KR102275098B1 - System and method for predicting of intensity modulated treatment plan - Google Patents

Abstract

An embodiment of the present invention is an intensity-modulated radiation treatment planning system including one or more memories and a processor, wherein the intensity-modulated radiation treatment planning system includes patient's medical image information, manually established treatment plan fluence map information, and and a receiving unit for receiving dose distribution information, wherein the processor includes an organ division information generation unit for generating organ division information from the medical image information of the patient, the generated organ division information, and the manually established treatment plan fluence It provides an intensity-modulated radiation treatment plan prediction system, including a learning unit for generating a fluence map prediction model of a multi-leaf collimator by machine learning based on map information and the dose distribution information.

Description

SYSTEM AND METHOD FOR PREDICTING OF INTENSITY MODULATED TREATMENT PLAN

Embodiments of the present invention relate to a system and method for predicting an intensity modulated radiation treatment plan.

Radiation therapy is a method of delaying, inhibiting, or even annihilating the growth of malignant tissue by damaging or destroying a target tissue using high-energy waves such as X-rays and gamma rays or high-energy particles such as electron beams and proton beams. Radiation therapy is used not only for cancer, but also for the treatment of benign tumors, medical diseases, and some skin diseases. Recently, a radiosurgery method has been developed in which a large amount of radiation is irradiated and treated at once without an incisional operation, replacing the neurosurgical method of incision of the skull.

Recently, it has become so common that more than 60% of cancer patients receive radiation therapy. Radiation therapy is not only used to treat tumors on its own, but it is also used in conjunction with other surgical procedures to treat localized areas that cannot be removed by surgery, because the tumors are large and invasive, and thus reduce the size of the tumor. It can be used to make surgery easier or to destroy malignant cells left after surgery.

As such, before irradiating radiation to the patient's tumor tissue, a precise radiation treatment plan is required, such as determining the radiation location and radiation dose in consideration of the size and location of the tumor existing in the patient's current body tissue.

For such precise radiation treatment planning, Intensity Modulated Radiation Therapy (IMRT) programming method has been proposed. Intensity-controlled radiation therapy plan is attracting attention as a new radiation therapy that can protect normal organs while concentrating the planned dose on the lesion compared to the existing 3D conformal radiation therapy (3D-CRT). Advanced intensity control radiation therapy planning is possible through control of the shape of the irradiation area and change of the radiation intensity within the irradiation area, which is made by combining the arrangement of a complex multi-leaf collimator (MLC).

Conventionally, to determine the combination of the arrangement of multi-leaf collimators, after setting the location of the tumor using a tomography image, the optimal radiation treatment is performed by adjusting the location of radiation and the amount of radiation based on the experience and physical knowledge of the dose designer. A plan was drawn up. However, in this method, the quality of the treatment plan results depends on the capacity and experience of the dose designer and the clinical staff participating in the radiation treatment plan, and the treatment planning time and work increase. In particular, in the case of a high-level treatment plan with a complex tumor shape and close proximity to the surrounding normal organs, it is difficult to obtain an excellent treatment plan that lowers the dose to the surrounding normal organs and irradiates the tumor with high dose.

In order to solve the above problems, the present invention provides a system and method for intensity-modulated radiation treatment planning that can generate fluence map information of a multi-leaf collimator using machine learning.

An embodiment of the present invention is an intensity-modulated radiation treatment planning system including one or more memories and a processor, wherein the intensity-modulated radiation treatment planning system includes patient's medical image information, manually established treatment plan fluence map information, and and a receiving unit for receiving dose distribution information, wherein the processor includes an organ division information generation unit for generating organ division information from the medical image information of the patient, the generated organ division information, and the manually established treatment plan fluence It provides an intensity-modulated radiation treatment plan prediction system, including a learning unit for generating a fluence map prediction model of a multi-leaf collimator by machine learning based on map information and the dose distribution information.

In one embodiment of the present invention, the organ segmentation information generating unit derives a three-dimensional organ image from the medical image information of the patient, and divides the three-dimensional organ image in a plane according to a distance from a radiation source to obtain the organ segmentation information can create

In one embodiment of the present invention, the patient's medical image information is a two-dimensional organ image taken along a first direction of the target organ of the patient, and the organ segmentation information is the target organ of the patient in the first direction. It may be a two-dimensional long-term image divided along a second direction perpendicular to .

In one embodiment of the present invention, the patient's medical image information is magnetic resonance imaging (MRI), computed tomography (CT) image, positron emission tomography (PET) image, functional MRI (fMRI) image, fluoroscopy image, It may include image information of at least one of an ultrasound image and a single photon emission tomography (SPECT) image.

In one embodiment of the present invention, the organ division information generating unit generates a plurality of two-dimensional images by plane-segmenting the three-dimensional organ image, and sets unit areas of the plurality of two-dimensional images at a distance from the radiation source. It is possible to generate the long-term division information by correcting it.

In an embodiment of the present invention, the organ segmentation information generating unit is configured to correct the unit area of the plurality of 2D images with respect to the distance from the radiation source based on the manually established treatment plan fluence map information. Long-term segmentation information can be created.

In an embodiment of the present invention, the processor further includes a radiation treatment plan derivation unit for obtaining step-by-step position data of the multi-leaf collimator using the fluence map prediction model, organ division information, and dose distribution information of the multi-leaf collimator. may include

In an embodiment of the present invention, the radiation treatment plan derivation unit uses the fluence map prediction model of the multi-leaf collimator, organ division information, and dose distribution information to further add radiation intensity data corresponding to each step of the multi-leaf collimator. can be obtained

Another embodiment of the present invention includes the steps of receiving medical image information of a patient, treatment plan fluence map information and dose distribution information established manually, generating organ segmentation information from the received medical image information of the patient; Generating a fluence map prediction model of the multi-leaf collimator by machine learning based on the generated organ segmentation information, the manually established treatment plan fluence map information, and the dose distribution information, and predicting the fluence map of the multi-leaf collimator It provides a method for predicting an intensity-modulated radiation treatment plan, including deriving a radiation treatment plan including step-by-step position data of the multi-leaf collimator using a model, organ segmentation information, and dose distribution information.

In one embodiment of the present invention, the step of deriving the radiation treatment plan may include the radiation intensity data corresponding to each step of the multi-leaf collimator to derive the radiation treatment plan.

In one embodiment of the present invention, the generating of the organ segmentation information comprises deriving a three-dimensional organ image from the medical image information of the patient and dividing the three-dimensional organ image by plane according to the distance from the radiation source. It may include generating the long-term division information.

In one embodiment of the present invention, the patient's medical image information is a two-dimensional organ image taken along a first direction of the target organ of the patient, and the organ segmentation information is the target organ of the patient in the first direction. It may be a two-dimensional long-term image divided along a second direction perpendicular to .

In one embodiment of the present invention, the patient's medical image information is magnetic resonance imaging (MRI), computed tomography (CT) image, positron emission tomography (PET) image, functional MRI (fMRI) image, fluoroscopy image, It may include image information of at least one of an ultrasound image and a single photon emission tomography (SPECT) image.

In an embodiment of the present invention, the generating of the organ segmentation information includes generating a plurality of two-dimensional images by plane segmenting the three-dimensional organ image, and calculating unit areas of the plurality of two-dimensional images as the radiation source. It may include generating the long-term segmentation information by correcting the distance from the .

In an embodiment of the present invention, the generating of the organ segmentation information comprises calculating the unit area of the plurality of two-dimensional images based on the manually established treatment plan fluence map information with respect to the distance from the radiation source. The long-term segmentation information may be generated by correcting it.

An embodiment of the present invention provides a computer program stored in a medium for executing any one of the methods described above using a computer.

Other aspects, features and advantages other than those described above will become apparent from the following drawings, claims, and detailed description of the invention.

Intensity-modulated radiation treatment plan prediction system and method according to the present invention by generating long-term segmentation information, which is learning data, using medical image information of a patient, and training using this as learning data, driving data of a multi-leaf collimator as a radiation treatment plan can be printed automatically. Through this, an ideal fluence map can be obtained while omitting the optimization process repeatedly performed by the treatment planner for the existing radiation treatment plan, and the quality of the treatment plan can be maintained while minimizing the time and effort for establishing a treatment plan.

1 is a diagram illustrating an example of a network environment according to an embodiment of the present invention.
2 is a block diagram illustrating an internal configuration of a server according to an embodiment of the present invention.
3 is a time-series diagram illustrating a method for predicting an intensity-modulated radiation treatment plan according to an embodiment of the present invention.
4 is a conceptual diagram schematically illustrating a method for predicting an intensity-modulated radiation treatment plan according to an embodiment of the present invention.
5 is a conceptual diagram for explaining a driving method of the multi-leaf collimator.
6 is a flowchart illustrating a process for generating long-term division information in a time-series information generating unit.
7 to 9 are diagrams schematically illustrating a process for generating the organ division information of FIG. 6 .
10 is a diagram illustrating an example of data obtained after reconstructing a three-dimensional organ image with organ segmentation information.
11 is a diagram comparing a fluence map (clinical) obtained from a manually generated treatment plan with a fluence map (Generated) obtained through an embodiment of the present invention.

Since the present invention can apply various transformations and can have various embodiments, specific embodiments are illustrated in the drawings and described in detail in the detailed description. Effects and features of the present invention, and a method of achieving them, will become apparent with reference to the embodiments described below in detail in conjunction with the drawings. However, the present invention is not limited to the embodiments disclosed below and may be implemented in various forms.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings, and when described with reference to the drawings, the same or corresponding components are given the same reference numerals, and the overlapping description thereof will be omitted. .

In the following embodiments, terms such as first, second, etc. are used for the purpose of distinguishing one component from another, not in a limiting sense.

In the following examples, the singular expression includes the plural expression unless the context clearly dictates otherwise.

In the following embodiments, terms such as include or have means that the features or components described in the specification are present, and the possibility of adding one or more other features or components is not excluded in advance.

In the following embodiments, when it is said that a part such as a film, region, or component is on or on another part, it is not only when it is directly on the other part, but also another film, region, component, etc. is interposed therebetween. Including cases where there is

In the drawings, the size of the components may be exaggerated or reduced for convenience of description. For example, since the size and thickness of each component shown in the drawings are arbitrarily indicated for convenience of description, the present invention is not necessarily limited to the illustrated bar.

Where certain embodiments are otherwise feasible, a specific process sequence may be performed different from the described sequence. For example, two processes described in succession may be performed substantially simultaneously, or may be performed in an order opposite to the order described.

In the following embodiments, when a film, region, or component is connected, other films, regions, and components are interposed between the films, regions, and components as well as when the films, regions, and components are directly connected. It also includes cases where it is indirectly connected. For example, in this specification, when it is said that a film, a region, a component, etc. are electrically connected, not only the case where the film, a region, a component, etc. are directly electrically connected, but also other films, regions, and components are interposed therebetween. Indirect electrical connection is also included.

1 is a diagram illustrating an example of a network environment according to an embodiment of the present invention.

The network environment of FIG. 1 shows an example including a plurality of user terminals 110 , 120 , 130 , 140 , a server 150 , and a network 160 . 1 is an example for the description of the invention, and the number of user terminals or the number of servers is not limited as in FIG. 1 .

The plurality of user terminals 110 , 120 , 130 , and 140 may be a fixed terminal implemented as a computer device or a mobile terminal. The plurality of user terminals 110 , 120 , 130 , and 140 may be terminals of an administrator who controls the server 150 . Examples of the plurality of user terminals 110 , 120 , 130 , 140 include a smart phone, a mobile phone, a navigation device, a computer, a notebook computer, a digital broadcasting terminal, a personal digital assistant (PDA), and a portable multimedia player (PMP). ), and tablet PCs. For example, the user terminal 1110 may communicate with other user terminals 120 , 130 , 140 and/or the server 150 through the network 160 using a wireless or wired communication method.

The communication method is not limited, and not only a communication method using a communication network (eg, a mobile communication network, a wired Internet, a wireless Internet, a broadcasting network) that the network 160 may include, but also short-range wireless communication between devices may be included. For example, the network 160 may include a personal area network (PAN), a local area network (LAN), a campus area network (CAN), a metropolitan area network (MAN), a wide area network (WAN), and a broadband network (BBN). , the Internet, and the like. In addition, the network 160 may include any one or more of a network topology including a bus network, a star network, a ring network, a mesh network, a star-bus network, a tree or a hierarchical network, etc. not limited

The server 150 is a computer device or a plurality of computer devices that communicates with a plurality of user terminals 110 , 120 , 130 , 140 and the network 160 to provide commands, codes, files, contents, services, etc. can be implemented.

For example, the server 150 may provide a file for installing an application to the user terminal 1110 connected through the network 160 . In this case, the user terminal 1110 may install the application using the file provided from the server 150 . In addition, the service provided by the server 150 by accessing the server 150 under the control of the operating system (OS) and at least one program (eg, a browser or an installed application) included in the user terminal 1110 content can be provided. As another example, the server 150 may establish a communication session for data transmission/reception, and route data transmission/reception between the plurality of user terminals 110 , 120 , 130 , and 140 through the established communication session.

According to an embodiment of the present invention, the server 150 may automatically establish a radiation treatment plan based on deep learning. At this time, the server 150 sets the dose distribution including information on the maximum and minimum dose of a specific organ and the dose tolerance in the partial volume required by the treatment planner, and determines the location information of the multi-leaf collimator or the intensity of the irradiation beam according to this. Radiation treatment planning can be performed by inverse planning. The above-mentioned radiation treatment plan uses a multileaf collimator (MLC) to divide the irradiated surfaces (segments) each having a uniform intensity, and fixed intensity modulation to treat each divided irradiated surface in the form of an independent irradiation beam. It may be performed by a static IMRT, or by a dynamic intensity modulation radiation treatment plan (dynamic IMRT) implemented by changing the radiation dose rate and the speed of the multi-leaf collimator.

A system and method for predicting a radiation treatment plan using the server 150 of the present invention is a multi-leaf collimator based on deep learning by analyzing the patient's medical image information, the manually established treatment plan fluence map information, and the dose distribution information. It is characterized by predicting the fluence map of

2 is a block diagram illustrating an internal configuration of a server according to an embodiment of the present invention.

In FIG. 2 , the internal configuration of the server 150 will be described. The server 150 may include a memory 221 , a processor 222 , a receiver 223 , and an input/output interface 224 . The memory 221 is a computer-readable recording medium and may include a random access memory (RAM), a read only memory (ROM), and a permanent mass storage device such as a disk drive. Also, the memory 221 may store an operating system and at least one program code (eg, a code for a browser installed and driven in the user terminal 1110 and the above-described application). These software components may be loaded from a computer-readable recording medium separate from the memory 221 using a drive mechanism. The separate computer-readable recording medium may include a computer-readable recording medium such as a floppy drive, a disk, a tape, a DVD/CD-ROM drive, and a memory card.

In another embodiment, the software components may be loaded into the memory 221 through the receiving unit 223 instead of a computer-readable recording medium. For example, the at least one program is a program installed by files provided through the network 160 by a file distribution system (eg, the above-described server 150 ) for distributing installation files of developers or applications (eg, a program) may be loaded into the memory 221 based on the above-described application).

The processor 222 may be configured to process instructions of a computer program by performing basic arithmetic, logic, and input/output operations. Here, the 'processor' may refer to a data processing device embedded in hardware, for example, having a physically structured circuit to perform a function expressed as a code or an instruction included in a program. As an example of the data processing device embedded in the hardware as described above, a microprocessor, a central processing unit (CPU), a processor core, a multiprocessor, an application-specific integrated (ASIC) Circuit) and a processing device such as an FPGA (Field Programmable Gate Array) may be included, but the scope of the present invention is not limited thereto. The command may be provided to the processor 222 by the memory 221 or the receiving unit 223 . For example, the processor 222 may be configured to execute a received instruction according to program code stored in a recording device such as the memory 221 . The processor 222 may include an organ division information generating unit 111 , a learning unit 112 , and a radiation treatment plan deriving unit 113 .

The receiver 223 may provide a function for the user terminal 1110 and the server 150 to communicate with each other through the network 160 , and may be another user terminal (eg, user terminal 2 120 ). . Alternatively, it may provide a function for communicating with another server. For example, a request generated by the processor of the user terminal 1110 according to a program code stored in a recording device such as a memory may be transmitted to the server 150 through the network 160 under the control of the communication module. Conversely, a control signal, command, content, file, etc. provided under the control of the processor 222 of the server 150 passes through the receiver 223 and the network 160 through the communication module of the user terminal 1110 . It may be received by the user terminal 1110 . For example, a control signal or command of the server 150 received through the communication module may be transmitted to a processor or a memory.

Alternatively, the receiver 223 may provide a function for communicating with another external device. In this case, the other external device may be, for example, a device for generating medical image information used for a radiation treatment plan. For example, the external device may be a Computed Tomography (CT) device, a Magnetic Resonance Imaging (MRI) device, a Positron Emission Tomography (PET) device, a computed tomography simulator (CT Simulator), or a Computed Radiography (CR) device. The external device may be a device that provides clinical information or radiation treatment plan information of patients to the server 150 . Such an external device may be singular or plural.

The input/output interface 224 may be a means for interfacing with an input/output device. For example, the input device may include a device such as a keyboard or mouse, and the output device may include a device such as a display for displaying a communication session of an application. As another example, the input/output interface 224 may be a means for an interface with a device in which functions for input and output are integrated into one, such as a touch screen. As a more specific example, the processor 222 of the server 150 processes the command of the computer program loaded in the memory 221 , the service screen or content configured using the data provided by the user terminal 2 120 is input/output. may be displayed on the display via interface 224 .

Also, in other embodiments, the server 150 may include more components than those of FIG. 2 . However, there is no need to clearly show most of the prior art components. For example, the user terminal 1110 is implemented to include at least a portion of the above-described input/output interface 224 or other components such as a transceiver, a global positioning system (GPS) module, a camera, various sensors, a database, and the like. It may include more elements.

Hereinafter, with reference to the drawings, a method for planning intensity modulation radiation treatment according to an embodiment of the present invention will be described in detail.

3 is a view showing a method for predicting an intensity-modulated radiation treatment plan in a time series according to an embodiment of the present invention, Figure 4 is a diagram schematically illustrating a method for predicting an intensity-modulated radiation treatment plan according to an embodiment of the present invention It is a conceptual diagram, and FIG. 5 is a conceptual diagram for explaining a driving method of the multi-leaf collimator.

Referring back to Figures 2, 3, and 4, the intensity modulation radiation treatment planning method according to an embodiment of the present invention may follow the following process.

First, according to an embodiment of the present invention, the intensity modulation radiation treatment plan system receives the patient's medical image information, the manually established treatment plan fluence map information, and the dose distribution information by the receiving unit 223 (S10). ).

Next, the organ segmentation information generating unit 111 generates organ segmentation information using the patient's medical image information (S20). More specifically, the organ segmentation information generating unit 111 constructs a three-dimensional organ image of a target organ using the patient's medical image information, and divides the constructed three-dimensional organ image into two-dimensional images by plane segmentation information. will create

Next, the learning unit 112 generates a fluence map prediction model of the multi-leaf collimator by machine learning based on the generated organ segmentation information, the manually established treatment plan fluence map information, and the dose distribution information (S30) . Intensity modulated radiation treatment plan (IMRT) is irradiated by adjusting the intensity of radiation in each step while changing the positions of the leaves 171 of the multi-leaf collimator 170 . Here, as shown in FIG. 5, the fluence map of the multi-leaf collimator is an opening formed by the leaves 171 of the multi-leaf collimator 170 moving step by step when using an intensity modulated radiation treatment plan (IMRT). It may be a fluence map f1 for (A1).

Next, the radiation treatment plan derivation unit 113 uses the fluence map prediction model of the multi-leaf collimator and the organ division information and dose distribution information of the new patient or the organ division information and dose distribution information newly acquired from the existing patient to obtain the multi-leaf collimator. A radiation treatment plan including step-by-step location data is derived (S40).

In other words, the intensity-controlled radiation treatment plan prediction method according to an embodiment of the present invention having the above-described process is, as shown in FIG. 4 , the patient's medical image information (LD1) and the manually established treatment plan fluence map information ( By training using LD2) as learning data, it is possible to automatically output the driving data of the multi-leaf collimator as a radiation treatment plan. Through this, an ideal fluence map can be obtained while omitting the optimization process repeatedly performed by the treatment planner for the existing radiation treatment plan.

Hereinafter, the steps of S10 to S40 will be described in more detail.

As described above, the receiver 223 receives the patient's medical information, the manually established treatment plan fluence map information, and the dose distribution information. Here, the medical image information may be information generated by the medical imaging apparatus. For example, radiation through a medical imaging device such as a Computed Tomography (CT) device, a Magnetic Resonance Imaging (MRI) device, a Positron Emission Tomography (PET) device, a CT Simulator, or a Computed Radiography (CR) device. By photographing a living body to be investigated, medical image information may be generated. The generated medical image information is converted into a medical digital image communication standard (DICOM, Digital Imaging and Communications in Medicine) and is represented in the form of a DICOM RT (Radiation Therapy) data file.

The dose distribution information may include information on the maximum and minimum dose of a specific organ required by the treatment planner, and the permissible dose in the partial volume. In addition, the manually established treatment plan fluence map (MD, see FIG. 7 ) may be the optimal dose distribution information established manually based on the radiation surface passing through the center of the target organ.

6 is a flowchart illustrating a process of generating long-term division information in the long-term division information generation unit 111 in a time series, and FIGS. 7 to 9 are schematically shown to explain the process of generating the long-term division information of FIG. it is one drawing 10 is a diagram illustrating an example of data obtained after reconstructing a three-dimensional organ image with organ segmentation information.

Referring to FIG. 6 , first, the organ segmentation information generating unit 111 derives a 3D organ image from the patient's medical image information ( S210 ). Here, the patient's medical image information is magnetic resonance imaging (MRI), computed tomography (CT) image, positron emission tomography (PET) image, functional MRI (fMRI) image, fluoroscopic image, ultrasound image, or single photon emission tomography. It may include image information of at least one of the SPECT images. The patient's medical image information may be a two-dimensional organ image taken along the first direction (y-direction) of the patient's target organ. For example, when the patient's medical image information is an image by computed tomography (CT), the medical image information may include tomographic images captured while scanning along the first direction (y-direction). The organ segmentation information generating unit 111 constructs a three-dimensional organ image by using the two-dimensional organ images.

Next, the organ segmentation information generating unit 111 generates organ segmentation information including a plurality of 2D images by plane segmenting the 3D organ image according to the distance from the radiation source 10 ( S220 ). In this case, the organ segmentation information may be two-dimensional organ images obtained by dividing the target organ of the patient along the second direction (z direction) perpendicular to the first direction (y direction). In other words, the organ division information includes 2D images f0 and f1 reconstructed by plane division of the 3D organ image with respect to the direction (z direction) in which radiation is irradiated from the radiation source 10 . Accordingly, the organ segmentation information may also include distance information from the radiation source.

Next, the organ segmentation information generating unit 111 generates organ segmentation information by correcting the unit area of the plurality of 2D images with respect to the distance from the radiation source 10 ( S230 ). 7 and 8 , the radiation irradiated from the radiation source 10 is not irradiated in a uniform area according to the distance, but is irradiated in a form that spreads as it moves away from the radiation source 10 . That is, even when irradiated from the same radiation source 10, the size (u') of the radiation source covering the fluence map adjacent to the radiation source 10 is the size (u') of the radiation source covering the fluence map at the center of the target organ. smaller than the size (u). In other words, the size of the radiation source covering the fluence map is inversely proportional to the distance from the radiation source.

[Equation]

　

Here, D is the distance from the radiation source 10 to the center of the target organ, and R is the distance from the radiation source 10 to the first region f1 (R<D).

On the other hand, referring to FIGS. 7 and 9 , the organ division information generating unit 111 calculates the unit area of a plurality of 2D images from the radiation source 10 based on the manually established treatment plan fluence map information MD. Long-term segmentation information can be generated by correcting the distance of . The manually established treatment plan fluence map information (MD) is fluence map information at the center of the target organ as described above. In other words, the organ segmentation information generating unit 111 is a unit of a two-dimensional image in a region adjacent to or far from the radiation source 10 based on the unit area irradiated from the radiation source 10 at the center of the target organ. area can be corrected.

As shown in FIG. 9 , the unit area P1 of the first area f1 adjacent to the radiation source 10 is greater than the area f0 passing through the center of the target organ 10 is the second area f2 farther from the radiation source 10 . ) is smaller than the unit area (P2). That is, the total amount of radiation irradiated to the first region f1 and the second region f2 is the same, but the first region f1 is irradiated with a large amount of radiation in a narrow region, and the second region f2 is irradiated with a large amount of radiation. is irradiated with a small amount of radiation over a large area. Accordingly, the organ segmentation information generating unit 111 determines the unit area of the two-dimensional images with respect to the distance from the radiation source 10 in order to derive more accurate fluence map information in all regions based on the coordinate system of the target organ. Long-term segmentation information may be generated by linearly corrected, and may be provided as learning data of the learning unit 112 to be described later.

Next, the learning unit 112 may generate a fluence map prediction model of the multi-leaf collimator based on the characteristics extracted from the generated organ segmentation information, the manually established treatment plan fluence map information, and the dose distribution information. The learning unit 112 learns the fluence map prediction model of the multi-leaf collimator based on deep learning, and deep learning uses a combination of several non-linear transformation methods to obtain high-level abstractions (a large amount of data or complex data). It is defined as a set of machine learning algorithms that attempt to summarize key contents or functions in data. The learning unit 421 includes, for example, a deep neural network (DNN), a convolutional neural network (CNN), a recurrent neural network (RNN), and a deep belief among deep learning models. Networks, DBN) may be used. The learning unit 112 according to an embodiment of the present invention is characterized in that it does not learn from the patient's medical image information, but learns using the organ segmentation information newly reconstructed using the same.

The learning unit 112 may generate a first feature vector from organ segmentation information, generate a second vector from manually established treatment plan fluence map information, and generate a third vector from dose distribution information. The first to third vectors may be values for factors that directly or indirectly influence the fluence map prediction model of the multi-leaf collimator. The first vector to the third vector may be a set or a combination of values rather than a single value. The learner 112 may generate a fluence map prediction model of the multi-leaf collimator that derives position data of the leaves of the multi-leaf collimator as a result value using the first to third vectors.

The learning unit 112 includes: Logistic regression, Decision tree, Nearest-neighbor classifier, Kernel discriminate analysis, Neural network, Support Vector Machine, Algorithms and/or methods (techniques) such as random forest and boosted tree may be used.

The learning unit 112 may use algorithms and/or methods (techniques) such as linear regression, regression tree, kernel regression, support vector regression, and deep learning to predict the treatment plan.

In addition, the learning unit 112 may use algorithms and/or methods (techniques) such as principal component analysis, non-negative matrix factorization, independent component analysis, manifold learning, and SVD for vector calculation.

The learning unit 112 may use algorithms and/or methods (techniques) such as k-means, hierarchical clustering, mean-shift, and self-organizing maps (SOMs) for grouping information.

The learning unit 112 may use an algorithm and/or method (technique) such as bipartite cross-matching, n-point correlation two-sample testing, and minimum spanning tree for data comparison.

However, the above-described algorithm and/or method (technique) are exemplary and the spirit of the present invention is not limited thereto.

Next, the radiation treatment plan derivation unit 113 may obtain step-by-step position data of the multi-leaf collimator using the fluence map prediction model of the multi-leaf collimator, organ division information, and dose distribution information. In addition, the radiation treatment plan deriving unit 113 may derive radiation intensity data corresponding to each step of the multi-leaf collimator together.

11 is a diagram comparing a fluence map (clinical) obtained from a manually generated treatment plan with a fluence map (Generated) obtained through an embodiment of the present invention.

Referring to Figure 11, it can be confirmed that the fluence map obtained from the intensity modulation radiation treatment plan prediction system and method according to an embodiment of the present invention is almost the same as the fluence map obtained from the manually generated treatment plan. . Intensity-modulated radiation treatment plan prediction system and method according to an embodiment of the present invention, by using the patient's medical image information to generate long-term segmentation information, which is learning data, and use this as learning data to train, multi-leaf as a radiation treatment plan Collimator drive data can be automatically output. Through this, an ideal fluence map can be obtained while omitting the optimization process repeatedly performed by the treatment planner for the existing radiation treatment plan, and the quality of the treatment plan can be maintained while minimizing the time and effort for establishing a treatment plan.

The embodiment according to the present invention described above may be implemented in the form of a computer program that can be executed through various components on a computer, and such a computer program may be recorded in a computer-readable medium. In this case, the medium may be to store a program executable by a computer. Examples of the medium include a hard disk, a magnetic medium such as a floppy disk and a magnetic tape, an optical recording medium such as CD-ROM and DVD, a magneto-optical medium such as a floppy disk, and those configured to store program instructions, including ROM, RAM, flash memory, and the like.

Meanwhile, the computer program may be specially designed and configured for the present invention, or may be known and used by those skilled in the computer software field. Examples of the computer program may include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like.

The specific implementations described in the present invention are only examples, and do not limit the scope of the present invention in any way. For brevity of the specification, descriptions of conventional electronic components, control systems, software, and other functional aspects of the systems may be omitted. In addition, the connection or connection members of the lines between the components shown in the drawings exemplarily represent functional connections and/or physical or circuit connections, and in an actual device, various functional connections, physical connections that are replaceable or additional may be referred to as connections, or circuit connections. In addition, unless there is a specific reference such as "essential" or "importantly", it may not be a necessary component for the application of the present invention.

Therefore, the spirit of the present invention should not be limited to the above-described embodiments, and the scope of the spirit of the present invention is not limited to the scope of the scope of the present invention. will be said to belong to

10: radiation source
111: long-term division information generation unit
112: study department
113: Radiation treatment plan derivation unit
150: server
160: network
170: multi-leaf collimator
171: leaf

Claims (16)

In the intensity-modulated radiation treatment planning system comprising one or more memory and a processor,
The intensity-modulated radiation treatment planning system,
Includes; a receiver for receiving the patient's medical image information, the manually established treatment plan fluence map information, and dose distribution information;
The processor is
an organ segmentation information generation unit for generating organ segmentation information from the patient's medical image information; and
A learning unit that generates a fluence map prediction model of a multi-leaf collimator by machine learning based on the generated organ segmentation information, the manually established treatment plan fluence map information, and the dose distribution information;
The long-term division information generating unit,
A three-dimensional organ image is derived from the medical image information of the patient, a plurality of two-dimensional images are generated by plane segmenting the three-dimensional organ image according to a distance from a radiation source, and the unit area of the plurality of two-dimensional images is the Intensity-modulated radiation treatment plan prediction system for generating the long-term segmentation information by correcting the distance from the radiation source. delete According to claim 1,
The patient's medical image information is a two-dimensional organ image taken along the first direction of the target organ of the patient,
The organ segmentation information is a two-dimensional organ image obtained by segmenting the target organ of the patient along a second direction perpendicular to the first direction, intensity-modulated radiation treatment plan prediction system. 4. The method of claim 3,
The patient's medical image information is magnetic resonance imaging (MRI), computed tomography (CT) image, positron emission tomography (PET) image, functional MRI (fMRI) image, fluoroscopic image, ultrasound image, or single photon emission tomography method. (SPECT) Intensity modulation radiation treatment plan prediction system, including at least any one of the image information of the image. delete According to claim 1,
The long-term division information generating unit,
Intensity-modulated radiation treatment plan prediction system for generating the long-term segmentation information by correcting the unit area of the plurality of two-dimensional images with respect to the distance from the radiation source based on the manually established treatment plan fluence map information. According to claim 1,
The processor is
A radiation treatment plan derivation unit for obtaining step-by-step position data of the multi-leaf collimator using the fluence map prediction model of the multi-leaf collimator, organ segmentation information, and dose distribution information; further comprising, intensity-modulated radiation treatment plan prediction system. 8. The method of claim 7,
The radiation treatment plan derivation unit further acquires radiation intensity data corresponding to each step of the multi-leaf collimator using the fluence map prediction model, organ segmentation information, and dose distribution information of the multi-leaf collimator, intensity-modulated radiation treatment plan prediction system. Receiving the patient's medical image information, the manually established treatment plan fluence map information, and dose distribution information;
generating organ segmentation information from the received medical image information of the patient;
generating a fluence map prediction model of a multi-leaf collimator by machine learning based on the generated organ segmentation information, the manually established treatment plan fluence map information, and the dose distribution information; and
Deriving a radiation treatment plan including step-by-step position data of the multi-leaf collimator using the fluence map prediction model of the multi-leaf collimator, organ segmentation information, and dose distribution information;
The step of generating the long-term division information comprises:
deriving a three-dimensional organ image from the patient's medical image information;
generating a plurality of two-dimensional images by plane-segmenting the three-dimensional organ image; and
Comprising; generating the organ segmentation information by correcting the unit area of the plurality of two-dimensional images with respect to the distance from the radiation source; Containing, intensity modulation radiation treatment plan prediction method. 10. The method of claim 9,
The step of deriving the radiation treatment plan is,
Including radiation intensity data corresponding to each step of the multi-leaf collimator to derive the radiation treatment plan, intensity-modulated radiation treatment plan prediction method. delete 10. The method of claim 9,
The patient's medical image information is a two-dimensional organ image taken along the first direction of the target organ of the patient,
The organ segmentation information is a two-dimensional organ image obtained by segmenting the target organ of the patient along a second direction perpendicular to the first direction, intensity-modulated radiation treatment plan prediction method. 13. The method of claim 12,
The patient's medical image information is magnetic resonance imaging (MRI), computed tomography (CT) image, positron emission tomography (PET) image, functional MRI (fMRI) image, fluoroscopic image, ultrasound image, or single photon emission tomography method. (SPECT) Intensity modulation radiation treatment plan prediction method, including at least any one of the image information of the image. delete 10. The method of claim 9,
The step of generating the long-term division information comprises:
Intensity-modulated radiation treatment plan prediction method for generating the long-term segmentation information by correcting the unit area of the plurality of two-dimensional images with respect to the distance from the radiation source based on the manually established treatment plan fluence map information. A computer program stored in a medium for executing the method of any one of claims 9, 10, 12, 13 and 15 using a computer.

Images