patient care plan; creating a computerized workflow to include a course of multiple radiation therapy sessions; performing instructions on the oncology computer system to generate control parameters for a radiation therapy apparatus to provide the radiation treatment in accordance with the workflow during the course of sessions; obtaining computerized treatment data after initiating the course of sessions; processing the computerized treatment data, using the processor circuit, to determine an indication of delivery or effect of the radiation treatment during the course of sessions based on the initial patient care plan relative to the workflow; using the indication of delivery or effect of the radiation treatment to adapt the patient care plan; and managing the workflow for the patient using the adapted patient care plan as the patient proceeds through a course of sessions.

A system and method for automatically generating radiation therapy treatment plans including one or more processors configured to capture geometries of organs at risk and a target volume specific to a subject, and use a shape-based algorithm to mine (152) a knowledgebase (38) of previously constructed treatment plans for similar geometries to the subject. The system and method interfaces (154) dosimetric information from a plan with a similar geometry as a patient specific starting point for a progressive tuning optimization algorithm resulting in fewer iterations. The progressive tuning algorithm (156, 158, 162) generates an optimized treatment plan. The optimized plan is evaluated against treatment goals. Trade-off plans are generated (164) create alternative plans according to unmet treatment goals.

An apparatus for determining a contour of a treatment region in a patient includes a computer processor to receive input regarding a contour of at least one organ-at-risk (OAR) adjacent to the treatment region; receive input regarding an initial contour of the treatment region; predict a radiation toxicity to the at least one OAR based on the contour of the at least one OAR, the initial contour of the treatment region, and a radiation treatment regimen; determine whether the predicted radiation toxicity exceeds a threshold; and determine a contour of the treatment region by iteratively modifying the initial contour of the treatment region, and any subsequent modified contours of the treatment region, until a stopping condition is satisfied. The stopping condition can be a preselected number of iterations or that the predicted radiation toxicity using the contour in place of the initial contour is first calculated is below said threshold.

A control circuit forms a radiation therapy treatment plan by automatically generating a base dose that references dosing information from multiple sources and then using that base dose to optimize a radiation therapy treatment plan. That radiation therapy treatment plan is then used to administer radiation therapy to a patient. That automatically generated base dose can represent any or all of earlier radiation therapy treatments for the patient, a same fraction as a dose presently being optimized per the radiation therapy treatment plan, and future planned fractions for the patient.

A control circuit operably couples to a memory having a plurality of different radiation treatment applications stored therein, a data store having patient data stored therein, and at least a first user interface comprising a first display and a first user input interface. The control circuit can be configured to present simultaneously, via the first display and for a given patient, at least two workspaces that each correspond to a different one of the radiation treatment applications wherein the workspaces are using patient data from the data store for the given patient. The control circuit can also be configured to present, via the user input interface, a bookmark capture opportunity, such that a user of the apparatus can selectively create a bookmark that captures a present state for both of the at least two workspaces and hence for the radiation treatment applications that correspond to the at least two workspaces.

Disclosed herein are systems and methods for iteratively training artificial intelligence models using reinforcement learning techniques. With each iteration, a training agent applies a random radiation therapy treatment attribute corresponding to the radiation therapy treatment attribute associated with previously performed radiation therapy treatments when an epsilon value indicative of a likelihood of exploration and exploitation training of the artificial intelligence model satisfies a threshold. When the epsilon value does not satisfy the threshold, the agent generates, using an existing policy, a first predicted radiation therapy treatment attribute, and generates, using a predefined model, a second predicted radiation therapy treatment attribute. The agent applies one of the first predicted radiation therapy treatment attribute or the second predicted radiation therapy treatment attribute that is associated with a higher reward. The agent iteratively repeats training the artificial intelligence model until the existing policy satisfies an accuracy threshold.

Systems and methods for radiation treatment planning based on a model of planning strategies knowledge including treatment states and treatment actions are disclosed. According to an aspect, a method includes receiving geometric characterization data of a target volume for radiation treatment of a patient. The method also includes receiving geometric characterization data of at least one organ at risk proximate the target volume. Further, the method includes constructing a model for applying a predetermined radiation dosage to the target volume based on the received data. The model includes treatment states and associated treatment actions selectable to implement at each state. The method includes presenting information about at least one treatment state, the treatment actions associated with the at least one treatment state, and the rewards associated with the treatment actions associated with the at least one treatment state. The method also includes reconstructing the model.

Methods for dose or treatment planning for a radiotherapy system including a radiotherapy unit are provided. A spatial dose delivered can be changed by adjusting beam shape settings, and the delivered radiation is determined using an optimization problem that steers the delivered radiation according to objectives reflecting criteria for regions of interest including at least one of: targets to be treated during treatment of the patient, organs at risk and/or healthy tissue. The method includes determining an inner set of voxels and providing a first frame description for the inner set of voxels, where the first frame description reflects criteria for the inner set of voxels. Determining an outer set of voxels encompassing the target volume and the inner set of voxels and a frame description for the outer set of voxels is provided where each reflecting criteria for the outer set of voxels. The frame descriptions are then used in the optimization problem that steers the delivered radiation.

Example methods for adaptive radiotherapy treatment planning using deep learning engines are provided. One example method may comprise obtaining treatment image data associated with a first imaging modality and planning image data associated with a second imaging modality. The treatment image data may be acquired during a treatment phase of a patient. Also, planning image data associated with a second imaging modality may be acquired prior to the treatment phase to generate a treatment plan for the patient. The method may also comprise: in response to determination that an update of the treatment plan is required, processing, using the deep learning engine, the treatment image data and the planning image data to generate output data for updating the treatment plan.