Treatment support system and treatment support method

Abstract

Provided is a treatment support system capable of easily determining a radiation treatment device desirable for a patient. An arithmetic processing device 101 predicts a treatment effect index for each treatment modality of a radiation treatment device 202 by using a region of interest which corresponds to a part related to the irradiation of a patient with radioactive ray in a three-dimensional fluoroscopic image showing the patient, and also using parameter information which indicates a value of a parameter for the prediction of a treatment effect index for evaluating the effect of a radiation treatment for each treatment modality of the radiation treatment device 202 for performing the radiation treatment.

Description

Treatment support system and treatment support method

[0001] The present disclosure relates to a treatment support system and a treatment support method.

[0002] Particle beam therapy using proton beams or carbon beams has the advantage of being able to adjust the depth of the energy peak, known as the Bragg peak, to match the depth of the tumor, and is therefore superior in terms of dose concentration to the tumor compared to conventional X-ray therapy. This is expected to reduce exposure to normal tissue and organs at risk (OARs) and the likelihood of side effects.

[0003] However, particle therapy systems for performing particle therapy require large equipment such as accelerators and rotating gantries, which require a large installation area, and are generally more expensive than X-ray therapy systems for performing X-ray therapy. Therefore, only a limited number of facilities can possess particle therapy systems, and therefore the number of patients who can receive particle therapy at one time is limited.

[0004] Therefore, rather than administering particle beam therapy to all cancer patients, it is necessary to identify patients who can fully benefit from particle beam therapy and preferentially guide them to particle beam therapy. Note that, for example, in cases with small tumor volumes, it is known that conventional X-ray therapy can sufficiently reduce exposure to OAR and provide treatment with the same level of effectiveness as particle beam therapy.

[0005] As a judgment index for determining whether a patient is suitable for particle beam therapy, the NTCP (Normal tissue complication probability) of OAR has attracted attention (see Non-Patent Document 1). The NTCP of OAR is the probability that irreversible damage will occur in an OAR adjacent to a tumor due to irradiation with radiation, and is generally calculated using a mathematical model constructed based on past treatment data. The mathematical model uses, as input, a three-dimensional dose distribution in the patient's body calculated using treatment planning software for formulating a radiation therapy treatment plan. According to a typical mathematical model, the NTCP is calculated using Equation 1. Here, t is Furthermore, D ｍａｘ V ｎｐ teeth, Here, V is the volume of the entire OAR, v is the volume per voxel, and k is the number of the voxel included in the OAR. ｐ , m ｐ and T.D. ５０ is a parameter determined for each OAR, type of side effect, and case, and is calculated based on past clinical data, etc.

[0006] Also, D' ｋ is the converted total irradiation dose in voxel k, and the single irradiation dose is d ｒｅｆ = 2 [Gy], the total irradiation dose D ｋ and indicates the total irradiation dose required to obtain the same biological effect as irradiation with a total number of fractions N. According to the linear-quadratic model (LQ model), which is a general evaluation model for evaluating the relationship between radiation dose and biological effect, the total irradiation dose D ｋ And when the total number of divisions is N, the probability that a normal tissue cell contained in voxel k will survive is λ ｋ teeth, Here, α and β are parameters that indicate the radiosensitivity of normal tissue cells, and can be determined by in vitro radiation irradiation experiments. Based on these relationships, the converted total irradiation dose D' ｋ is calculated by the following equations 5 and 6.

[0007] NTCP calculated from the dose distribution of X-ray therapy is ｘ , the NTCP calculated from the dose distribution of particle beam therapy is NTCP ｐ Then, the difference between them, the NTCP difference (ΔNTCP = NTCP ｘ -NTCP ｐ For patients with a large NTCP difference, particle therapy is considered to be advantageous, and preferential treatment with particle therapy is considered. On the other hand, for patients with a small NTCP difference, X-ray therapy is considered to be equivalent to particle therapy, and preferential treatment with X-ray therapy is considered.

[0008] Yoshimura T, Kinoshita R, Onodera S, Toramatsu C, Suzuki R, Ito YM, Takao S, Matsuura T, Matsuzaki Y, Umegaki K, Shirato H, Shimizu S. NTCP modeling analysis of acute hematologic toxicity in whole pelvic radiation therapy for gynecologic malignancies - A dosimetric comparison of IMRT and spot-scanning proton therapy (SSPT). Phys Med. 2016 Sep;32(9):1095-102. doi: 101016 / j.ejmp.2016.08.007. Epub 2016 Aug 25. PMID: 27567089.

[0009] As described above, the NTCP difference is a useful indicator for determining whether a patient is suitable for particle beam therapy. However, in order to calculate the NTCP difference, it is necessary to prepare treatment plans for both X-ray therapy and particle beam therapy and to obtain a three-dimensional dose distribution in the patient's body by simulation from these treatment plans, which requires time and effort.

[0010] Furthermore, there are various irradiation methods for X-ray therapy, such as 3D-CRT (Three Dimensional Conformal Radiotherapy), IMRT (Intensity Modulated Radiotherapy), and VMAT (Volummetric Modulated Arc Therapy), and each of these irradiation methods requires a different treatment modality (type of radiation therapy device). Therefore, similar to the determination index for determining whether a patient is suitable for particle beam therapy, the difference in NTCP for each treatment modality can be used as a judgment index for determining the treatment modality suitable for a patient. However, since it is necessary to obtain the three-dimensional dose distribution within the patient's body for each treatment modality, additional time and effort are required.

[0011] An object of the present disclosure is to provide a treatment support system and a treatment support method that make it possible to easily determine an appropriate radiation therapy device for a patient.

[0012] A treatment support system according to one aspect of the present disclosure is a treatment support system that predicts the effect of radiation therapy on a patient, and includes a calculation processing device that predicts the treatment effect index for each type of radiation therapy device using a region of interest in a three-dimensional fluoroscopic image of the patient that corresponds to a part of the patient that is related to radiation irradiation of the patient and parameter information that indicates, for each type of radiation therapy device that performs the radiation therapy, the values ​​of parameters for predicting the treatment effect index that evaluates the effect of the radiation therapy.

[0013] According to the present disclosure, it becomes possible to easily determine the appropriate radiation therapy device for a patient.

[0014] Fig. 1 is a diagram illustrating a treatment selection support system according to an embodiment of the present disclosure; Fig. 2 is a flowchart for explaining an example of treatment selection processing by the treatment selection support system according to an embodiment of the present disclosure; Fig. 3 is a diagram illustrating an example of a GUI used by an operator in the treatment selection processing; Fig. 4 is a diagram illustrating another example of a GUI used by an operator in the treatment selection processing; Fig. 5 is a diagram for explaining an example of an algorithm of a DVH prediction model;

[0015] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.

[0016] FIG. 1 is a diagram illustrating a treatment selection support system according to an embodiment of the present disclosure. The treatment selection support system 100 illustrated in FIG. 1 is a treatment support system configured with devices capable of various information processing, for example, information processing devices such as computers. The treatment selection support system 100 is installed in a treatment facility, such as a hospital, that performs radiation therapy. In the example of FIG. 1, facilities A to C are shown as treatment facilities, and the treatment selection support system 100 at facility A is shown, but facilities B and C also have systems equivalent to the treatment selection support system 100 installed therein.

[0017] The treatment selection support system 100 includes a processing device 101, an input device 102, a display device 103, a memory 104, a database 105, and a communication device 106. The processing device 101 is connected to the input device 102, the display device 103, the memory 104, the database 105, and the communication device 106. The connection method for connecting these devices is not particularly limited, and may be, for example, a connection method via a LAN (Local Area Network) or a WAN (Wide Area Network) such as the Internet.

[0018] The arithmetic processing device 101 is a processor such as a CPU (Central Processing Unit), a GPU (Graphic Processing Unit), or an FPGA (Field-Programmable Gate Array), and constitutes a control unit that controls the entire treatment selection support system 100.

[0019] The input device 102 is a device, such as a mouse and keyboard, that receives various information from an operator who operates the treatment selection support system 100. The operator is, for example, a doctor or a medical physicist. The display device 103 is a device, such as a display, that displays various information such as the calculation results calculated by the arithmetic processing device 101.

[0020] The memory 104 and the database 105 are composed of the same or different recording media, and record a program (computer program) that defines the operation of the arithmetic processing device 101, as well as various information used and generated by the arithmetic processing device 101. The recording media may be, for example, a magnetic storage medium such as a hard disk drive (HDD), a semiconductor storage medium such as a random access memory (RAM), a read-only memory (ROM), and a solid state drive (SSD), or a combination of an optical disk and an optical disk drive, such as a digital versatile disk (DVD). When the treatment selection support system 100 starts operating (e.g., when the power is turned on), the arithmetic processing device 101 reads the program from the recording medium and executes the read program to perform various processes.

[0021] The communication device 106 is a communication interface that is communicatively connected to external devices. In the example of Fig. 5, the communication device 106 is connected to an information system 201 and a radiotherapy device 202 installed in the facility (facility A) as external devices, and an external database 203 installed outside the facility.

[0022] The information system 201 is a computer device that performs various information processing related to radiation therapy and stores various information used and generated in the information processing. For example, the information system 201 performs a planning process to create a treatment plan for radiation therapy as information processing, and stores the treatment plan, which is the result of the planning process, and treatment planning CT (Computed Tomography) images, which are treatment planning images for creating the treatment plan.

[0023] The radiation therapy device 202 is a device that actually performs radiation therapy on a patient. There are multiple types of radiation therapy devices 202, which are treatment modalities, and in Figure 1, two radiation therapy devices 202A and 202B with different treatment modalities are shown as the radiation therapy device 202. Examples of treatment modalities include a proton therapy device, a carbon ion therapy device, a 3D-CRT device, an IMRT device, and a VMAT device. Note that radiation therapy devices 202 with different treatment modalities may be installed in each treatment facility.

[0024] The external database 203 is a storage device that stores parameter information indicating the values ​​of parameters used to predict the value of a therapeutic effect index that evaluates the effect of radiation therapy on a patient. The external database 203 is configured, for example, from a recording medium, similar to the memory 104 and the database 105. In this embodiment, the therapeutic effect indexes are a DVH (Dose-Volume Histogram) and an NTCP. The values ​​of the parameters used to predict the value of the therapeutic effect index differ for each treatment modality, and the parameter information indicates the parameter values ​​for each treatment modality.

[0025] FIG. 2 is a flowchart illustrating an example of treatment selection processing by the treatment selection support system 100. FIGS. 3 to 5 are diagrams illustrating an example of a GUI (Graphical User Interface) 300 used by an operator in the treatment selection processing. The GUI 300 of this embodiment has three tabs (a planning image prediction tab 301, a ROI (Region of Interest) extraction tab 306, and a DVH prediction tab 307). FIG. 3 shows the GUI 300 with the planning image prediction tab 301 selected, FIG. 4 shows the GUI 300 with the ROI extraction tab 306 selected, and FIG. 5 shows the GUI 300 with the DVH prediction tab 307 selected. The GUI 300 is displayed on the display device 103.

[0026] In the treatment selection process, first, the arithmetic processing device 101 acquires a diagnostic CT image as a diagnostic image for diagnosing a patient's disease (step S1). For example, the operator uses the input device 102 to select the planning image prediction tab 301 of the GUI 300 shown in Fig. 3, and then presses the file input button 302 to select a diagnostic CT image to be used in the treatment selection process. The arithmetic processing device 101 acquires the selected diagnostic CT image.

[0027] The arithmetic processing unit 101 records the acquired diagnostic CT image in the memory 104 or the database 105 (step S2).

[0028] 3, the arithmetic processing device 101 creates a predicted planning CT image, which is a predicted planning image obtained by predicting a treatment planning CT image for creating a treatment plan for the patient, based on the diagnostic CT image recorded in the memory 104 or the database 105. The arithmetic processing device 101 records the predicted planning CT image in the memory 104 or the database 105 (step S3).

[0029] The planning predicted CT image is created using a planning image prediction model implemented by an application program recorded in the memory 104 or the database 105. The planning image prediction model is a model that predicts a treatment planning CT image from a diagnostic CT image and outputs the predicted planning CT image. The planning image prediction model is constructed, for example, by machine learning such as deep learning.

[0030] Since treatment planning CT images are used to simulate radiation exposure to patients during treatment planning, they are captured in a different position than diagnostic CT images, such as with the patient's arms raised. Furthermore, to perform accurate simulations, treatment planning CT images must capture the entire area from the body surface to the target. Therefore, the field of view (FOV) of treatment planning CT images and diagnostic CT images may differ. Furthermore, diagnostic CT images may use filters to improve the visibility of the patient's internal structures. Therefore, the pixel values ​​of each voxel may differ between treatment planning CT images and diagnostic CT images. Furthermore, diagnostic CT images and treatment planning CT images may have different resolutions due to protocols at each medical institution, or the settings of the CT imaging device (such as the X-ray source tube settings) used during imaging may differ.

[0031] 3, the arithmetic processing unit 101 displays the diagnostic CT image and the planning-use predicted CT image recorded in the memory 104 or the database 105 as a diagnostic CT image 304 and a predicted CT image 305 on the GUI 300 (step S4), allowing the operator to check the diagnostic CT image and the planning-use predicted CT image.

[0032] Thereafter, when the operator selects ROI extraction tab 306 on GUI 300, GUI 300 is displayed in the state shown in Fig. 4. Then, when the operator presses extraction start button 401 on GUI 300, processing unit 101 extracts a predicted region of interest that predicts a region of interest corresponding to a part of the patient that is related to radiation irradiation from the planning prediction CT image recorded in memory 104 or database 105, and records predicted region of interest information indicating the predicted region of interest in memory 104 or database 105 (step S5).

[0033] The region of interest includes a target region corresponding to the target, which is the area to be irradiated with radiation, and an exclusion region corresponding to the OAR, which is the area to avoid irradiation with radiation. For example, in radiotherapy for prostate cancer, the region corresponding to the tumor in the prostate corresponds to the target region, and the region corresponding to the rectum and bladder corresponds to the exclusion region. Furthermore, extraction of the region of interest is performed using an automatic region of interest extraction model implemented by an application program recorded in memory 104 or database 105. The automatic region of interest extraction model predicts a region of interest from a predicted planning CT image and outputs three-dimensional position information of the predicted region of interest, which is the predicted region of interest, as predicted region of interest information. The automatic region of interest extraction model is constructed, for example, by machine learning such as deep learning.

[0034] Next, the arithmetic processing device 101 displays the GUI 300 by superimposing the predicted region of interest indicated by the predicted region of interest information recorded in the memory 104 or the database 105 as a region of interest 402 on the predicted planning CT image 403 (step S6), as shown in FIG. 4 . At this time, the arithmetic processing device 101 may further display detailed information 404 regarding the predicted region of interest, as shown in FIG. 4 . The detailed information includes, for example, the type of the predicted region of interest (target region or exclusion region) and the name of the part corresponding to the predicted region of interest. This allows the operator to confirm the predicted region of interest. Note that the treatment selection support system 100 may be configured so that the operator can use the input device 102 to instruct modifications to the shape and position of the region of interest 402. In this case, the arithmetic processing device 101 modifies the shape and position of the predicted region of interest in accordance with the modification instructions.

[0035] Next, when the operator selects the DVH prediction tab 307 of the GUI, the GUI 300 is displayed as shown in FIG. When the operator presses the prediction button 501 on the GUI 300, the arithmetic processing device 101 calculates a predicted DVH, which predicts the DVH as a three-dimensional dose distribution in the patient's body formed by radiation therapy, for each treatment modality based on the predicted region of interest information recorded in the memory 104 or the database 105. Furthermore, the arithmetic processing device 101 calculates a predicted NTCP, which predicts the NTCP from the predicted DVH, for each treatment modality. The arithmetic processing device 101 then records predicted DVH information indicating the predicted DVH for each treatment modality and predicted NTCP information indicating the predicted NTCP for each treatment modality in the memory 104 or the database 105 (step S7). The DVH and NTCP are treatment effect indices, and the predicted DVH and predicted NTCP are prediction results of the treatment effect indices.

[0036] The predicted DVH is calculated using a DVH prediction model implemented by an application program stored in memory 104 or database 105. The DVH prediction model is prepared for each treatment modality, and predicts the DVH for the corresponding treatment modality from the predicted region of interest information and outputs it as the predicted DVH. The DVH prediction model will be described in more detail later (see FIG. 6 ).

[0037] Next, as shown in FIG. 5 , the arithmetic processing device 101 displays the predicted DVH and predicted NTCP for each treatment modality recorded in the memory 104 or the database 105 on the GUI 300 as treatment effect information 510 (step S8). This allows the operator to check the predicted DVH and predicted NTCP, which are treatment effect indicators of radiation therapy, for each treatment modality. The predicted DVH is calculated for each region of interest, and the operator can switch the displayed predicted DVH using the tag 502. Furthermore, the predicted NTCP is calculated for each type and grade of side effect of radiation therapy, and the operator can switch the displayed predicted NTCP using the tag 503.

[0038] Next, the operator uses the input device 102 to select a reference modality, which is a reference treatment modality, for the NTCP list 507, which is a list of predicted NTCPs for each treatment modality displayed on the GUI 300 (step S9). The reference modality can be selected arbitrarily by the operator, but for example, the treatment modality of the radiation therapy device 202 installed in the operator's own facility is selected. In the example shown in Figure 5, "X-ray therapy system A," which is a 3D-CRT device, is selected as the reference modality.

[0039] Next, the arithmetic processing device 101 acquires reference modality information indicating the reference modality selected by the operator and records the reference modality information in the memory 104 or the database 105. Furthermore, the arithmetic processing device 101 calculates an NTCP difference, which is the difference between the predicted NTCP of the comparison modality and the predicted NTCP of the reference treatment modality, for each comparison modality, which is a treatment modality other than the reference modality. Then, the arithmetic processing device 101 records NTCP difference information indicating the NTCP difference for each treatment modality in the memory 104 or the database 105 (step S10).

[0040] The arithmetic processing device 101 then displays the NTCP difference information recorded in the memory 104 or the database 105 as the NTCP difference 508 on the display device 103 (step S11), and terminates the process. This allows the operator to confirm the NTCP difference between the radiotherapy device 202 at the facility, which has been set as the reference modality, and other treatment modalities. In step S11, the arithmetic processing device 101 may suggest to the operator that the comparison modality be adopted by displaying an icon 509 for a comparison modality whose NTCP difference exceeds a predetermined reference value. In this case, the operator can more easily confirm a treatment modality that is expected to have a higher therapeutic effect than the treatment modality at the facility. Alternatively, the operator may return to the process of step S9 and change the reference modality.

[0041] Fig. 6 is a diagram for explaining an example of an algorithm of a DVH prediction model that calculates a predicted DVH in the processing of step S7 in Fig. 2. Note that the method of implementing the DVH prediction model is not limited to the algorithm below, and machine learning such as deep learning may also be used. The arithmetic processing device 101 performs the following processing in accordance with the algorithm of the DVH prediction model.

[0042] The arithmetic processing device 101 first calculates an enlarged target area 602 by enlarging the target area 601. Specifically, as shown in FIG. 6A, the arithmetic processing device 101 first sets the target area 601 as a target area, and defines a plurality of points P ｉ （０） Set the point P ｉ （０） It is desirable that the numbers are arranged evenly (at equal intervals). Also, i is an integer from 1 to N, and N is P ｉ （０） The total number of points P ｉ （０） A normal vector r to the surface of the region of interest, passing from the inside to the outside of the region of interest at ｉ （０） Furthermore, the calculation processing unit 101 calculates the ｉ （０） Let r be the normal vector passing through that point. ｉ （０） Point P moved by ｉ （１） (=P ｉ （０） +r ｉ （０） ) is calculated as the enlarged target region 602.

[0043] Then, every time the calculation processing device 101 calculates the enlarged target area 602 according to the following formula (2), it performs the above process again using the enlarged target area as the target area, and calculates the enlarged target area 602 by dividing each point P ｉ （ｊ） Let r be the normal vector passing through that point. ｉ （ｊ） Point P moved by ｉ （ｊ＋１） (=P ｉ（０） +r ｉ （０） ) is repeatedly generated, where j indicates the number of times the target region 601 is generated. ｉ （ｊ） The absolute value of each point P ｉ （ｊ） and is, for example, about 2 mm to 10 mm. ｉ （ｊ） The absolute value of j may differ depending on the protocol of each facility, etc. The maximum value of the number of expansions j is, for example, determined in advance.

[0044] Next, as shown in FIG. 6B, the arithmetic processing device 101 calculates an OV (Overlap Volume) area where the enlarged target area 602 and the OAR 603 overlap for each enlargement count j. Furthermore, the arithmetic processing device 101 calculates a differential OV area OV, which is the difference between the OV area in the j-th enlargement target area 602 and the OV area in the j-1th enlargement target area 602. ｊ 604 is calculated, and the difference OV area OV ｊ 604 and its volume dV ｊ and are recorded in the memory 104 or the database 105. When j=0, the difference OV region ΔOV ｊ 604 is the area where the target area 601 and the OAR 603 overlap.

[0045] Next, the arithmetic processing unit 101 calculates the difference OV region ΔOV recorded in the memory 104 or the database 105. ｊ Volume dV per 604 ｊ Using the above, the predicted DVHV(x) is calculated according to the following equation 6. Here, dV(x) is is.

[0046] Also, D ｊ is the difference OV area ΔOV ｊ the average dose delivered to 604, σ ｊ is the differential OV area ΔOV ｊ604. These parameters are used to predict the therapeutic effect index, and their values ​​vary for each treatment modality and can be calculated from treatment plans of previously performed treatments. Parameter information indicating the parameter values ​​for each treatment modality is stored, for example, in the external database 203, as described above. The parameter information may include parameter values ​​corresponding to radiation therapy devices located in different facilities.

[0047] As shown in FIG. 5 , when the operator presses the modality selection button 504 in step S7, a modality selection window 505 is displayed. The operator uses the modality selection window 505 to select a treatment modality for which the predicted result of the treatment effect index is to be confirmed, and when the operator presses the close button 506, the arithmetic processing device 101 accesses the external database 203 via the communication device 106. The arithmetic processing device 101 then obtains parameter values ​​corresponding to the treatment modality selected in the modality selection window 505 from the parameter information stored in the external database 203, and uses the parameter values ​​to calculate the predicted DVH. Also, as shown in FIG. 5 , even if the treatment system is the same, treatments at different facilities can be registered as different types of treatment modalities. Furthermore, similar treatment modalities, such as treatment modalities with the same irradiation method, may be registered as a single treatment modality.

[0048] In addition, as shown in Equation 6 and Equation 7, in this embodiment, the difference OV region ΔOV ｊ The histogram of the dose applied to 604 is modeled using a Gaussian distribution, although the distribution used for modeling can be changed as appropriate depending on actual data from previous experiments, and may be, for example, a Cauchy distribution or a Landau distribution.

[0049] In the above-described embodiment, DVH and NTCP were calculated as the therapeutic effect index. However, the therapeutic effect index is not limited to this example. For example, DVH and NTCP alone may be used, or other indices such as TCP (Tumor Control Probability) may be used, or a combination of these may be used. Calculation of a therapeutic effect index such as TCP may require patient information about the patient in addition to a predictive planning CT image. In this case, the arithmetic processing device 101 calculates the therapeutic effect index using the patient information. The patient information is stored, for example, in the information system 201 and acquired by the arithmetic processing device 101 as needed. The patient information may also include, for example, age, gender, lifestyle history such as smoking and drinking habits, test results of various tests (blood tests, biopsies, genetic tests, etc.), and whether or not concomitant medication such as anticancer drugs and immune checkpoint inhibitors is being used.

[0050] In this embodiment, the arithmetic processing device 101 generates a prediction planning CT image from a diagnostic CT image and predicts a treatment effect index based on the prediction planning CT image. However, the image used to predict a treatment effect index is not limited to the prediction planning CT image. For example, other 3D fluoroscopic images, such as treatment planning CT images actually used to create a patient's treatment plan, can be used as the image used to predict a treatment effect index. If a region of interest has already been set for the treatment planning CT image, the arithmetic processing device 101 can omit the process of predicting the region of interest and use the already set region of interest to predict the treatment effect index.

[0051] Furthermore, the treatment selection support system 100 may evaluate the appropriateness of the treatment modality actually selected by the operator.

[0052] After the operator selects a treatment modality, a treatment planning CT image is actually acquired using a treatment planning CT device (not shown). The information system 201 then sets a region of interest on the treatment planning CT image according to instructions from the operator. Furthermore, the information system 201 uses treatment planning software corresponding to the selected treatment modality to create a treatment plan with an optimized dose distribution within the patient, so as to deliver a high dose as uniformly as possible to the target region and to minimize exposure to the OAR. The information system 201 calculates the DVH based on the dose distribution and the region of interest according to the treatment plan. The information system 201 stores the treatment planning CT image, region of interest information indicating the region of interest, and the DVH. If a treatment modality not available at the patient's facility is selected, the patient is referred to another facility equipped with that treatment modality, and the above-described series of procedures are performed at that facility.

[0053] The arithmetic processing device 101 of the treatment selection support system 100 acquires treatment planning CT images, region of interest information, and DVHs from the information system 201 and stores them in the memory 104 or the database 105. The arithmetic processing device 101 calculates the degree of agreement between the treatment planning CT images, region of interest information, and DVHs and the predicted planning CT images, predicted region of interest information, and predicted DVHs calculated in the treatment selection process described with reference to FIG. 2 . The arithmetic processing device 101 determines whether the degree of agreement is equal to or greater than a predetermined reference value. If the degree of agreement is less than the reference value, it is considered that the accuracy of the prediction of the treatment planning CT images, predicted region of interest, and predicted DVHs by the treatment selection support system 100 is insufficient. In this case, the arithmetic processing device 101 outputs an alert, such as by displaying a message on the GUI 300 indicating that there is doubt about the validity of the selected treatment modality, to prompt the operator to reconsider the treatment modality. The arithmetic processing device 101 may output an alert when at least one of these degrees of match is below a reference value, or may output an alert when all of these degrees of match are below a reference value.

[0054] As described above, according to this embodiment, the arithmetic processing device 101 predicts a therapeutic effect index for each therapeutic modality of the radiation therapy device 202 by using a region of interest corresponding to a part of the patient related to radiation irradiation in a 3D fluoroscopic image of the patient and parameter information indicating, for each therapeutic modality of the radiation therapy device 202 that performs radiation therapy, the values ​​of parameters for predicting a therapeutic effect index for evaluating the effect of radiation therapy. Therefore, since it is possible to predict a therapeutic effect index for each therapeutic modality, it is not necessary to obtain a 3D dose distribution in the patient's body by simulation for each therapeutic modality. This makes it possible to easily determine the appropriate radiation therapy device for the patient.

[0055] In this embodiment, the arithmetic processing device 101 generates, as a 3D fluoroscopic image for predicting a treatment effect index, a predicted planning CT image, which is a treatment planning CT image for formulating a radiation therapy treatment plan based on a diagnostic CT image for diagnosing a patient's disease. This makes it possible to predict a treatment effect index for each treatment modality before capturing a treatment planning CT image, i.e., before formulating a treatment plan. Generally, the treatment modality is already determined when a treatment planning CT image is captured. Therefore, predicting a treatment effect index for each treatment modality before capturing a treatment planning CT image, as in this embodiment, is very useful for easily determining an appropriate radiation therapy device for a patient.

[0056] In addition, in this embodiment, the calculation processing device 101 outputs an alert if the degree of agreement between the treatment planning CT image, region of interest information, and DVH used when actually creating the treatment plan and the planning predicted CT image, predicted region of interest information, and predicted DVH predicted in the treatment selection process is less than a reference value.

[0057] In this embodiment, the arithmetic processing device 101 extracts a region of interest from the planning prediction CT image, which eliminates the need for the operator to set a region of interest in order to predict the treatment effect index, thereby reducing the effort required to determine the appropriate radiation therapy device 202 for the patient.

[0058] In this embodiment, the arithmetic processing device 101 acquires parameter information for predicting the therapeutic effect index from an external database 203 that stores parameter information indicating the values ​​of parameters corresponding to multiple types of radiotherapy devices 202 installed in multiple facilities A to C. This makes it possible to determine whether a radiotherapy device 202 with a therapeutic modality that is not present in the facility itself is appropriate, thereby making it possible to provide more appropriate radiotherapy for the patient.

[0059] In this embodiment, the therapeutic effect index includes at least one of DVH, NTCP, and TCOP. The arithmetic processing device 101 may further use patient information about the patient to predict the therapeutic effect index. This makes it possible to predict an appropriate therapeutic effect index depending on the situation, etc.

[0060] The above-described embodiments of the present disclosure are merely illustrative examples of the present disclosure, and are not intended to limit the scope of the present disclosure to these embodiments alone. Those skilled in the art may implement the present disclosure in various other forms without departing from the scope of the present disclosure.

[0061] 100: Treatment selection support system 101: Processing unit 102: Input device 103: Display device 104: Memory 105: Database 106: Communication device 201: Information system 202: Radiation therapy device 203: External database

Claims

1. A treatment support system for predicting the effects of radiation therapy on a patient, comprising: a calculation processing device that predicts a treatment effect index for each type of radiation therapy device using a region of interest in a three-dimensional fluoroscopic image of the patient that corresponds to the area of ​​the patient that is related to the irradiation of radiation, and parameter information indicating parameter values ​​for predicting a treatment effect index evaluating the effects of the radiation therapy for each type of radiation therapy device that performs the radiation therapy. The treatment support system has a calculation processing device that predicts the treatment effect index for each type of radiation therapy device that performs the radiation therapy.

2. The treatment support system of claim 1, wherein the arithmetic processing device generates, as the three-dimensional fluoroscopic image, a predicted planning image that predicts a treatment planning image for formulating a treatment plan for the radiation therapy based on a diagnostic image for diagnosing the patient's disease.

3. The treatment support system of claim 2, wherein the arithmetic processing device calculates the degree of match between the treatment planning images used when the treatment plan was actually created and the three-dimensional fluoroscopic images, and outputs an alert if the degree of match is below a reference value.

4. The treatment support system according to claim 2, wherein the arithmetic processing device extracts the region of interest from the three-dimensional fluoroscopic image.

5. The treatment support system of claim 4, wherein the arithmetic processing device calculates a degree of correspondence between the region of interest set in a treatment planning image used when the treatment plan was actually created and the extracted region of interest, and outputs an alert if the degree of correspondence is less than a reference value.

6. The treatment support system of claim 2, wherein the arithmetic processing device calculates a degree of agreement between a treatment effect index calculated from the region of interest and the parameters set in a treatment planning image used when actually creating the treatment plan and the predicted treatment effect index, and outputs an alert if the degree of agreement is less than a reference value.

7. The treatment support system of claim 1, wherein the arithmetic processing device acquires the parameter information from a storage device that stores the parameter information indicating the values ​​of the parameters corresponding to multiple types of radiation therapy devices installed in multiple treatment facilities, and uses the parameter information to predict the treatment effect index.

8. The treatment support system according to claim 1, wherein the calculation processing device further uses patient information relating to the patient to predict the treatment effect index.

9. The treatment support system according to claim 1, wherein the treatment effect index includes at least one of a dose-volume histogram, a probability of normal tissue damage, and a probability of tumor control.

10. A treatment support method using a treatment support system that predicts the effects of radiation therapy on a patient, comprising: predicting a treatment effect index for each type of radiation therapy device using a region of interest in a three-dimensional fluoroscopic image of the patient that corresponds to a part of the patient that is related to radiation irradiation, and parameter information indicating parameter values ​​for predicting a treatment effect index evaluating the effect of the radiation therapy for each type of radiation therapy device that performs the radiation therapy.

Images