CN114681813A - Automatic planning system, automatic planning method and computer program product for radiation therapy - Google Patents
Automatic planning system, automatic planning method and computer program product for radiation therapy Download PDFInfo
- Publication number
- CN114681813A CN114681813A CN202011575203.5A CN202011575203A CN114681813A CN 114681813 A CN114681813 A CN 114681813A CN 202011575203 A CN202011575203 A CN 202011575203A CN 114681813 A CN114681813 A CN 114681813A
- Authority
- CN
- China
- Prior art keywords
- information
- tps
- radiation therapy
- medical image
- neural network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
- A61N5/00—Radiation therapy
- A61N5/10—X-ray therapy; Gamma-ray therapy; Particle-irradiation therapy
- A61N5/103—Treatment planning systems
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
- A61N5/00—Radiation therapy
- A61N5/10—X-ray therapy; Gamma-ray therapy; Particle-irradiation therapy
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
- A61N5/00—Radiation therapy
- A61N5/10—X-ray therapy; Gamma-ray therapy; Particle-irradiation therapy
- A61N5/1001—X-ray therapy; Gamma-ray therapy; Particle-irradiation therapy using radiation sources introduced into or applied onto the body; brachytherapy
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
- A61N5/00—Radiation therapy
- A61N5/10—X-ray therapy; Gamma-ray therapy; Particle-irradiation therapy
- A61N5/103—Treatment planning systems
- A61N5/1039—Treatment planning systems using functional images, e.g. PET or MRI
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
- A61N5/00—Radiation therapy
- A61N5/10—X-ray therapy; Gamma-ray therapy; Particle-irradiation therapy
- A61N5/1048—Monitoring, verifying, controlling systems and methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H20/40—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to mechanical, radiation or invasive therapies, e.g. surgery, laser therapy, dialysis or acupuncture
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
- A61N5/00—Radiation therapy
- A61N5/10—X-ray therapy; Gamma-ray therapy; Particle-irradiation therapy
- A61N2005/1092—Details
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10081—Computed x-ray tomography [CT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30096—Tumor; Lesion
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Public Health (AREA)
- Radiology & Medical Imaging (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Pathology (AREA)
- Veterinary Medicine (AREA)
- Medical Informatics (AREA)
- Animal Behavior & Ethology (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Urology & Nephrology (AREA)
- Artificial Intelligence (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Surgery (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Quality & Reliability (AREA)
- Radiation-Therapy Devices (AREA)
Abstract
The invention provides a radiation therapy automatic planning system, which comprises an information input unit, a parameter generation unit, a dose prediction unit, an evaluation unit and an output unit, wherein the information input unit is used for inputting patient information comprising medical image information, contour drawing information and prescription dose information; the parameter generating unit generates initialization parameters according to the medical image information, the contour drawing information and the prescription dose information based on the first neural network model and outputs the initialization parameters to the TPS, so that the TPS can determine a radiotherapy plan by using the initialization parameters; the dose prediction unit predicts dose distribution of a radiotherapy plan generated by the TPS according to the medical image information and the contour drawing information based on the second neural network model; the evaluation unit compares the dose distribution of the radiation treatment plan determined by the TPS with the predicted dose distribution to generate an evaluation result; the output unit outputs the radiation treatment plan generated by the TPS according to the evaluation result. The present invention can improve the efficiency of designing a radiation treatment plan.
Description
Technical Field
The present invention relates to the field of radiotherapy technology, and in particular, to an automatic radiotherapy planning system, an automatic radiotherapy planning method, and a computer program product.
Background
Radiotherapy, which utilizes radiation to treat diseases, is one of the important means for tumor therapy, and has great significance for improving human health and prolonging human life.
In general, the dose distribution of a radiation treatment plan depends on the dose objectives defined by the physicist. Traditional forward treatment planning, highly dependent on the experience of the physicist, has been gradually replaced by inverse planning. The physicist submits the dose prescription to the computer, which uses dose calculation and optimization techniques to meet the set target and organ-at-risk dose requirements. To ensure the safety and reliability of the Treatment, a Treatment Planning System (TPS) is followed. In TPS, a radiation plan is made for a patient by modeling a radiation source and the patient, plan implementation is simulated, and when the radiation plan meets clinical requirements, clinical treatment is carried out. Specifically, the radiation therapy planning process includes acquiring patient target and normal tissue delineation results, planning beam and radiation types, dose calculation, plan evaluation, selecting to continue optimizing the irradiation technique or completing plan design into a plan verification step based on the evaluation results.
In order to enable the TPS to reversely calculate the beam-emitting mode of the therapy apparatus from the desired dose distribution, a physicist needs to design the number of radiation fields, an objective function and relevant weights thereof and submit the number to the TPS for optimization, and the TPS obtains the optimal beam intensity distribution of each radiation field through repeated iterative operation of a computer according to factors such as the position of a tumor in a body, tissue nonuniformity, the position of key tissues, the number of radiation fields and the like, so that the spatial dose distribution actually formed in the body is closest to the prescription dose of a doctor. After the dose distribution in the patient has been calculated from the TPS, the treatment plan is evaluated, which is also usually completely dependent on the experience of the physicist. If the requirements are not met, corresponding parameters in the treatment plan design process need to be adjusted to carry out the design, simulation plan, plan evaluation and the like of the treatment process again.
How to design appropriate parameters such as field, objective function and weight for different target region positions, shapes and dosage requirements is highly dependent on the experience of physicists. Repeating the steps until the treatment requirement is met. In order to reduce the irradiated dose of normal tissues as much as possible on the premise of ensuring the dose of the target area, a physicist needs to continuously adjust and optimize a given dose target, a great deal of manpower and time are consumed in the process, and an optimal scheme is difficult to ensure. Although the physicist may keep some templates for fine-tuning based on previous successful planning experience, the final satisfactory results are still dependent on the experience of the physicist. The primary physicist is likely to need to repeat the design optimization 5 to 7 times to obtain a planned design that meets the prescription requirements, with time varying from half an hour to hours, and may not even be able to complete the planned design at all.
In addition, due to inevitable experience and level differences, radiation treatment plans designed for the same case vary greatly in quality from hospital to hospital and from physicist to physicist, thereby potentially reducing the probability of tumor control in some patients or increasing the probability of unnecessary normal tissue complications.
The prior art still suffers from deficiencies in the complexity and time-consuming nature of radiation treatment planning.
Disclosure of Invention
The invention aims to provide an automatic radiation therapy planning system, which is used for improving the efficiency of designing a radiation therapy plan.
According to a first aspect of the present invention, there is provided a radiation therapy automatic planning system, wherein the radiation therapy automatic planning system comprises an information input unit, a parameter generation unit, a dose prediction unit, an evaluation unit and an output unit, wherein the information input unit is arranged at least for inputting patient information, the patient information comprising medical image information, contouring information and prescription dose information; the parameter generating unit is set to be capable of generating initialization parameters for the TPS according to the medical image information, the contour drawing information and the prescription dose information based on the first neural network model and outputting the initialization parameters to the TPS so that the TPS can determine a radiotherapy plan by utilizing the initialization parameters; the dose prediction unit is arranged to predict a dose distribution of a radiation therapy plan generated by the TPS for the patient from the medical image information and the contouring information based on the second neural network model; the evaluation unit is arranged to compare the dose distribution of the radiation treatment plan determined by the TPS with the dose distribution predicted by the dose prediction unit and to generate an evaluation result; and an output unit is arranged to be able to output the radiation therapy plan generated by the TPS as a final radiation therapy plan in dependence on the evaluation result.
According to an exemplary embodiment of the invention, the parameter generation unit is arranged to be able to determine an objective function for the TPS, the initialization parameter comprising objective function information representing said objective function.
According to an exemplary embodiment of the present invention, the objective function information includes a vector composed of a plurality of objective function items with weights, the objective function items being objective function items that the TPS can provide; and/or the objective function information comprises a vector formed by combining all objective function items with weights of each region to be calculated in the medical image, wherein the objective function items are objective function items which can be provided by the TPS.
According to an exemplary embodiment of the present invention, the first neural network model performs feature extraction with medical image information and contour information including image data of a plurality of successive medical image slices as input; and/or the first neural network model is arranged to be able to extract at least one of the following features: features related to density information displayed by the medical image, features related to the size of the target contour and the organ-at-risk contour, and features related to the positional relationship between the target contour and the organ-at-risk contour.
According to an exemplary embodiment of the invention, the second neural network model takes as input medical image information and contouring information of the image data comprising a plurality of successive medical image slices, the predicted dose distribution map of each medical image slice being the output; and/or in the second neural network model, in order to distinguish all the target regions from the organ-at-risk regions, the input contouring information needs to be preprocessed in the following way: pixels within the target and each of the organs at risk in the medical image are assigned unique tag values, and/or pixels of portions of the medical image where the target overlaps the organs at risk use the sum of the tag values of the overlapping target and organs at risk as the tag value.
According to an exemplary embodiment of the invention, the information input unit is further arranged for inputting portal information, wherein: the portal information comprises portal number information and/or portal direction information; and/or the parameter generation unit is arranged to be able to input the portal information into the first neural network model and for generating initialization parameters for the TPS; and/or the dose prediction unit is arranged to be able to input the portal information into the second neural network model and to be used for predicting the dose distribution of the radiation therapy plan generated by the TPS for said patient.
According to an exemplary embodiment of the invention, the output unit is arranged to be able to: outputting the radiation treatment plan generated by the TPS as a final radiation treatment plan when the evaluation result reaches a preset requirement; and outputting prompt information when the evaluation result does not meet the preset requirement, wherein the prompt information indicates that the initialization parameters for the TPS need to be adjusted to generate a new radiation treatment plan.
According to an exemplary embodiment of the invention, the first neural network model and/or the second neural network model is a convolutional neural network deep learning model; and/or the medical image information comprises CT image information of the patient; and/or the contouring information comprises contouring information of the target region and the organ at risk.
According to a second aspect of the present invention, there is provided a radiation therapy automatic planning method using the radiation therapy automatic planning system of the present invention, wherein the method includes the steps of: inputting patient information, wherein the patient information comprises medical image information, contour drawing information and prescription dose information; generating initialization parameters for the TPS according to the medical image information, the contour drawing information and the prescription dose information based on the first neural network model, and outputting the initialization parameters to the TPS so that the TPS can determine a radiation treatment plan by using the initialization parameters; predicting, based on a second neural network model, a dose distribution of a radiation therapy plan generated by the TPS for the patient according to the medical image information and the contouring information; comparing the dose distribution of the radiation treatment plan determined by the TPS with the dose distribution predicted by the dose prediction unit and generating an evaluation result; and outputting the radiation treatment plan generated by the TPS as a final radiation treatment plan when the evaluation result reaches a preset requirement.
According to a third aspect of the invention, a computer program product is provided, comprising computer program instructions, wherein the computer program instructions, when executed by one or more processors, cause the processors to perform a radiation therapy automatic planning method according to the invention.
The invention has the positive effects that: with the automatic planning system for radiation treatment and the corresponding automatic planning method and computer program product according to the invention, initialization parameters for the TPS can be automatically selected for generating a radiation treatment plan, which can also be automatically evaluated. Thus, the time required for the physicist to call up the template and repeatedly modify the initialization parameters for the TPS may be reduced, and possibly even eliminated, thereby speeding up and optimizing the process of generating a clinically useful radiation treatment plan.
Automation of the radiation treatment planning process can improve the quality of the plan while reducing the time spent, and reduce the quality of the plan variation caused by human factors. In addition, the dose distribution generated by automatic planning can be used as the starting point of manual planning, and is optimized and adjusted according to the specific requirements of the physician on the basis of the starting point.
Drawings
The principles, features and advantages of the present invention will be better understood by describing the invention in more detail below with reference to the accompanying drawings. The drawings comprise:
FIG. 1 schematically illustrates an automatic radiation treatment planning system according to an exemplary embodiment of the present invention; and
fig. 2 schematically illustrates an automatic planning method for radiation therapy according to an exemplary embodiment of the present invention.
Detailed Description
In order to make the technical problems, technical solutions and advantageous effects of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings and exemplary embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the scope of the invention.
The present invention will be described in detail below by taking a radiation therapy plan for a brain tumor as an example. However, it will be appreciated by those skilled in the art that the present application is not only applicable to the treatment of brain tumors, but also to radiation therapy programs for other conditions such as lung, prostate, esophageal, mesothelioma, head and neck, central nervous system, gynecological and gastrointestinal tumors.
Brain tumor is also called intracranial tumor, is a kind of slow disease and gradually aggravated brain disease. Brain tumors include primary brain tumors and brain metastases, and the treatment means includes complete excision by surgery, but extensive and complete excision by surgery is extremely difficult due to the unresectable nature of normal brain tissue and the wide invasive growth of malignant tumors into the cranium. With the application of Stereotactic Body Radiotherapy (SBRT), the survival time of brain tumor patients is significantly prolonged. SBRT is a generic term for a type of radiotherapy technique that focuses high-energy rays focused by multiple sources, multiple beams or multiple fields in three-dimensional space on a target area in vivo by using a stereotactic technique and a special ray device, and accurately causes the well-defined lesion tissue volume to be damaged, so that the lesion tissue is irradiated by high dose and the amount of surrounding normal tissues is reduced, thereby obtaining a high clinical effect and few adverse reactions. SBRT has the characteristics of high accuracy, high fractionated dose, high conformality and few treatment times, namely three-high and one-low. SBRT requires very precise target volume, very high demands on doctors and physicists, and a manual planning procedure is very time consuming. Although the present invention is described herein with reference to SBRT as an example, it will be understood by those skilled in the art that the present invention is not limited to SBRT, but is also applicable to other radiotherapy techniques, such as IMRT, IGRT, etc.
Currently, radiation treatment plans may be generated by the TPS 20. By the physicist giving the number of fields, the objective function and its associated weights, the TPS20 uses a specific algorithm to give a radiation treatment plan that meets the dose distribution requirements, which can be used for clinical treatment after evaluation by the radiation treating physician and physicist, and plan validation. The planning process of radiation therapy planning is very complicated and lengthy, and firstly requires a clinician to trace a tumor target area and an organ at risk on a medical image of a patient, such as a CT image, in combination with other diagnosis results, and then to give a planning design to a physicist after an upper clinician confirms the outline and dose prescription of the treatment target area and the organ at risk. In order to enable the TPS20 to reversely calculate the beam-out mode of the therapeutic machine from the desired dose distribution, a physicist is required to create an auxiliary structure in advance, arrange the treatment mode and the radiation field, design an objective function and relevant weights thereof to form an equation set, the radiation field and the objective function are adjusted according to an optimization result through repeated optimization of the TPS20 until the minimum value of the equation set is found out, each parameter value of the radiation field of the radiotherapy corresponding to the minimum value is a final required solution, and a corresponding radiotherapy plan can be made. The radiation treatment plan is available for clinical treatment after evaluation by the radiation treatment physician and physicist and plan validation.
Fig. 1 schematically illustrates an automatic radiation therapy planning system according to an exemplary embodiment of the present invention. The automatic planning system for radiotherapy comprises an information input unit 11, a parameter generation unit 12, a dose prediction unit 13, an evaluation unit 14 and an output unit 15, wherein the information input unit 11 is arranged at least for inputting patient information, the patient information comprising medical image information, contouring information and prescription dose information; the parameter generating unit 12 is configured to generate initialization parameters for the TPS20 according to the medical image information, the contour delineation information and the prescription dose information based on the first neural network model, and output the initialization parameters to the TPS20 so that the TPS20 can determine a radiation treatment plan using the initialization parameters; the dose prediction unit 13 is arranged to be able to predict a dose distribution of a radiation treatment plan generated by the TPS20 for said patient from the medical image information and the contouring information based on the second neural network model; the evaluation unit 14 is arranged to compare the dose distribution of the radiation treatment plan determined by the TPS20 with the dose distribution predicted by the dose prediction unit 13 and to generate an evaluation result; and the output unit 15 is arranged to be able to output the radiation therapy plan generated by the TPS20 as a final radiation therapy plan in dependence on the evaluation result.
With the automatic planning system for radiation treatment according to the present invention, the initialization parameters for the TPS20 can be automatically selected for generating the radiation treatment plan and also automatically evaluated according to the medical image information, the contouring information and the prescription dose information. This can improve the efficiency of designing a radiation therapy plan.
The automated radiation treatment planning system may reduce the time required for a physicist to call up templates and repeatedly modify initialization parameters for the TPS20, and may even omit this step, thereby speeding up and optimizing the process of generating a clinically useful radiation treatment plan.
By automating the radiation treatment planning optimization process, the planning quality can be improved while the time spent is reduced, and the planning quality difference caused by human factors is reduced. In addition, the dose distribution generated by automatic planning can be used as the starting point of manual planning, and is optimized and adjusted according to the specific requirements of the physician on the basis of the starting point.
Due to inevitable experience and level differences, radiation treatment plans designed for the same case vary greatly in quality from hospital to hospital and from physicist to physicist at all levels, thereby potentially reducing the probability of tumor control in some patients or increasing the probability of unnecessary normal tissue complications. This variability problem can be addressed using an automated radiation treatment planning system. The automatic radiation therapy planning system can improve the working efficiency of a physicist and help the physicist to find and improve poor plans in time. In addition, the sophisticated automatic planning model trained by the high-level hospital with the high-quality plan can be transplanted to the same type planning system of other hospitals, and the patient individualized optimized parameter reference value is provided so as to assist in improving the planning quality of the hospitals, and finally, the difference of the planning design quality between the hospitals and physicists is reduced and the patients benefit.
The first neural network model may be implemented using a deep learning convolutional neural network, for example. The first neural network model may be selected from VGG, Resnet, densenert, and the like.
By means of the first neural network model, at least one, in particular all, of the following features can be extracted from the input CT image: features related to density information displayed by the CT image, features related to the size of the target contour and the organ-at-risk contour, and features related to the positional relationship between the target contour and the organ-at-risk contour. The first neural network model may be particularly set as a 3D or 2.5D network model taking into account the spatial positional relationship of the target region to the organs at risk, and feature extraction is performed with medical image information including image data of a plurality of successive medical image slices and silhouette delineation information as input. Successive medical image slices may be spaced apart at a predetermined interval. Thereby, the first neural network model may acquire three-dimensional information contained in the medical image information. Specifically, the 3D network model performs feature extraction using the entire 3D CT image, i.e., medical image information and target region delineation results of all medical image slices including the entire scan result, as input. Better initialization parameters can be obtained through the first neural network model in 3D. In order to reduce the computational effort, a 2.5D network model can also be used, which likewise takes as input medical image information and contouring information of image data of all medical image slices comprising a plurality of successive medical image slices, but not completely comprising the entire scan result. The first neural network model can correspond to initialization parameters for the TPS20 based on the learning of these features.
To train the first neural network model, brain tumor SBRT radiation therapy planning data, e.g., SBRT radiation therapy planning data for 300 brain tumor patients, may be collected and a radiation therapy planning database constructed. The radiation therapy planning data may include patient data, patient medical images, physician prescriptions, diagnostic data, DVH (dose volume histogram), isodose curve preset dose maps, objective functions and related parameters, target volume and organ-at-risk contour maps, and the like. To ensure the quality of the data, the collected data may be screened and expert verified. The expert group may consist of not less than 5 radiotherapy physicians and physicists who have a radiotherapy experience of more than 15 years. Optionally, each case of data is carefully checked and verified by an expert group, so that all the information of each case of data is accurate and can completely meet the requirements of clinical use.
Optionally, a small sample learning is adopted based on the collected data, and the training strategy adjusts the convolutional neural network in a transfer learning manner.
Optionally, data enhancement is performed on the collected data, samples are expanded, and the network is trained to acquire the universality characteristic under small data.
In an exemplary embodiment, the parameter generation unit 12 is for example arranged to be able to determine an objective function for the TPS20, the initialization parameters comprising objective function information representing said objective function. Thus, the parameter generation unit 12 may help the physicist skip the steps of calling the template, iteratively modifying the objective function and weights, thereby speeding up and optimizing the process of generating a clinically useful radiation treatment plan.
The TPS20 may employ, for example, the Monaco planning system. The automated radiation treatment planning system has a data interface for data exchange with the TPS20, for example a Monaco planning system. The Monaco planning system may provide 10 objective function terms for generating a radiation therapy plan, including three biological value function terms (Target EUD, Serial, Parallel) and 7 physical value function terms (Target Penalty, Quadratic overlay, Quadratic Underdose, Maximum overlay, overlay DVH, Underdose DVH, Confromlity). Each objective function term has a corresponding private parameter. In order to control the number of parameters and ensure the stability of the algorithm prediction, for example, the parameters in each objective function item are optionally set to default values provided by the Monaco planning system.
Alternatively, the objective function information may include a vector combined by all weighted objective function terms of each region to be calculated in the medical image, which are the objective function terms that the TPS20 can provide. In the objective function information, all the areas to be calculated need to be arranged in a fixed order, and all the objective functions of each area to be calculated are also arranged in a fixed order. The vector can be expressed as: (w)11f11(u11),...w1nf1n(u1n),w21f21(u21),...w2nf2n(u2n),...wm1fm1(um1),...wmnfmn(umn) W represents a weight, f (u) represents an objective function term, u represents a private function of the objective function term, n represents the number of objective function terms provided by the TPS20, m represents the number of regions to be computed, w represents a weight, f (u) represents an objective function term, u represents a private function of the objective function term, m represents a number of regions to be computed, w represents a number of regions to be computed, andx1fx1(ux1),......wxnfxn(uxn) Representing the respective objective function terms with weights determined for the xth region to be calculated.
For example, a respective objective function term is set for each region to be calculated (target volume or organ), each objective function term may have a corresponding weight value and dose value. And combining all the optional target function items with weights of each area to be calculated into a vector, wherein the weight values of the unselected target functions are set to be 0. All vectors of the region to be calculated are combined into a target function vector as an output of the parameter generation unit 12.
The output layer of the first neural network model is set to be a full connection layer with the same length as the target function vector, so that the selection of the target function and the setting of parameter values such as corresponding weights can be realized.
The second neural network model is also implemented, for example, using a deep learning convolutional neural network. The second neural network model may employ some typical split network models, such as a U-net model.
With the second neural network model, a predicted dose distribution of the radiation treatment plan generated by the TPS20 for the patient may be output from the input CT images and contouring information. To this end, a second neural network model may be trained using a radiation therapy plan database.
The predicted dose distribution can be represented by a dose distribution map. The dose profile may be predicted for each pixel value.
Alternatively, the second neural network model may be set to be a 3D or 2.5D network model, for example, which takes as input CT images including a plurality of successive medical image slices, contour delineation information of the target region and the organ at risk, and as output a predicted dose distribution map for each medical image slice. In order to distinguish all target regions from the organs-at-risk regions, the input contouring information needs to be preprocessed in a manner that may include: pixels within the target and each organ at risk are assigned unique tag values, and the sum of the tag values is used as a tag in the portion where the target and organ at risk overlap, and this tag value is also unique within all tags.
The accuracy of the second neural network model of the dose prediction unit 13 can be evaluated using the mean absolute error.
After the evaluation unit 14 compares the dose distribution of the radiation treatment plan determined by the TPS20 with the dose distribution predicted by the dose prediction unit 13 and generates an evaluation result, the output unit 15 is arranged to be able to: outputting the radiation therapy plan generated by the TPS20 as a final radiation therapy plan when the evaluation result reaches a preset requirement; and outputting prompt information when the assessment result does not meet a predetermined requirement, the prompt information indicating that initialization parameters for the TPS20 need to be adjusted to generate a new radiation treatment plan.
The prompt message may be displayed, for example, by a screen. Based on the prompt information, the physicist may further adjust the initialization parameters for the TPS20 to generate a new radiation treatment plan. The evaluation unit 14 may then compare the new radiation treatment plan with the dose distribution predicted by the dose prediction unit 13 and generate an evaluation result. This process may be repeated until the evaluation results reach a predetermined requirement.
Optionally, the information input unit 11 is further configured to input the portal information including portal number information and/or portal direction information set by the physicist. The parameter generation unit 12 may be arranged to be able to input portal information into the first neural network model and for generating initialisation parameters for the TPS 20. Alternatively or additionally, the dose prediction unit 13 may be arranged to be able to input portal information into the second neural network model and to be used for predicting the dose distribution of the radiation treatment plan generated by the TPS20 for said patient.
In an alternative embodiment according to the present invention, the automatic planning system for radiation treatment may comprise only the information input unit 11, the parameter generation unit 12 and the optional output unit 15, and not the dose prediction unit 13 and the evaluation unit 14. The information input unit 11 is arranged at least for inputting patient information including medical image information, contouring information and prescription dose information; the parameter generating unit 12 is configured to generate initialization parameters for the TPS20 according to the medical image information, the contouring information and the prescription dose information based on the first neural network model, and output the initialization parameters to the TPS20 so that the TPS20 can determine a radiation therapy plan using the initialization parameters; the output unit 15 is arranged to be able to output the final radiation treatment plan. The evaluation and adjustment of the radiation treatment plan can be carried out manually without the dose prediction unit 13 and the evaluation unit 14. In this case, the efficiency of designing the radiation therapy plan can still be improved by the parameter generation unit 12.
Fig. 2 schematically illustrates a radiation therapy automatic planning method using a radiation therapy automatic planning system according to an exemplary embodiment of the present invention. The method comprises the following steps:
the method comprises the following steps: inputting patient information, wherein the patient information comprises medical image information, contour drawing information and prescription dose information;
step two: generating initialization parameters for the TPS20 according to the medical image information, the contour drawing information and the prescription dose information based on the first neural network model, and outputting the initialization parameters to the TPS20, so that the TPS20 can determine a radiotherapy plan by using the initialization parameters;
step three: predicting, based on a second neural network model, a dose distribution of a radiation treatment plan generated by the TPS20 for the patient from the medical image information and the contouring information;
step four: comparing the dose distribution of the radiation treatment plan determined by the TPS20 with the dose distribution predicted by the dose prediction unit 13 and generating an evaluation result; and
step five: based on the evaluation, the radiation treatment plan generated by the TPS20 is output as a final radiation treatment plan.
In the embodiment shown in fig. 2, the CT image, the RT structure storing the delineation information, and the number of shots and direction information set by the physicist are used as input, and the first neural network model and the second neural network model are input accordingly. The initialization parameters output from the first neural network model may include weights and corresponding proprietary parameter values for the objective function terms, and the predicted optimal dose distribution map is output from the second neural network model. The initialization parameters determined by the first neural network model are input into the Monaco planning system, so that a corresponding radiation treatment plan is generated. In step four, the dose profile of the radiation treatment plan is compared to the predicted optimal dose profile. If the dose distribution map of the radiation treatment plan is very close to the optimal dose distribution, the current radiation treatment plan is output, otherwise, the parameters for the TPS20 may be adjusted to generate the radiation treatment plan again until the evaluation result reaches the predetermined requirement.
In addition, the present invention also provides a computer program product comprising computer program instructions which, when executed by one or more processors, cause the processors to perform the automatic planning method for radiation treatment according to the present invention. The computer program instructions may be stored in a computer readable storage medium, which may include, for example, any electronic, magnetic, optical, or other physical storage device. For example, the computer-readable storage medium may be: RAM, volatile memory, non-volatile memory, flash memory, a storage drive (e.g., a hard drive), a solid state drive, any type of storage disk (e.g., an optical disk), or similar storage media, or a combination thereof.
In order to improve the efficiency of radiotherapy plan design and reduce artificially-caused plan quality differences, the invention provides an automatic radiotherapy planning system and realizes the following advantages:
acquiring the association from the image to a target function by utilizing the strong characteristic learning capability of the first neural network model in extracting medical image information such as the volumes, the overlapping relation, the relative distance and the like of the target area and the organs at risk, and realizing the automatic initialization of the radiotherapy plan;
providing a solution of radiation treatment plan initialization parameter vectorization output, constructing a convolutional neural network model, and realizing end-to-end learning training;
the second neural network model is used to predict the optimal dose distribution, which is compared to the results provided by the TPS20 for evaluation of the set planning parameters, guiding the physicist to further adjust the parameters.
Although specific embodiments of the invention have been described herein in detail, they have been presented for purposes of illustration only and are not to be construed as limiting the scope of the invention. Various substitutions, alterations, and modifications may be devised without departing from the spirit and scope of the present invention.
Claims (10)
1. A radiation therapy automatic planning system, wherein the radiation therapy automatic planning system comprises an information input unit (11), a parameter generation unit (12), a dose prediction unit (13), an evaluation unit (14), and an output unit (15),
an information input unit (11) is arranged at least for inputting patient information, the patient information comprising medical image information, contouring information and prescription dose information;
the parameter generation unit (12) is arranged to generate initialization parameters for the TPS (20) from the medical image information, the contouring information and the prescription dose information based on the first neural network model and output the initialization parameters to the TPS (20) so that the TPS (20) can determine a radiation treatment plan using the initialization parameters;
a dose prediction unit (13) is arranged to be able to predict a dose distribution of a radiation treatment plan generated by the TPS (20) for the patient from the medical image information and the contouring information based on the second neural network model;
the evaluation unit (14) is arranged to compare the dose distribution of the radiation treatment plan determined by the TPS (20) with the dose distribution predicted by the dose prediction unit (13) and to generate an evaluation result; and
an output unit (15) is arranged to be able to output the radiation therapy plan generated by the TPS (20) as a final radiation therapy plan in dependence on the evaluation result.
2. Radiation treatment automatic planning system according to claim 1, wherein the parameter generation unit (12) is arranged to be able to determine an objective function for the TPS (20), the initialization parameters comprising objective function information representing the objective function.
3. Radiation therapy automatic planning system according to claim 2, wherein the objective function information comprises a vector combined of a plurality of objective function terms with weights, which objective function terms are ones that the TPS (20) can provide; and/or
The objective function information comprises a vector combined by all weighted objective function terms of each region to be calculated in the medical image, which are the objective function terms that the TPS (20) can provide.
4. The radiation therapy automatic planning system of any one of claims 1-3, wherein the first neural network model performs feature extraction with medical image information and contouring information comprising image data of a plurality of successive medical image slices as inputs; and/or
The first neural network model is arranged to enable extraction of at least one of the following features: features related to density information displayed by the medical image, features related to the size of the target contour and the organ-at-risk contour, and features related to the positional relationship between the target contour and the organ-at-risk contour.
5. The radiation therapy automatic planning system of any one of claims 1-3,
the second neural network model takes medical image information and contour drawing information of image data comprising a plurality of continuous medical image layers as input, and takes a predicted dose distribution map of each medical image layer as output; and/or
In a second neural network model, to distinguish all target regions from the organ-at-risk regions, the input contouring information is preprocessed as follows: pixels within the target and each of the organs at risk in the medical image are assigned unique tag values, and/or pixels of portions of the medical image where the target overlaps the organs at risk use the sum of the tag values of the overlapping target and organs at risk as the tag value.
6. The radiation therapy automatic planning system according to any one of claims 1-3, wherein the information input unit (11) is further arranged for inputting portal information, wherein:
the portal information comprises portal number information and/or portal direction information; and/or
The parameter generation unit (12) is arranged to be able to input portal information into the first neural network model and for generating initialization parameters for the TPS (20); and/or
The dose prediction unit (13) is arranged to be able to input portal information into the second neural network model and to be used for predicting a dose distribution of a radiation therapy plan generated by the TPS (20) for said patient.
7. The radiation therapy automatic planning system according to any of claims 1-3, wherein the output unit (15) is arranged to be able to:
outputting the radiation therapy plan generated by the TPS (20) as a final radiation therapy plan when the evaluation result reaches a preset requirement; and
when the assessment does not meet the predetermined requirements, prompt information is output indicating that the initialization parameters for the TPS (20) need to be adjusted to generate a new radiation treatment plan.
8. The radiation therapy automatic planning system of any one of claims 1-3,
the first neural network model and/or the second neural network model is a convolutional neural network deep learning model; and/or
The medical image information comprises CT image information of the patient; and/or
The contouring information includes contouring information of the target region and the organ at risk.
9. A radiation therapy automatic planning method using the radiation therapy automatic planning system according to any one of claims 1-8, wherein the method comprises the steps of:
inputting patient information, wherein the patient information comprises medical image information, contour drawing information and prescription dose information;
generating initialization parameters for the TPS (20) from the medical image information, the contouring information and the prescription dose information based on the first neural network model and outputting the initialization parameters to the TPS (20) so that the TPS (20) can determine a radiation treatment plan using the initialization parameters;
predicting a dose distribution of a radiation treatment plan generated by the TPS (20) for the patient from the medical image information and the contouring information based on a second neural network model;
comparing the dose distribution of the radiation therapy plan determined by the TPS (20) with the dose distribution predicted by the dose prediction unit (13) and generating an evaluation result; and
when the evaluation result reaches a predetermined requirement, the radiation therapy plan generated by the TPS (20) is outputted as a final radiation therapy plan.
10. A computer program product comprising computer program instructions, wherein the computer program instructions, when executed by one or more processors, cause the processors to perform the radiation therapy automatic planning method of claim 9.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011575203.5A CN114681813B (en) | 2020-12-28 | 2020-12-28 | Automatic radiation therapy planning system, automatic radiation therapy planning method, and storage medium |
| PCT/CN2021/130360 WO2022142770A1 (en) | 2020-12-28 | 2021-11-12 | Automatic radiation treatment planning system and method, and computer program product |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011575203.5A CN114681813B (en) | 2020-12-28 | 2020-12-28 | Automatic radiation therapy planning system, automatic radiation therapy planning method, and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114681813A true CN114681813A (en) | 2022-07-01 |
| CN114681813B CN114681813B (en) | 2023-07-14 |
Family
ID=82130416
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011575203.5A Active CN114681813B (en) | 2020-12-28 | 2020-12-28 | Automatic radiation therapy planning system, automatic radiation therapy planning method, and storage medium |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN114681813B (en) |
| WO (1) | WO2022142770A1 (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115100185A (en) * | 2022-07-22 | 2022-09-23 | 深圳市联影高端医疗装备创新研究院 | Image processing method, image processing device, computer equipment and storage medium |
| CN116072263A (en) * | 2023-03-06 | 2023-05-05 | 福建自贸试验区厦门片区Manteia数据科技有限公司 | Planning parameter prediction device based on radiotherapy |
| CN116844734A (en) * | 2023-09-01 | 2023-10-03 | 福建自贸试验区厦门片区Manteia数据科技有限公司 | Method and device for generating dose prediction model, electronic equipment and storage medium |
| CN117011245A (en) * | 2023-07-11 | 2023-11-07 | 北京医智影科技有限公司 | Automatic sketching method and device for rectal cancer tumor area fusing MR information to guide CT |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115300811B (en) * | 2022-08-08 | 2024-01-05 | 中山大学肿瘤防治中心（中山大学附属肿瘤医院、中山大学肿瘤研究所） | Dose distribution determining method and device based on machine learning |
| CN115938591B (en) * | 2023-02-23 | 2023-05-09 | 福建自贸试验区厦门片区Manteia数据科技有限公司 | Dose distribution interval determining device based on radiotherapy and electronic equipment |
| CN116344001B (en) * | 2023-03-10 | 2023-10-24 | 中南大学湘雅三医院 | Medical information visual management system and method based on artificial intelligence |
| CN116612853B (en) * | 2023-07-17 | 2023-09-26 | 中国医学科学院肿瘤医院 | Radiotherapy verification plan dose generation method, radiotherapy verification plan dose generation system, electronic equipment and storage medium |
| CN116779173B (en) * | 2023-08-24 | 2023-11-24 | 北京大学第三医院（北京大学第三临床医学院） | Radiation therapy dose prediction system and method based on artificial intelligence |
| CN117323584B (en) * | 2023-10-18 | 2024-03-29 | 迈胜医疗设备有限公司 | Electronic device for radiation treatment plan adjustment, radiation treatment system and related apparatus |
| CN117717723A (en) * | 2024-02-08 | 2024-03-19 | 福建自贸试验区厦门片区Manteia数据科技有限公司 | Portal information determining device, processor and electronic equipment |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180133516A1 (en) * | 2016-11-15 | 2018-05-17 | Xi'an Cyber Medical Technology Co., Ltd. | Method and system of generating radiation treatment plan |
| CN108717866A (en) * | 2018-04-03 | 2018-10-30 | 陈辛元 | A kind of method, apparatus, equipment and the storage medium of the distribution of prediction radiotherapy planning dosage |
| CN109843377A (en) * | 2016-09-07 | 2019-06-04 | 医科达有限公司 | System and method for predicting the learning model of the radiotherapeutic treatment plan of radiation therapy dose distribution |
| CN109979564A (en) * | 2017-12-28 | 2019-07-05 | 北京连心医疗科技有限公司 | A kind of intelligence radiotherapy planning method, equipment and storage medium |
| CN110197709A (en) * | 2019-05-29 | 2019-09-03 | 广州瑞多思医疗科技有限公司 | A kind of 3-dimensional dose prediction technique based on deep learning Yu priori plan |
| US20200075148A1 (en) * | 2018-08-31 | 2020-03-05 | The Board Of Regents Of The University Of Texas System | Deep learning based dosed prediction for treatment planning and quality assurance in radiation therapy |
| CN111128340A (en) * | 2019-12-25 | 2020-05-08 | 上海联影医疗科技有限公司 | Radiotherapy plan generation device, radiotherapy plan generation apparatus, and storage medium |
| CN111432879A (en) * | 2017-12-08 | 2020-07-17 | 医科达有限公司 | Determining bundle model parameters using deep convolutional neural networks |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101968671B1 (en) * | 2018-01-17 | 2019-04-12 | 사회복지법인 삼성생명공익재단 | Intelligent automated method and system for setting radiation therapy plan |
-
2020
- 2020-12-28 CN CN202011575203.5A patent/CN114681813B/en active Active
-
2021
- 2021-11-12 WO PCT/CN2021/130360 patent/WO2022142770A1/en active Application Filing
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109843377A (en) * | 2016-09-07 | 2019-06-04 | 医科达有限公司 | System and method for predicting the learning model of the radiotherapeutic treatment plan of radiation therapy dose distribution |
| US20180133516A1 (en) * | 2016-11-15 | 2018-05-17 | Xi'an Cyber Medical Technology Co., Ltd. | Method and system of generating radiation treatment plan |
| CN111432879A (en) * | 2017-12-08 | 2020-07-17 | 医科达有限公司 | Determining bundle model parameters using deep convolutional neural networks |
| CN109979564A (en) * | 2017-12-28 | 2019-07-05 | 北京连心医疗科技有限公司 | A kind of intelligence radiotherapy planning method, equipment and storage medium |
| CN108717866A (en) * | 2018-04-03 | 2018-10-30 | 陈辛元 | A kind of method, apparatus, equipment and the storage medium of the distribution of prediction radiotherapy planning dosage |
| US20200075148A1 (en) * | 2018-08-31 | 2020-03-05 | The Board Of Regents Of The University Of Texas System | Deep learning based dosed prediction for treatment planning and quality assurance in radiation therapy |
| CN110197709A (en) * | 2019-05-29 | 2019-09-03 | 广州瑞多思医疗科技有限公司 | A kind of 3-dimensional dose prediction technique based on deep learning Yu priori plan |
| CN111128340A (en) * | 2019-12-25 | 2020-05-08 | 上海联影医疗科技有限公司 | Radiotherapy plan generation device, radiotherapy plan generation apparatus, and storage medium |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115100185A (en) * | 2022-07-22 | 2022-09-23 | 深圳市联影高端医疗装备创新研究院 | Image processing method, image processing device, computer equipment and storage medium |
| CN116072263A (en) * | 2023-03-06 | 2023-05-05 | 福建自贸试验区厦门片区Manteia数据科技有限公司 | Planning parameter prediction device based on radiotherapy |
| CN117011245A (en) * | 2023-07-11 | 2023-11-07 | 北京医智影科技有限公司 | Automatic sketching method and device for rectal cancer tumor area fusing MR information to guide CT |
| CN117011245B (en) * | 2023-07-11 | 2024-03-26 | 北京医智影科技有限公司 | Automatic sketching method and device for rectal cancer tumor area fusing MR information to guide CT |
| CN116844734A (en) * | 2023-09-01 | 2023-10-03 | 福建自贸试验区厦门片区Manteia数据科技有限公司 | Method and device for generating dose prediction model, electronic equipment and storage medium |
| CN116844734B (en) * | 2023-09-01 | 2024-01-16 | 福建自贸试验区厦门片区Manteia数据科技有限公司 | Method and device for generating dose prediction model, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114681813B (en) | 2023-07-14 |
| WO2022142770A1 (en) | 2022-07-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114681813B (en) | Automatic radiation therapy planning system, automatic radiation therapy planning method, and storage medium | |
| CN109069858B (en) | Radiotherapy system and computer readable storage device | |
| CN107072624B (en) | System and method for automated treatment planning | |
| WO2021048143A1 (en) | Systems and methods for automatic treatment planning and optimization | |
| US9782607B2 (en) | Systems and methods for radiation treatment planning using combined imaging and treatment dose | |
| CN108815721B (en) | Irradiation dose determination method and system | |
| EP3436148B1 (en) | Adaptive radiation therapy planning | |
| CN112770811A (en) | Method and system for radiation therapy treatment planning using a deep learning engine | |
| CN114025837A (en) | FLASH therapy treatment planning and oncology information system with dose rate prescription and dose rate mapping | |
| US11986672B2 (en) | Radiation treatment planning for delivering high dose rates to spots in a target | |
| JP2009545394A (en) | Biologically guided adaptive treatment planning | |
| US10342994B2 (en) | Methods and systems for generating dose estimation models for radiotherapy treatment planning | |
| EP4126214B1 (en) | Automatically-planned radiation-based treatment | |
| US11478210B2 (en) | Automatically-registered patient fixation device images | |
| CN115666717A (en) | Method and apparatus for deriving and using virtual volume structures to predict potential collisions in delivering therapeutic radiation | |
| CN110706779B (en) | Automatic generation method of accurate target function of radiotherapy plan | |
| CN113941100A (en) | Method and apparatus for generating deliverable radiotherapy plan according to three-dimensional spatial dose distribution | |
| Wang | Automated Generation of Radiotherapy Treatment Plans Using Machine Learning Methods | |
| Hajinezhad et al. | Automated Learning of Radiation Therapy Dose Distribution for Head and Neck Cancer | |
| Suvira | Dosimetric differences between scheduled and adapted plans generated from ethos adaptive radiotherapy for patients with prostate cancer | |
| Niebuhr | Biological Dose Accumulation in Image-guided Radiotherapy | |
| Thongphiew | ON-LINE RE-OPTIMIZATION OF PROSTATE IMRT PLAN FOR ADAPTIVE RADIATION THERAPY–A FEASIBILITY STUDY AND IMPLEMENTATION | |
| CN117282040A (en) | Intelligent offline self-adaptive radiotherapy method and system based on time sequence prediction model | |
| Sheng | Dose-guided automatic IMRT planning: a feasibility study |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |