JP6351011B2 - Positioning device and method of operating positioning device - Google Patents

Description

The present invention relates to a positioning device that positions a patient when a treatment is performed on the patient, and an operation method of the positioning device .

  In radiation therapy in which radiation such as X-rays, electron beams, and particle beams is irradiated to an affected area such as cancer, it is necessary to accurately irradiate the affected area with radiation. When performing radiotherapy using a radiotherapy apparatus, a treatment plan is formulated prior to treatment. When performing radiotherapy, it is necessary to perform positioning so that the position of the patient coincides with that at the time of treatment planning. Conventionally, positioning of a patient in a radiotherapy apparatus has been performed by operating a position of a medical table on which a patient is placed while an engineer visually confirms a position indicated by a laser marking device. In recent years, it has been proposed to position a patient using an X-ray fluoroscopic image (see Patent Document 1 to Patent Document 3).

  In such radiotherapy, the patient is restrained on the treatment table in the same posture from the positioning of the patient to the end of the treatment, which is a stress on the patient. In order to reduce such stress on the patient, it is required to position the patient faster. Further, in order to allow more patients to have a treatment opportunity with the radiation therapy apparatus, it is necessary to increase the positioning speed of the patient and shorten the use time of the radiation therapy apparatus per patient.

  In the patient positioning device of Patent Document 2, it is proposed to reduce the number of repetitions of positioning by learning the rough positioning position of the treatment table based on the past treatment table positioning data of the patient.

  Moreover, in the patient positioning using the X-ray image, the geometric arrangement of the X-ray imaging system in the treatment apparatus is reproduced on the computer, and the three-dimensional image data collected by the X-ray CT apparatus at the time of treatment planning is used. DRR (Digital Reconstructed Radiography) which is virtual perspective projection is executed. Then, by evaluating the similarity between the actual X-ray image and the DRR image (image registration), a deviation amount between the current position of the patient and the position at the time of treatment planning is calculated. Since DRR has a huge calculation cost, Patent Document 3 proposes an automatic positioning apparatus that speeds up the positioning of a patient in radiation therapy by reducing the calculation cost necessary for the execution of DRR. .

JP 2007-282877 A JP 2010-57810 A JP 2013-99431 A

  The similarity between the DRR image to be compared in the image registration and the actual X-ray image of the patient is generally a non-convex function with respect to the projection geometry at the time of creating the DRR image. For this reason, a solution obtained by optimizing an evaluation scale for evaluating the similarity between both images may be a local solution. On the other hand, in order to prevent falling into a local solution at the time of optimization calculation, it is effective to bring the initial parameter of the projection geometry at the start of optimization closer to the optimal solution. That is, it is required to bring the actual patient imaging geometry closer to the DRR projection geometry at the stage of coarse adjustment, which is positioning before starting the optimization calculation.

  The rough positioning of the patient before the start of the conventional optimization calculation is performed by, for example, an adjustment using a laser marking device or an operator via the input device with the DRR image displayed on the display and the patient's actual fluoroscopic image. These operations are performed manually by the operator, such as overlaying, and the confirmation is an operation that depends on the visual observation of the operator. For this reason, an individual difference in judgment by the operator may become a difference in positioning accuracy. However, it is not preferable that the positioning accuracy to be achieved in relation to the treatment position varies depending on the skill level of the operator. Furthermore, the conventional method described above tends to increase the time required for positioning.

  Moreover, in the patient positioning apparatus of patent document 2, since the positioning is performed based on the past treatment table positioning data, the first treatment by divided irradiation or the treatment by one irradiation such as stereotactic radiation irradiation is performed. In the case of treatment of a patient who does not have past treatment table positioning data, such as when it is completed, manual positioning has to be performed as in the past. In such a case, problems such as individual differences in the operator's judgment in positioning the patient, and a long time for positioning occur.

The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a positioning device capable of automatically positioning a patient at a high speed and an operating method of the positioning device .

According to a first aspect of the present invention, there is provided a DRR image creation unit that reproduces a geometric arrangement of an X-ray imaging system on a computer and obtains a DRR image by virtually performing perspective projection on previously collected X-ray CT image data; ,
The X-ray CT so that the evaluation function for evaluating the degree of coincidence between the DRR image obtained by the DRR image creation unit and the actual captured image obtained by fluoroscopy or imaging the patient with an X-ray imaging system is maximized. Optimizing perspective projection parameters relating to rotation / translation of image data and acquiring the X-ray CT image data, and the actual patient position when the patient is seen or imaged by an X-ray imaging system And an optimization unit that calculates the amount of deviation from the position, and outputs the amount of deviation calculated by the optimization unit to the movement control means of the treatment table for radiation therapy on which the patient is placed. Then, before starting the optimization calculation for optimizing the perspective projection parameter by the optimization unit, the coarse positional deviation amount of the patient is calculated from the deviation amount of the DRR image and the captured image, and the coarse positional deviation is calculated. It includes an initial parameter adjustment unit for changing the initial position of the perspective projection parameters based on the initial parameter adjustment section, the integral of the captured image with 1-dimensional integral profile integrating the DRR image in a predetermined direction in a predetermined direction The one-dimensional integration profile is calculated, and the coarse positional deviation amount is calculated by performing comparison between the DRR image and the one-dimensional integration profile in the same direction of the captured image .

In the second invention, the initial parameter adjustment unit calculates a similarity between histograms using the one-dimensional integral profile as a histogram in a comparison between the one-dimensional integral profiles in the same direction of the DRR image and the captured image. The position where the similarity between the histograms is the maximum is set as the initial position of the perspective projection parameter.

According to a third aspect of the present invention, a DRR image creating step for obtaining a DRR image by reproducing a geometric arrangement of an X-ray imaging system on a computer and virtually performing perspective projection on previously collected X-ray CT image data; In order to maximize the evaluation function for evaluating the degree of coincidence between the DRR image obtained in the DRR image creation step and the actual captured image obtained by seeing through or imaging the patient with an X-ray imaging system Optimizing perspective projection parameters related to rotation / translation of CT image data, the position of the patient when the X-ray CT image data is collected, and the actual patient's position when the patient is seen or imaged by the X-ray imaging system and optimization step of calculating a shift amount of the position, the shift amounts calculated in the optimization step, the output to the couch movement control means for radiotherapy mounted with the patient Patients A method of operating a positioning device, before starting the optimization operation for optimizing the perspective projection parameter by the optimization step, the shift amount of the captured image and the DRR image that includes an output step that the calculating a coarse position deviation amount, seen including an initial parameter adjustment step of changing the initial position of the perspective projection parameters based on the coarse position deviation amount, the initial parameter adjustment step, integrating said DRR image in a predetermined direction By calculating a one-dimensional integration profile obtained by integrating the captured image and a one-dimensional integral profile obtained by integrating the captured image in a predetermined direction, and performing a comparison between the DRR image and the captured image in the same direction. The coarse positional deviation amount is calculated .

In a fourth invention, the initial parameter adjustment step calculates a similarity between histograms using the one-dimensional integral profile as a histogram in a comparison between the one-dimensional integral profile in the same direction of the DRR image and the captured image. The position where the similarity between the histograms is the maximum is set as the initial position of the perspective projection parameter.

According to the first and third aspects of the invention, the initial position of the perspective projection parameter can be automatically obtained using the DRR image and the photographed image before the optimization unit starts the optimization calculation. As a result, it is possible to reduce the labor of the operator for manual positioning as in the prior art. In addition, since the rough displacement amount between the DRR image and the captured image is automatically calculated, for example, the conventional manual rough positioning adjustment step can be omitted even during the initial treatment in the divided irradiation treatment. , Treatment throughput can be improved. Furthermore, by adopting a configuration in which the operator's manual positioning operation can be omitted, it is possible to achieve highly reproducible patient positioning accuracy that does not depend on the operator's skill level, and it is necessary for patient positioning. Time can be shortened. And, since the patient's restraint time is shortened, it prevents the patient from maintaining the same posture during positioning and suppresses the deterioration of positioning accuracy, and the patient suffers from having to maintain the same posture. It becomes possible to reduce.

Further, according to the first and third inventions, since the comparison is made between the one-dimensional integral profiles in the same direction of the DRR image and the photographed image, the two-dimensional DRR image and the photographed image are directly compared. It is possible to search for a position having the highest similarity between images at a higher speed. Furthermore, the integration can reduce adverse effects on positioning accuracy due to noise components of each image.

According to the second and fourth inventions, since the histogram intersection is used for calculating the similarity between histograms using a one-dimensional integral profile as a histogram in comparison between a DRR image and a captured image, Search for a position that has the highest robustness and high similarity even with differences in image quality and differences and density differences caused by the presence of non-patient elements such as treatment tables and collimators in each image. Is possible.

1 is a schematic view of a radiation therapy apparatus 10 to which a positioning device 40 according to the present invention is applied. It is a flowchart which shows the procedure of the positioning of the patient 57. FIG. It is a flowchart which shows the procedure of the initial parameter adjustment before starting an optimization calculation. It is a schematic diagram explaining the one-dimensional integration profile of the pixel value of the X direction of the picked-up image 101, and a Y direction. 3 is a schematic diagram illustrating a one-dimensional integration profile of pixel values in the X direction and the Y direction of a DRR image 102. FIG. 3 is a graph showing a one-dimensional integration profile in the X direction of a captured image 101 and a DRR image 102; 4 is a graph showing a one-dimensional integration profile in the Y direction of a captured image 101 and a DRR image 102; FIG. 7 is a graph for explaining a normalized integration profile obtained by normalizing the graph of FIG. 6 and a histogram intersection HI. FIG. 8 is a graph for explaining a normalized integration profile and a histogram intersection HI obtained by normalizing the graph of FIG. 7. It is a graph which shows the change of the histogram intersection HI when the normalized integration profile of FIG. 8 is moved. It is a graph which shows the change of the histogram intersection HI when the normalized integration profile of FIG. 9 is moved. It is a graph which shows the normalization integral profile of the X direction after initial parameter adjustment. It is a graph which shows the normalization integration profile of the Y direction after initial parameter adjustment. It is a schematic diagram which shows the superimposed image 103 of the picked-up image 101 and DRR image 102 after initial parameter adjustment.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a schematic diagram of a radiotherapy apparatus 10 to which a positioning apparatus 40 according to the present invention is applied.

  This radiotherapy apparatus 10 performs radiotherapy for a patient 57 placed on a treatment table 27. The radiotherapy apparatus 10 includes a therapy control apparatus 30 that controls the operation of the entire therapy apparatus, and a positioning apparatus 40 according to the present invention. Prepare. The treatment control device 30 and the positioning device 40 are connected to be communicable with each other. Further, the treatment control device 30 and the positioning device 40 obtain three-dimensional X-ray CT image data including a patient DB (database) 16 for storing patient information and an affected part of the patient 57 before treatment via the network 17. The X-ray CT apparatus 14 and the treatment planning apparatus 15 for formulating a treatment plan based on the CT image data collected by the X-ray CT apparatus 14 are connected. Note that the CT image data collected by the X-ray CT apparatus 14 and the treatment plan for the patient 57 formulated by the treatment planning apparatus 15 are stored in the patient DB 16.

  The treatment control device 30 includes a CPU that executes logical operations, a ROM that stores an operation control program necessary for controlling the device, a RAM that temporarily stores data and the like during control, and the like. Control the whole. The treatment control device 30 includes a treatment beam irradiation control unit 31, a treatment table movement control unit 32, and an X-ray fluoroscopic imaging control unit 33 as functional configurations.

  The radiation therapy apparatus 10 includes a horizontal irradiation port 21 and a vertical irradiation port 22 that irradiate a therapeutic beam such as an electron beam. Treatment beam irradiation control from the horizontal irradiation port 21 and the vertical irradiation port 22 is performed by the treatment beam irradiation control unit 31. Further, the treatment table 27 of the radiation therapy apparatus 10 can be moved and rotated in six axis directions by the treatment table movement control unit 32.

  The radiotherapy apparatus 10 detects an X-ray detector 23 for detecting X-rays irradiated from the X-ray tube 25 and passed through the patient 57 and an X-ray irradiated from the X-ray tube 26 and passed through the patient 57. And an X-ray imaging system capable of fluoroscopic imaging and imaging from two directions. The operation of the X-ray imaging system is controlled by the X-ray fluoroscopic imaging control unit 33. As the X-ray detector 23 and the X-ray detector 24, for example, an image intensifier (II) or a flat panel detector (FPD) is used.

  In this radiotherapy apparatus 10, the horizontal irradiation port 21 and the vertical irradiation port 22 are fixed indoors. Then, the X-ray detector 24 is moved away from the horizontal irradiation port 21 and an imaging position on the front surface of the horizontal irradiation port 21 facing the X-ray tube 26 and the patient 57 (position indicated by a virtual line in FIG. 1). It can move between positions (positions indicated by solid lines in FIG. 1). Similarly, the X-ray detector 23 is spaced from the imaging position (position indicated by the phantom line in FIG. 1) of the vertical irradiation port 22 facing the X-ray tube 25 and the patient 57 from the vertical irradiation port 22. It can move between the retracted positions (positions indicated by solid lines in FIG. 1).

  The positioning device 40 is a computer that includes a CPU that executes logical operations, a ROM that stores a program for realizing positioning of a patient 57, which will be described later, a RAM that temporarily stores calculation results, and the like. 47 is connected. The display unit 47 also functions as a graphical user interface (GUI) of the radiotherapy apparatus 10, and includes, for example, a liquid crystal display panel having a touch panel function. The display unit 47 collects a fluoroscopic / photographed image (hereinafter referred to as a radiographed image) based on the X-rays detected by the X-ray detector 23 and the X-ray detector 24, and is collected in advance by the X-ray CT apparatus 14. A three-dimensional CT image is displayed. Further, the display unit 47 displays other information such as a DRR image created from CT image data, which will be described later, radiation irradiation information, and X-ray imaging information.

  The positioning device 40 has, as functional configurations, a DRR image creation unit 41 that creates a DRR image based on CT image data, an optimization unit 43 that optimizes a perspective projection parameter and calculates a positional deviation amount of a patient 57, and an optimum And an initial parameter adjustment unit 42 that changes the initial position of the perspective projection parameter at the start of optimization before the optimization unit 43 optimizes the perspective projection parameter related to the rotation / translation of the CT image data.

  The DRR image creation unit 41 reads out CT image data collected in advance by the X-ray CT apparatus 14 from the patient DB 16, reproduces the geometric arrangement of the X-ray imaging system on the computer, and applies this CT image data to the CT image data. A DRR image is created by virtually performing perspective projection. That is, it is assumed that the X-ray tube 25 of the radiotherapy apparatus 10 is irradiated with X-rays from the X-ray tube 25 and the X-ray tube 26 is irradiated with X-rays. A two-dimensional image is calculated. These two-dimensional images are DRR images, and are created by accumulating voxel values of each CT image data along a virtual X-ray path.

  The initial parameter adjustment unit 42 sets the initial projection geometry of the DRR image created by the first virtual perspective projection in the DRR image creation unit 41, that is, the initial position of the perspective projection parameter related to the rotation / translation of the CT image data. Adjustment is performed based on the coarse positional deviation amount calculated by comparing the image and the captured image. Then, the adjusted perspective projection parameter optimizes an evaluation scale (evaluation function) for evaluating the degree of coincidence between the two images so that the optimization unit 43 calculates a more accurate deviation amount between the DRR image and the captured image. As the initial parameter.

  Next, positioning of the patient 57 in the radiation therapy apparatus 10 described above will be described. FIG. 2 is a flowchart showing a procedure for positioning the patient 57. FIG. 3 is a flowchart showing a procedure for adjusting the initial parameters before the optimization calculation is started. FIG. 4 is a schematic diagram illustrating a one-dimensional integration profile of the captured image 101, and FIG. 5 is a schematic diagram illustrating a one-dimensional integration profile of the DRR image 102. 4 schematically shows a still image of the head of the patient 57 based on the detection value of the X-ray detector 23, and the head image of the patient 57 is shown by a solid line. . The DRR image 102 in FIG. 5 has an X-ray path from the X-ray tube 25 to the X-ray detector 23 shown in FIG. 1 with respect to the CT image data of the head of the patient 57 acquired in advance by the X-ray CT apparatus 14. A two-dimensional image obtained by executing DRR along the line is schematically shown, and a head image of the patient 57 is indicated by a broken line.

  When performing radiotherapy, the patient 57 is positioned by moving the treatment table 27 with respect to the horizontal irradiation port 21 and the vertical irradiation port 22 so that the treatment beam is accurately irradiated to the affected area. When positioning the patient 57, CT image data and other treatment plan information are acquired from the patient DB 16, and a DRR image 102 is created (step S1). This treatment plan information includes parameters such as the irradiation direction of radiation as a treatment beam.

  Subsequently, the patient 57 is irradiated with X-rays from the X-ray tube 25 and the X-ray tube 26 in a state where the X-ray detector 23 and the X-ray detector 24 are arranged at positions indicated by virtual lines in FIG. 57 current captured images 101 are acquired (step S2).

  Thereafter, before the optimization unit 43 starts the optimization operation for optimizing the perspective projection parameters, the DRR image 102 created in step S1 and the captured image 101 obtained in step S2 are used to perform fluoroscopy. An initial parameter which is an initial position of the projection parameter is adjusted (step S3). The captured image 101 shown in FIG. 4 and the DRR image 102 shown in FIG. 5 are displayed on the display unit 47 as necessary.

  The initial parameters are adjusted according to the procedure shown in FIG. First, a one-dimensional integration profile is created by integrating the captured image 101 and the DRR image 102 in predetermined directions (step S31). In FIGS. 4 and 5, the integration direction is indicated by an arrow. In these figures, a one-dimensional integration profile that is integrated in the horizontal direction (X direction) and the vertical direction (Y direction) of the image is calculated.

  FIG. 6 is a graph showing a one-dimensional integration profile in the X direction of the captured image 101 and the DRR image 102, and FIG. 7 is a graph showing a one-dimensional integration profile in the Y direction of the captured image 101 and the DRR image 102. The horizontal axis of the graph is a pixel, and the vertical axis is an integrated value. 6 and 7, the one-dimensional integration profile of the captured image 101 is indicated by a solid line, and the one-dimensional integration profile of the DRR image 102 is indicated by a broken line.

  As shown in FIGS. 6 and 7, a one-dimensional integration profile of a captured image 101 obtained by imaging a patient 57 on a treatment table 27 of the radiation therapy apparatus 10, and a one-dimensional integration profile of a DRR image 102 created from CT image data Then, it does not agree at all. This is because there is a difference in the position and angle between the patient 57 on the treatment table 27 of the radiotherapy apparatus 10 and the patient 57 when the CT image data is collected by the X-ray CT apparatus 14, and the captured image 101 and the DRR image. This is because there is a difference in image quality and a difference in density with respect to 102, and a difference due to the presence or absence of reflection in the image of members constituting the apparatus such as the treatment table 27.

  In this way, normalization for unifying the scale is performed on the one-dimensional integral profiles of the captured image 101 and the DRR image 102 that do not match each other. This normalization is performed so that each one-dimensional integral profile is regarded as a histogram, and the similarity between the one-dimensional integral profiles can be evaluated by a histogram intersection HI described later. That is, by the following equation (1), normalization is performed so that the area of the one-dimensional integration profile H1 [i] of the captured image 101 matches the area of the one-dimensional integration profile H2 [i] of the DRR image 102 (step S32). ).

  FIG. 8 is a graph for explaining a normalized integral profile obtained by normalizing the graph of FIG. 6 and the histogram intersection HI. FIG. 9 is a graph for explaining a normalized integration profile and a histogram intersection HI obtained by normalizing the graph of FIG. 8 and 9, the normalized integration profile of the captured image 101 is indicated by a solid line, and the normalized integration profile of the DRR image 102 is indicated by a broken line.

  This normalization also serves to equalize the difference in density between the captured image 101 and the DRR image 102. That is, on the one-dimensional integration profile, since the influence of the difference in image quality between images is small, it is possible to align the density of patient images with practical accuracy. Here, since the DRR image 102 is a simulation of the captured image 101 using CT image data, it is possible to adjust the density only by adjusting the gain of the X-ray dose without adjusting the density distribution width of the CT value. Is possible. If it is necessary to adjust the density distribution width of the CT value, the density between images may be made uniform by using a conventional WL (Window Level) / WW (Window Width) process.

  Next, a position where the one-dimensional integration profile after normalization in the same direction of the captured image 101 and the DRR image 102 (hereinafter referred to as a normalized integration profile) is most similar is searched. In this search, the similarity between both normalized integral profiles of the captured image 101 and the DRR image 102 is obtained by the following equation (2) using the histogram intersection HI (step S33).

  Due to the normalization described above, in FIG. 8, the area of the histogram of the captured image 101 surrounded by the solid line and the area of the histogram of the DRR image 102 surrounded by the broken line are the same. Similarly, in FIG. 9, the area of the histogram of the captured image 101 surrounded by the solid line and the area of the histogram of the DRR image 102 surrounded by the broken line are the same. As shown by hatching in FIGS. 8 and 9, the histogram intersection HI is the ratio of the area of the area where the histograms of the captured image 101 and the DRR image 102 overlap with the area of the histogram. Therefore, the histogram intersection HI takes a value of 0 or more and 1 or less, and the closer to 1, the higher the degree of coincidence of histograms.

  The position where the degree of coincidence of the histograms is the highest is the position where the normalized integration profiles of the captured image 101 and the DRR image 102 are most similar. And it can be estimated that this position is the position where both images overlap most. Therefore, a position where the degree of coincidence of histograms is highest, that is, a position where the histogram intersection HI is maximum is searched (step S34). The search in step S34 is performed by moving the position of the normalized integration profile of the other image along the horizontal axis of the graphs of FIGS. 8 and 9 with respect to the position of the normalized integration profile of one image. -This is done by calculating the intersection HI.

  FIG. 10 is a graph showing changes in the histogram intersection HI when the normalized integration profile of FIG. 8 is moved. FIG. 11 is a graph showing the histogram intersection when the normalized integration profile of FIG. 9 is moved. It is a graph which shows the change of section HI. The horizontal axis of the graph is the number of moving pixels, and the vertical axis is the histogram intersection HI.

  When the position of the normalized integral profile of the other image (DRR image 102) is shifted in the horizontal axis direction of the graphs of FIGS. 8 and 9 with respect to the position of the normalized integral profile of one image (captured image 101) The histogram intersection HI is as shown in FIGS. In the graph shown in FIG. 10, since the peak of the histogram intersection HI is clear, the position of the normalized integration profile where the histogram intersection HI is maximum can be easily searched. The calculation of the histogram / intersection HI using the normalized integration profile can sufficiently reduce the calculation cost compared to the case of using the image itself. Therefore, for the search for the maximum value of the histogram / intersection HI, exhaustive calculation may be performed, or a one-dimensional optimization method such as the Brent method may be used.

  On the other hand, as in the change of the histogram / intersection HI shown in FIG. 11, it may be difficult to search for the position of the normalized integration profile where the peak is flat and the histogram / intersection HI is maximum. For example, when there is a difference in the reflection of the treatment table 27 or the like between images to be compared, the histogram intersection HI changes as shown in FIG. Thus, when there are a plurality of maximum values of the histogram intersection HI, the center of the flat portion may be set as the position of the maximum value.

  When the search for the maximum value of the histogram intersection HI is completed, the position at which the histogram intersection HI is maximum is set as the initial position of the perspective projection parameter (step S35).

  FIG. 12 is a graph showing normalized integral profiles in the X direction of the captured image 101 and the DRR image 102 at positions where the histogram intersection HI is maximum. FIG. 13 is a graph showing normalized integration profiles in the Y direction of the captured image 101 and the DRR image 102 at the position where the histogram intersection HI is maximum. FIG. 14 is a schematic diagram showing a superimposed image 103 of the captured image 101 and the DRR image 102 after adjusting the initial parameters at the position where the histogram intersection HI shown in FIGS. 12 and 13 is maximum. In FIG. 14, the head image of the patient 57 derived from the captured image 101 is indicated by a solid line, and the head image of the patient 57 derived from the DRR image 102 is indicated by a broken line.

  As shown in FIG. 12, the normalized integral profiles in the X direction of the captured image 101 and the DRR image 102 are similar to each other. As shown in FIG. 13, the normalized integral profiles in the Y direction are also similar to each other. It should be noted that the normalized integral profile of the captured image 101 and the normalized integral profile of the DRR image 102 are different from each other due to differences in the position and angle of the patient 57 and the like, so the histograms do not match completely. . That is, the position where the normalized integration profiles are the most similar does not always exactly match the position where the captured image 101 and the DRR image 102 overlap. However, the purpose of adjusting the initial parameters (step S3) is to set the perspective projection geometry to an actual X in order to prevent the solution at the time of optimization calculation from falling into a local solution in the subsequent optimization (step S4). It is to be closer to the geometry of the radiography system. That is, from the viewpoint of adjusting the initial parameters before starting the optimization calculation, the normalized integral profiles of the captured image 101 and the DRR image 102 obtained by the above-described procedure (steps S31 to S34) are the most similar. It can be said that setting the position to be used as the initial position before the optimization calculation is in a sufficiently practical category as the accuracy of the coarse adjustment of the position of the patient 57.

  In the example with reference to FIGS. 4 to 14 described above, the captured image 101 and the DRR image 102 are 2 in the XY plane with the X-ray irradiation axis from the X-ray tube 25 to the X-ray detector 23 as the image center. Although the case of a three-dimensional image has been described, a two-dimensional image on the YZ plane with the X-ray irradiation axis from the X-ray tube 26 to the X-ray detector 24 as the image center is also the initial stage before optimization calculation by the same procedure. The position can be determined automatically.

  When the adjustment of the initial parameters is completed, the superimposed image 103 of the captured image 101 and the DRR image 102 shown in FIG. 14 is displayed on the display unit 47 as the image after the coarse position adjustment. Then, optimization of the perspective projection parameters is executed, and a positional deviation amount between the current position of the patient 57 and the position of the patient 57 at the time of obtaining the CT image data is calculated (step S5). The displacement amount is converted into a movement amount of the treatment table 27, and the treatment table 27 is moved by the action of the treatment table movement control unit 32 (step S6).

  After the treatment table 27 moves, a captured image is acquired again, and the operator visually confirms the image displayed on the display unit 47, and the amount of deviation between the captured image 101 and the DRR image 102 indicates the treatment beam irradiation position. Confirmation by automatic determination as to whether or not the error is within the allowable range is executed (step S7). After confirming that the patient 57 has been properly positioned, the X-ray detector 23 and the X-ray detector 24 are moved to the positions indicated by the solid lines in FIG. 1 to irradiate the affected area of the patient 57 with the treatment beam. Radiotherapy is performed (step S8).

  Conventionally, the DRR image is created in the first step of the optimization calculation. In the above-described embodiment, the DRR image 102 is created before the adjustment of the initial parameters. The DRR image 102 can be used as it is for the optimization calculation.

  In the embodiment described above, an example in which the positioning device 40 is applied to the radiotherapy device 10 has been described. However, for example, the positioning device of the present invention is applied to an X-ray fluoroscopic table used when performing cardiac catheter therapy. It is also possible to match the three-dimensional image of the heart previously acquired by the X-ray CT apparatus with the current patient position.

The present invention relates to a positioning device for positioning a patient and a method for operating the positioning device in the field of radiation therapy in which radiation such as X-rays, electron beams, and particle beams is applied to an affected area such as cancer. With the above applicability.

DESCRIPTION OF SYMBOLS 10 Radiotherapy apparatus 14 X-ray CT apparatus 15 Treatment planning apparatus 16 Patient DB
17 Network 21 Horizontal irradiation port 22 Vertical irradiation port 23 X-ray detector 24 X-ray detector 25 X-ray tube 26 X-ray tube 27 Treatment table 30 Treatment control device 31 Treatment beam irradiation control unit 32 Treatment table movement control unit 33 X-ray Fluoroscopic imaging control unit 40 Positioning device 41 DRR image creation unit 42 Initial parameter adjustment unit 43 Optimization unit 47 Display unit 57 Patient 101 Captured image 102 DRR image 103 Superposed image

Claims (4)

A DRR image creation unit that reproduces the geometric arrangement of the X-ray imaging system on a computer and obtains a DRR image by virtually performing perspective projection on previously collected X-ray CT image data;
The X-ray CT so that the evaluation function for evaluating the degree of coincidence between the DRR image obtained by the DRR image creation unit and the actual captured image obtained by fluoroscopy or imaging the patient with an X-ray imaging system is maximized. Optimizing perspective projection parameters relating to rotation / translation of image data and acquiring the X-ray CT image data, and the actual patient position when the patient is seen or imaged by an X-ray imaging system An optimization unit for calculating a deviation amount from
A positioning device that outputs a shift amount calculated by the optimization unit to a movement control means of a treatment table for radiation therapy on which a patient is placed ,
Before starting the optimization operation for optimizing the perspective projection parameter by the optimization unit, the patient's coarse positional deviation amount is calculated from the deviation amount between the DRR image and the captured image, and based on the coarse positional deviation amount An initial parameter adjusting unit for changing the initial position of the perspective projection parameter ,
The initial parameter adjustment unit calculates a one-dimensional integration profile obtained by integrating the DRR image in a predetermined direction and a one-dimensional integration profile obtained by integrating the photographed image in a predetermined direction, respectively, and the same direction of the DRR image and the photographed image A positioning device that calculates the coarse positional deviation amount by executing a comparison between the one-dimensional integral profiles . The positioning device according to claim 1 ,
The initial parameter adjustment unit calculates a similarity between histograms using the one-dimensional integral profile as a histogram in a comparison between the one-dimensional integral profiles in the same direction of the DRR image and the captured image, and the similarity between the histograms A positioning apparatus in which a position where the degree is maximum is an initial position of the perspective projection parameter. A DRR image creation step of obtaining a DRR image by reproducing a geometric arrangement of an X-ray imaging system on a computer and virtually performing perspective projection on previously collected X-ray CT image data;
The X-ray CT so that the evaluation function for evaluating the degree of coincidence between the DRR image obtained in the DRR image creation step and the actual captured image obtained by fluoroscopy or imaging the patient with an X-ray imaging system is maximized. Optimizing perspective projection parameters relating to rotation / translation of image data and acquiring the X-ray CT image data, and the actual patient position when the patient is seen or imaged by an X-ray imaging system An optimization step for calculating the amount of deviation from
An output step of outputting the shift amount calculated in the optimization step to a movement control means of a treatment table for radiation therapy on which a patient is placed;
A method of operating a positioning device comprising :
Before starting the optimization operation for optimizing the perspective projection parameter by the optimization step, the patient's coarse positional deviation amount is calculated from the deviation amount between the DRR image and the captured image, and based on the coarse positional deviation amount. the initial parameter adjustment step of changing the initial position of the perspective projection parameter Te seen including,
The initial parameter adjustment step calculates a one-dimensional integration profile obtained by integrating the DRR image in a predetermined direction and a one-dimensional integration profile obtained by integrating the photographed image in a predetermined direction, respectively, and the same direction of the DRR image and the photographed image. A method for operating a positioning device, wherein the coarse positional deviation amount is calculated by performing a comparison between the one-dimensional integral profiles of the two . The method of operating a positioning device according to claim 3 ,
The initial parameter adjustment step calculates similarity between histograms using the one-dimensional integral profile as a histogram in the comparison between the one-dimensional integral profiles in the same direction of the DRR image and the captured image, and the similarity between the histograms An operation method of a positioning device, wherein a position where the degree is maximum is an initial position of the perspective projection parameter.

Images