Abstract: 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.

Abstract: These teachings facilitate the administration of therapeutic radiation to a heterogeneous patient volume using a radiation beam source. More particularly, these teachings provide for determining a cross-sectional size of a radiation beam as corresponds to that radiation beam source and also for determining density information corresponding to the aforementioned heterogeneous body. These teachings then provide for generating a three-dimensional radiation dose calculation for the heterogeneous body using a control circuit configured as a convolution/superposition based dose calculator using a three-dimensional energy-spreading kernel. By one approach, these teachings provide for the calculator scaling total energy released per mass as a function of the cross-sectional size and energy of the radiation beam and the aforementioned density information.

Abstract: Systems and methods for augmenting a training data set with annotated pseudo images for training machine learning models. The pseudo images are generated from corresponding images of the training data set and provide a realistic model of the interaction of image generating signals with the patient, while also providing a realistic patient model. The pseudo images are of a target imaging modality, which is different than the imaging modality of the training data set, and are generated using algorithms that account for artifacts of the target imaging modality. The pseudo images may include therein the contours and/or features of the anatomical structures contained in corresponding medical images of the training data set. The trained models can be used to generate contours in medical images of a patient of the target imaging modality or to predict an anatomical condition that may be indicative of a disease.

Abstract: 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.

Abstract: A computing system comprising a central processing unit (CPU), and memory coupled to the CPU and having stored therein instructions that, when executed by the computing system, cause the computing system to execute operations to generate a radiation treatment plan. The operations include accessing a minimum prescribed dose to be delivered into and across the target, determining a number of beams and directions of the beams, and determining a beam energy for each of the beams, wherein the number of beams, the directions of the beams, and the beam energy for each of the beams are determined such that the entire target receives the minimum prescribed dose. The operations further include prescribing a dose rate and optimizing dose rate constraints for FLASH therapy, and displaying a dose rate map of the FLASH therapy.

Abstract: A computer-implemented method of performing a treatment fraction of radiation therapy comprises: determining a current position of a target volume of patient anatomy; based on the current position of the target volume, computing an accumulated dose for non-target tissue proximate the target volume; determining that the accumulated dose is less than a current value for a dose budget of the non-target tissue; and in response to the accumulated dose being less than the current value for the dose budget, applying a treatment beam to the target volume while the target volume is in the current position.

Abstract: Nominal values of parameters, and perturbations of the nominal values, that are associated with previously defined radiation treatment plans are accessed. For each treatment field of the treatment plans, a field-specific planning target volume (fsPTV) is determined based on those perturbations. At least one clinical target volume (CTV) and at least one organ-at-risk (OAR) volume are also delineated. Each OAR includes at least one sub-volume that is delineated based on spatial relationships between each OAR and the CTV and the fsPTV for each treatment field. Dose distributions for the sub-volumes are determined based on the nominal values and the perturbations. One or more dose prediction models are generated for each sub-volume. The dose prediction model(s) are trained using the dose distributions.

Abstract: Embodiments of the present disclosure are directed to methods and computer systems for converting datasets into three-dimensional (“3D”) mesh surface visualization, displaying the mesh surface on a computer display, comparing two three-dimensional mesh surface structures by blending two primary different primary colors to create a secondary color, and computing the distance between two three-dimensional mesh surface structures converted from two closely-matched datasets. For qualitative analysis, the system includes a three-dimensional structure comparison control engine that is configured to convert dataset with three-dimensional structure into three-dimensional surfaces with mesh surface visualization. The control engine is also configured to assign color and translucency value to the three-dimensional surface for the user to do qualitative comparison analysis. For quantitative analysis, the control engine is configured to compute the distance field between two closely-matched datasets.

Abstract: A method for implementing a radiation therapy knowledge exchange starts with searching a database of cases studies and selecting a case study. The selected case study is downloaded. The downloaded case study is applied to a medical case, wherein the downloaded case is applied using deformable image registration to deform reference images of the downloaded case to medical images of the medical case. After application of the downloaded case study, the medical case is uploaded to the network, wherein uploading the medical case allows at least the submitting clinician to download, review, and edit at least a portion of the medical case to create a reviewed medical case. Finally, the reviewed medical case is downloaded and applied to the medical case to create a final medical case.

Abstract: Systems and methods for radiation treatment planning can include a computing system determining an estimate of a radiation dose distribution within an anatomical region of a patient, and determining a cost matrix representing an objective function, using the estimate of the radiation dose distribution. The computing system can project the cost matrix on each of a plurality of fluence planes. Each of the plurality of fluence planes can be associated with a corresponding gantry-couch orientation of a plurality of gantry-couch orientations of a medical linear accelerator. The computing system can determine, using projections of the cost matrix on each of the plurality of fluence planes, a sequence of gantry-couch orientations among the plurality of gantry-couch orientations representing a treatment path.

Abstract: A three-dimensional model for a patient fixation device that serves to immobilize at least a portion of a particular patient when capturing CT image information of that patient is accessed and then registered with the pixels that correspond to the patient fixation device in the CT image. The model can specify rules of movement for each of a plurality of structural elements that comprise the patient fixation device and that are capable of movement relative to one another. By one approach the aforementioned registration occurs on a part-by-part basis for each of the structural elements. Following registration, the CT image can be automatically segmented.

Abstract: Deep learning approaches automatically segment at least some breast tissue images while non-deep learning approaches automatically segment organs-at-risk. Both three-dimensional CT imaging information and two-dimensional orthogonal topogram imaging information can be used to determine virtual-skin volume. The foregoing imaging information can also serve to automatically determine a body outline for at least a portion of the patient. That body outline, along with the virtual-skin volume and registration information can serve as inputs to automatically calculate radiation treatment platform trajectories, collision detection information, and virtual dry run information of treatment delivery per the optimized radiation treatment plan.

Abstract: An imager includes: an array of imager elements configured to generate image signals based on radiation received by the imager; and circuit configured to perform readout of image signals, wherein the circuit is configured to be radiation hard. An imager includes: an array of imager elements configured to generate image signals based on the radiation received by the imager; and readout and control circuit coupled to the array of imager elements, wherein the readout and control circuit is configured to perform signal readout in synchronization with an operation of a treatment beam source.

Abstract: A method of imaging a region of patient anatomy having a target volume includes performing an autosegmentation of a high-contrast portion of a first reconstructed volume of the region to generate a three-dimensional (3D) representation of the high-contrast portion disposed within the region and generating a set of two-dimensional (2D) mask projections of the region by performing a forward projection process on the 3D representation, wherein each 2D mask projection in the set of 2D mask projections includes location information indicating pixels that are blocked by the high-contrast portion during the forward projection process performed on the 3D representation.

Abstract: New techniques are described herein for providing a user-friendly interface for adjusting dose distribution values during optimization of a radiation application plan. In an embodiment, a graphical user interface is provided that provides an image of a target area for a radiation application, and a graphical overlay of a dose distribution disposed over the target area that visually represents the optimized dose distribution according to the input dose parameters. In one or more embodiments, the dose distribution may be automatically calculated from input parameters supplied by a user through the graphical user interface prior to optimization. A user (such as a clinician, radiation oncologist, or radiation therapy operator, etc.) is able to modify the visualization of the dose distribution volume during optimization via a user input device in conjunction with the graphical user interface and have the modification adjusted in real-time.

Abstract: A method of determining treatment geometries for a radiotherapy treatment includes providing a patient model having one or more regions of interest (ROIs); defining a delivery coordinate space (DCS); for each beam's eye view (BEV) plane of each vertex in the DCS, and for each ROI, evaluating a dose of the ROI using transport solutions; evaluating a BEV scores of each pixel of the BEV plane using the doses of the one or more ROIs; determining one or more BEV regions in the BEV plane based on the BEV scores; determining a BEV region connectivity manifold based on the BEV regions; determining a set of treatment trajectories based on the BEV region connectivity manifold; and determining one or more IMRT fields. Each treatment trajectory defines a path through a set of vertices in the DCS. Each IMRT field defines a direction of incidence corresponding to a vertex in the DCS.

Abstract: Solutions are provided herein that specifically accounts for biological effects of tissue during radiation planning (such as treatment planning). In one or more embodiments, the biological effects may be calculated by accessing a knowledge base to determine reference data comprising at least one biological characteristic corresponding to the at least one organ, predicting a biological effect for the plurality of identified structures based on the biological characteristic corresponding to the at least one organ, and generating or modifying a radiation plan based on the biological effect. By incorporating biological data and fraction dose information, dose-estimation models can be created and trained to more accurately estimate dose absorption and effectiveness. Moreover, existing estimation models may be adapted to create dose estimations that account for the biological efficiency of target structures.

Abstract: Systems and methods for cloud-based scalable segmentation model training solutions including a computing interface by which a client/user/customer can upload and store training data in a storage device of a cloud-based network, provide access to the training data stored in the storage device, initiate a request for training a segmentation model, monitor the training of the segmentation model, and download the trained segmentation model, and a computing system operatively coupled with a client device through the computing interface and configured to pre-process the training data using a first set of computing resources of the cloud-based network, store the processed training data in a storage device of the cloud-based network, deploy, upon a training request from the client device, a training application on a second set of computing resources of the cloud-based network to train the segmentation model based on the processed training data, provide access to the client device to monitor the training, and provide

Abstract: A control circuit accesses topograms of a patient that include patient content that is beyond the portion of the patient that appears in the three-dimensional computed tomography (CT) images for that patient. The control circuit uses those topograms to derive a virtual volumetric structure representing at least some of the patient content that is beyond the aforementioned portion of the patient that appears in the 3D CT images. That virtual volumetric structure can then be used to predict potential collisions when assessing a radiation treatment plan for the patient that utilizes the aforementioned radiation treatment platform. By one approach the topograms include at least two substantially orthographic views of the aforementioned patient content.

Abstract: Example methods and systems for tomographic data analysis are provided. One example method may comprise: obtaining first three-dimensional (3D) feature volume data and processing the first 3D feature volume data using an AI engine that includes multiple first processing layers, an interposing forward-projection module and multiple second processing layers. Example processing using the AI engine may involve: generating second 3D feature volume data by processing the first 3D feature volume data using the multiple first processing layers, transforming the second 3D volume data into 2D feature data using the forward-projection module and generating analysis output data by processing the 2D feature data using the multiple second processing layers.