JP7577591B2 - Positioning device, particle beam therapy device, and positioning method - Google Patents

Description

The present invention relates to a positioning device, a particle beam therapy device, and a positioning method.

One method of treating cancer is radiation therapy, in which radiation is irradiated onto the patient. The radiation used in radiation therapy is broadly divided into uncharged particle beams, such as X-rays and gamma rays, and charged particle beams, such as proton beams and carbon beams. Treatment using the latter type of charged particle beams (hereafter also referred to as charged particle beams) is generally called particle beam therapy.

The dose delivered by uncharged particle beams decreases at a constant rate from shallow to deep positions inside the body. On the other hand, charged particle beams can form a dose distribution (Brugh curve) with an energy peak at a specific depth. Therefore, by aligning the peak with the position of the tumor, it is possible to significantly reduce the dose delivered to normal tissue deeper than the tumor.

In radiation therapy, irradiating the targeted tumor with the desired dose of a charged particle beam as accurately as possible leads to improved therapeutic effects. In order to accurately irradiate the tumor with a charged particle beam, it is necessary to create a treatment plan in advance using a treatment planning system, and then align the patient at the same position during actual irradiation as in the treatment plan. This alignment of the patient is called patient positioning.

One method of patient positioning in radiation therapy is to use fluoroscopic X-ray images (Digital Radiography: DR) of a patient taken from two directions using two sets of X-ray tubes and flat panel detectors arranged perpendicular to each other. The fluoroscopic X-ray image of the patient is compared with a projection processing image (hereinafter referred to as a pseudo-fluoroscopic X-ray image) created by placing a CT image at the time of treatment planning in a virtual space and calculating the amount of attenuation of X-rays entering the virtual flat panel detector from the virtual X-ray tube, and the amount of displacement required to match the patient's position in both images is determined. Here, the amount of displacement is determined so that a structure that serves as a landmark for positioning on the fluoroscopic X-ray image (hereinafter referred to as the structure to be positioned) matches the same part on the pseudo-fluoroscopic X-ray image. Generally, bones are used as landmark structures. The amount of displacement is determined by manual alignment by visual inspection by a medical professional on an image matching device, or automatic alignment using an automatic calculation function.

In general, fluoroscopic X-ray images capture structures other than the object being positioned, such as the patient's fixation devices and soft tissues, and even bones, which are the structures being positioned, may change position from day to day. In such a situation, it is expected that automatic calculation of positioning will not be accurate enough, since the structures of the fluoroscopic X-ray image and the pseudo-fluoroscopic X-ray image do not match across the entire image.

Therefore, in order to perform automatic calculation alignment that targets only the structure to be positioned on the fluoroscopic X-ray image, a medical professional sets in advance the area on the fluoroscopic X-ray image or pseudo-fluoroscopic X-ray image where the structure to be positioned exists as a Region of Interest (ROI), and during automatic alignment, only the area within the ROI is calculated as the calculation target, thereby performing alignment that focuses only on the structure within the ROI. The process of setting the ROI on an image is usually performed by a user who is a medical professional by drawing the ROI on the image.

Patent Document 1 describes a medical imaging device that uses a region of interest set as a dose evaluation target when formulating a treatment plan to calculate the degree of influence that radiation irradiated to a patient will have on the region of interest before it reaches the treatment target area inside the body, and generates a display image on which information on the degree of influence is superimposed to highlight and display areas with a high degree of influence. In addition, Patent Document 2 describes a radiation therapy system that uses a region of interest set in a treatment plan image to calculate feature amounts within the region of interest between a tomographic image acquired for positioning and the treatment plan image, and performs positioning calculations.

Patent Document 2 also describes a positioning system that calculates feature values related to the position of a region of interest set by a treatment planning device for treatment planning CT images and 3D tomographic images captured for bed positioning, and performs machine learning using these feature values as weak classifiers. Furthermore, it describes a radiation therapy system that determines the parts and regions to be aligned and the degree of alignment (weighting), creates an evaluation formula for positioning using the parts and regions and weighting determined by learning, and performs positioning using this evaluation formula.

JP 2019-162379 A JP 2014-212820 A

The technology disclosed in Patent Document 1 is effective in that it allows the effect on radiation dose to be visually confirmed.

However, the region of interest described here is a three-dimensional region set on a three-dimensional image (usually CT is used) for dose assessment of the irradiation target and surrounding dangerous organs during treatment planning, and is different from the bones used for positioning shown on the fluoroscopic image, and is different from the region of interest used for automatic positioning calculations.

Similarly, in the technology described in Patent Document 2, the region of interest is set as a three-dimensional region for dose assessment of the irradiation target and surrounding dangerous organs, and is used for automatic calculations; this differs from the region of interest that is set on a two-dimensional image for the bone to be positioned when positioning is performed using fluoroscopic X-ray images and pseudo-fluoroscopic X-ray images.

If these technologies could be used for automatic calculations by setting an ROI along the contour of the target bone (hereafter also referred to as the bone contour) on a pseudo-fluoroscopic X-ray image or a fluoroscopic X-ray image and weighting the ROI according to the interest level of the ROI, it would be possible to perform automatic calculations that increase the importance of the target bone relative to other structures that appear within the ROI.

However, the task of having a medical professional draw an ROI along the bone contour in advance and set weighting coefficients according to the degree of interest within the ROI means that the setting of areas and coefficients depends on the medical professional's experience, and drawing the ROI takes a long time. Furthermore, even if an ROI is drawn over a long period of time, it may be necessary to redraw the ROI depending on the degree of deformation of the target bone of the patient on a daily basis.

The present invention has been made in consideration of the above problems, and aims to provide a positioning device, particle beam therapy device, and positioning method that facilitates highly accurate patient positioning using conventional ROI setting methods.

In order to solve the above problem, a positioning device according to one aspect of the present invention is a positioning device that controls the position of a bed on which a patient is placed, creates a pseudo-fluoroscopic X-ray image by projecting a three-dimensional image used to create a treatment plan for irradiating a patient with radiation in radiation therapy onto a specified surface, obtains a first region of interest including the region of the patient's target structure in the pseudo-fluoroscopic X-ray image, calculates two-dimensional similarity information between the fluoroscopic X-ray image and the pseudo-fluoroscopic X-ray image based on the fluoroscopic X-ray image taken of the patient by X-ray fluoroscopy, the pseudo-fluoroscopic X-ray image, and the first region of interest, creates a second region of interest by applying a weighting coefficient within the first region of interest based on the two-dimensional similarity information, and calculates the amount of movement for moving the bed so that the second region of interest, the position of the target structure in the fluoroscopic X-ray image, and the position of the target structure in the pseudo-fluoroscopic X-ray image match.

The present invention makes it possible to realize a positioning device, particle beam therapy device, and positioning method that can easily perform highly accurate patient positioning using conventional ROI setting methods.

1 is a diagram showing an overall configuration of a particle beam therapy system according to a first embodiment; 11 is a flowchart showing a weighting ROI creation and update process of the positioning device according to the first embodiment. 11 is a diagram illustrating how weighting coefficients are assigned based on two-dimensional similarity information in the positioning device according to the first embodiment. FIG. 11A and 11B are diagrams illustrating a process of updating a weighted ROI in the positioning device according to the first embodiment. FIG. 11 is a diagram showing a flow of creating a weighted ROI using 4DCT in the positioning device according to the second embodiment.

The following describes an embodiment of the present invention with reference to the drawings.

Note that the following description and drawings are illustrative for explaining the present invention, and are omitted and simplified as appropriate for clarity of explanation. The present invention can be implemented in various other forms. Unless otherwise specified, each component may be singular or plural. In addition, in the drawings explaining the embodiment, the same reference numerals are given to parts having the same functions, and repeated explanations are omitted. In addition, the position, size, shape, range, etc. of each component shown in the drawings may not represent the actual position, size, shape, range, etc., in order to facilitate understanding of the invention. Therefore, the present invention is not limited to the position, size, shape, range, etc. disclosed in the drawings. In addition, when there are multiple identical or similar components, they may be described by adding different subscripts to the same reference numerals. However, when it is not necessary to distinguish between these multiple components, the subscripts may be omitted.

The particle beam therapy system of this embodiment is a group of devices for irradiating a particle beam to a patient, which is a target. The particle beam therapy system irradiates the patient with a particle beam after positioning the patient. In positioning the patient, the particle beam therapy system creates a weighted ROI from two-dimensional similarity information calculated using a pseudo fluoroscopic X-ray image (digitally reconstructed radiography: DRR) created by projecting three-dimensional patient image information, a ROI drawn in advance, and a fluoroscopic X-ray image taken on the day of treatment. Furthermore, the particle beam therapy system aligns the fluoroscopic X-ray image and the pseudo fluoroscopic X-ray image using the weighted ROI.

The positioning device used in the particle beam therapy system of this embodiment includes a pseudo-fluoroscopic X-ray image creation unit that creates a pseudo-fluoroscopic X-ray image by projecting a three-dimensional image used to create a treatment plan for irradiating a patient with radiation in radiation therapy onto a specified surface; a region of interest drawing unit that sets an area to be aligned on the pseudo-fluoroscopic X-ray image as an ROI; a two-dimensional similarity information calculation unit that calculates two-dimensional similarity information within the ROI between a fluoroscopic X-ray image of a patient taken by X-ray fluoroscopy and the pseudo-fluoroscopic X-ray image; a region of interest weighting coefficient calculation unit that weights the ROI to create a weighted ROI; an image matching unit that calculates the amount of movement required to move the bed using the pseudo-fluoroscopic X-ray image, the X-ray fluoroscopic image, and the weighted ROI so that the position of the target structure in the fluoroscopic X-ray image matches the position of the target structure in the pseudo-fluoroscopic X-ray image; and an image display unit that displays the result.

This embodiment is described in detail below.

Figure 1 is a diagram showing the overall configuration of a particle beam therapy system according to the first embodiment.

The particle beam therapy system A includes an accelerator 1, a beam transport device 2, a gantry 3, an irradiation nozzle 4, flat panel detectors 5A and 5B, X-ray tubes 6A and 6B, a bed 7, a robot arm 8, a treatment planning device 10, a communication device 11, a data server 12, a fluoroscopic X-ray imaging device 14, a bed control device 15, and a patient positioning device 20.

The patient positioning device (hereinafter simply referred to as the positioning device) 20 includes a pseudo-fluoroscopic X-ray image creation unit 21, an ROI drawing unit 22, a two-dimensional similarity information calculation unit 23, an ROI weighting processing unit 24, an image matching unit 25, and an image display unit 26.

When positioning the patient before particle beam therapy, the patient 9 is placed on the bed 7 and moved to a planned position planned in advance by the robot arm 8 connected to the bed 7. The planned position here refers to a position of the patient 9 in the radiation therapy room that reproduces the same patient position as in the treatment plan created in advance. The robot arm 8 can be driven in three translational directions and three rotational directions, and can position the bed 7 with the patient 9 on it at an appropriate position and angle. The procedure up to this point corresponds to patient positioning in particle beam therapy.

The particle beam used for treatment is generated in an accelerator 1, accelerated to an energy level suitable for treatment, and transported to a gantry 3 by a beam transport device 2. The gantry 3 has a rotation mechanism and is configured to be able to rotate 360° so that the particle beam can be irradiated to the affected area of the patient at any angle. In the gantry 3, the particle beam is deflected in an appropriate direction, passes through an irradiation nozzle 4, and is irradiated to the affected area of the patient 9.

The irradiation nozzle 4 incorporates a mechanism that changes the shape of the particle beam to match the shape of the affected area on the patient.

The treatment planning device 10 is a device that creates a treatment plan based on three-dimensional image information of the patient 9 before actually irradiating the patient with particle beams. In the treatment plan, the treatment planning device 10 acquires three-dimensional image information of the patient 9 from the data server 12 via the communication device 11, calculates the appropriate irradiation angle, irradiation shape, and irradiation amount of the particle beam for the three-dimensional image, and determines it as irradiation information. The determined irradiation information is stored in the data server 12 via the communication device 11.

The X-ray fluoroscopy imaging device 14 controls the flat panel detector 5A and the X-ray tube 6A, and the flat panel detector 5B and the X-ray tube 6B, which are arranged in mutually orthogonal directions, to obtain a fluoroscopy X-ray image. The obtained fluoroscopy X-ray image is sent to the positioning device 20.

The positioning device 20 is a device that calculates the amount of movement required to move the bed 7 from the fluoroscopic X-ray image and the pseudo-fluoroscopic X-ray image so that the position of the patient 9 when the treatment plan was created coincides with the position of the patient 9 placed on the bed 7 for irradiating the particle beam.

The positioning device 20 acquires irradiation information created during treatment planning and three-dimensional image information of the patient 9 from the data server 12 via the communication device 11.

The pseudo-fluoroscopic X-ray image creation unit 21 of the positioning device 20 creates a pseudo-fluoroscopic X-ray image by placing a three-dimensional image of the patient 9 in a virtual space that is the same as the actual fluoroscopic X-ray imaging system and performing a projection process.

The three-dimensional image information here includes information that indicates the shape and electron density of the patient in voxel units. Computed tomography (CT) images are generally used.

The ROI drawing unit 22 allows the user to draw on the screen as an ROI an area including a bone to be used for alignment on the pseudo-fluoroscopic X-ray image. The ROI that the user draws here may be created as an area drawn to include the structure to be positioned.

The two-dimensional similarity information calculation unit 23 calculates two-dimensional similarity information between the fluoroscopic X-ray image and the pseudo fluoroscopic X-ray image in the ROI. The two-dimensional similarity information is information obtained by converting a similarity index, which is generally used when calculating the similarity between images, into two dimensions and displaying it.
As an example of two-dimensional similarity information, a case where a zero-mean normalized cross correlation (ZNCC) coefficient is used as a similarity index will be described. The similarity of ZNCC can be calculated by formula (1). Here, g(i,j) is the DR image, f(i,j) is the DRR image, μ _g is the average luminance value of the DR image, and μ _f is the average luminance value of the DRR image. The numerator of formula (1) is a value obtained by calculating the difference between the average luminance value of the image for each pixel of each image and adding up the product for all pixels, so that formula (1) can be decomposed as a formula representing the similarity for each pixel and expressed as two-dimensional similarity information as in formula (2). This is used as two-dimensional similarity information in the present invention. When the similarity of each pixel calculated by formula (2) is expressed as an image, a part with a high degree of matching on the image is displayed as a high numerical value.

The ROI weighting processing unit 24 assigns a weighting coefficient to each pixel of the ROI based on the two-dimensional similarity information within the ROI calculated by the two-dimensional similarity information calculation unit 23 .

The image matching unit 25 performs image matching between the fluoroscopic X-ray image and the pseudo fluoroscopic X-ray image only in the portion included in the ROI, and calculates the amount of displacement with six degrees of freedom required to match the patient's position in both images. The amount of displacement calculated here is used by the positioning device 20 to control the bed control device 15 to move the bed 7. The image display unit 26 displays the matching image, the ROI image, and the two-dimensional similarity information.

The positioning device 20 is composed of a device capable of various information processing, for example an information processing device such as a computer. The information processing device has a processing element, a storage medium, and a communication interface, and further has an input unit such as a mouse and a keyboard, and a display unit such as a display, as necessary.

The computing element may be, for example, a CPU (Central Processing Unit) or an FPGA (Field-Programmable Gate Array).

The storage medium may be, for example, a magnetic storage medium such as a hard disk drive (HDD), or a semiconductor storage medium such as a random access memory (RAM), a read only memory (ROM), or a solid state drive (SSD). As a storage medium, a combination of an optical disk such as a digital versatile disk (DVD) and an optical disk drive may also be used. Furthermore, as a storage medium, other well-known storage media such as magnetic tape media may also be used.

The storage medium stores programs such as firmware. When the positioning device 20 starts to operate (for example, when the power is turned on), the computing element reads the programs from the storage medium and executes them to realize each of the sections 21 to 25 of the positioning device 20 and execute a series of controls over the entire device. In addition to the programs, the storage medium also stores data and the like required for each process of the positioning device 20.

In addition, the positioning device 20 of this embodiment may be configured using so-called cloud computing, in which information processing devices are configured to be able to communicate via a communication network.

In the particle beam therapy system A of this embodiment, an ROI is created in advance before patient positioning for particle beam therapy. The method of creating a weighted ROI and the method of patient positioning of this embodiment will be described below with reference to Figures 2 to 5.

Figure 2 is a flowchart showing the creation and update process of weighting ROI. Figure 3 is a diagram showing the assignment of weighting coefficients based on two-dimensional similarity information. Figure 4 is a diagram showing weighting ROIs that are updated sequentially over the course of treatment over multiple days. Figure 5 is a diagram showing the creation flow of weighting ROIs using 4DCT.

Note that for ease of explanation, Figures 3 and 4 only show a pseudo-fluoroscopic X-ray image from one direction, but in reality there are pseudo-fluoroscopic X-ray images from two directions, and the weighted ROI creation process is applied to the images from each direction.

As shown in FIG. 2, first, the pseudo-fluoroscopic X-ray image creation unit 21 of the positioning device 20 creates a pseudo-fluoroscopic X-ray image (Step 100).

Next, the ROI drawing unit 22 sets the area on the pseudo fluoroscopic X-ray image 201 where the structure to be used for positioning exists as the ROI 50 based on the user's operation (Step 101). The above processing is the work in advance of preparation for treatment.

As a method of setting the ROI 50, the ROI drawing unit 22 displays the pseudo-fluoroscopic X-ray image 201 on the screen, has the user fill in the area to be the ROI 50 by operating the mouse or the like, and sets the filled area as the ROI 50. There is also a method in which the ROI drawing unit 22 displays the pseudo-fluoroscopic X-ray image on the screen, has the user select a position in the area to be the ROI 50 by operating the mouse or the like, and sets an area within a predetermined distance from the selected position as the ROI 50. There is also a method in which the ROI drawing unit 22 displays the pseudo-fluoroscopic X-ray image 201 on the screen, has the user select a position in the area to be the ROI 50 by operating the mouse or the like, and sets a continuous area having pixel values within a certain range from the pixel value of the selected position as the ROI 50. The image display unit 26 displays the ROI 50 created by the user's drawing in the ROI drawing unit 22 on the screen, superimposed on the pseudo-fluoroscopic X-ray image.

The steps involved in positioning during treatment are explained below.

The image matching unit 25 of the positioning device 20 reads the ROI 50 created by the ROI drawing unit 22 (Step 102). After that, the X-ray fluoroscopy imaging device 14 acquires a fluoroscopy X-ray image of the patient 9 (Step 103).

The acquired fluoroscopic X-ray image is transmitted to the patient positioning device, and the image matching unit 25 automatically performs calculations for matching the position of the pseudo fluoroscopic X-ray image with the ROI (Step 104). Here, the normalized cross-correlation coefficient and mutual information, which are commonly used indices for measuring similarity, are used to calculate the amount of displacement of the six degrees of freedom (three translational directions and three rotational directions) of the pseudo fluoroscopic X-ray image or the fluoroscopic X-ray image required for the two images to match by optimization calculations. Other indices may also be used for the similarity.

It is determined whether the calculated similarity satisfies the convergence conditions set in advance, and if the degree of similarity does not satisfy the convergence conditions, a condition that satisfies the convergence conditions (a value of six degrees of freedom in total, three translational degrees of freedom and three rotational degrees of freedom) is searched for by optimization processing. Methods of optimization processing for multiple variables such as six degrees of freedom include the downhill simplex method and Powell's method, which can be used in this embodiment. However, the optimization processing is not limited to these, and other methods may also be used.

Then, the user visually checks whether the fluoroscopic X-ray image and the pseudo-fluoroscopic X-ray image within the ROI displayed on the screen by the image display unit 26 of the patient positioning device match (Step 105), and if they do not match, the user manually aligns the images so that they match (Step 106).

Then, the two-dimensional similarity information calculation unit 23 of the positioning device 20 calculates two-dimensional similarity information of both images (Step 107). Then, the ROI weighting processing unit 24 assigns a weighting factor to the ROI based on the calculated two-dimensional similarity information (Step 108).

An example of a method for assigning weighting coefficients to an ROI will be described with reference to FIG. 3. If image 100 is an image in which ROI 50 is superimposed on a pseudo fluoroscopic X-ray image, then by calculating two-dimensional similarity information within the ROI between image 100 and the fluoroscopic X-ray image, a two-dimensional similarity image with high similarity in areas with high coincidence can be created.

Here, the two-dimensional similarity image is superimposed on the pseudo-fluoroscopic X-ray image and displayed as image 101. In this example image 101, the target bone is the spine, and an image showing high similarity is created along the contour of the spine. Image 102 is created by enlarging the square area surrounded by ROI 50 in image 101 with dotted line 52, and a coefficient is assigned to each pixel to reflect the size of the two-dimensional similarity information.

For ease of explanation, the area on this image to which coefficients are assigned is shown as a grid area larger than the pixel size of the actual pseudo-fluoroscopic X-ray image. In reality, however, the pixel size when weighting coefficients are assigned to pixels within ROI 50 from the two-dimensional similarity information may be set to match the pixel size of the pseudo-fluoroscopic X-ray image, or may be set to a larger size.

Here, the weighting coefficients assigned to the pseudo-fluoroscopic X-ray image 102 in a grid pattern are shown as an image to which a coefficient is assigned: 0 outside the ROI, 1 within the ROI, 2 if the two-dimensional similarity is within the ROI and is medium, and 3 if it is high. The coefficient information within the ROI in the image is used for automatic calculation as a weighted ROI image.

Ｓｔｅｐ１０８で作成された重み付けＲＯＩ画像は、次回治療における位置決め前の疑似透視Ｘ線画像、ＲＯＩ情報の読み込み（Ｓｔｅｐ１０２）の際に新たなＲＯＩ画像として使用される。重み付けＲＯＩ画像が治療期間中に日々の位置決め時に用いられ、更新された重み係数を日ごとに積算した重み付けＲＯＩ画像を位置決めに用いることで、治療期間中に配置の変形のあるＲＯＩ内の背骨領域は徐々に低い重み係数が付与されて、変形の少ない背骨領域のみに注目した位置決めが可能になる。 As an example of calculating the similarity between images using weighted ROI images, when the normalized cross-correlation coefficient shown in formula (1) is used as the similarity index, the calculation formula for the similarity can be expressed as follows: Here, w(i,j) means the weighting coefficient at pixel (i,j). When the two-dimensional similarity information is set to a size different from the pixel size of the actual pseudo fluoroscopic X-ray image, the two-dimensional similarity information is interpolated to calculate and use the weighting coefficient w(i,j) at pixel (i,j).

The weighted ROI image created in Step 108 is used as a new ROI image when reading the pseudo fluoroscopic X-ray image and ROI information before positioning in the next treatment (Step 102). The weighted ROI image is used for daily positioning during the treatment period, and the weighted ROI image obtained by integrating the updated weighting coefficients each day is used for positioning. By using this weighted ROI image, the spinal region in the ROI that has been deformed in positioning during the treatment period is assigned a gradually lower weighting coefficient, making it possible to perform positioning that focuses on only the spinal region with little deformation.

An example of weighted ROI image update is explained using Figure 4.

Image 200, in which ROI 50 is drawn on a pseudo-fluoroscopic X-ray image, becomes an ROI with a high coefficient assigned along the contour of the target bone, the spine, as shown in image 201, at the end of positioning on the first day of treatment. When positioning is performed in a deformed state at the top of the spine within the ROI on the Nth day of treatment and 2D similarity information is calculated, as shown in image 202, a lower weighting coefficient is assigned to the contour of spine 53 at the top of the image within the ROI than to the other spine within the ROI. In this way, by accumulating and updating the weight within the ROI at any time during the treatment period, positioning can be performed with attention focused only on the target bone with the least deformation.

Therefore, this embodiment makes it easy to position the patient with high accuracy using conventional ROI setting methods.

In the above-mentioned first embodiment, the creation and updating of the weighting ROI begins when positioning on the first day is completed, but if a four-dimensional CT (hereinafter referred to as 4DCT) is acquired for treatment planning, it is possible to create the weighting ROI before the start of treatment. 4DCT here refers to CT images acquired for each respiratory phase by dividing one respiratory cycle of the patient into 10 phases (T00 to T90). Generally, of the 10 CT phases, the CT image for the expiratory phase (T40), in which respiratory movement is stable, is used as the CT for treatment planning.

Below, we will use Figure 5 to explain how to create a weighted ROI using 4DCT.

First, pseudo fluoroscopic X-ray images I _T00 -I _T90 are created for each respiratory phase CT (T00 to T90) of the 4D CT (Step 200).

Next, the user draws an ROI on the pseudo-fluoroscopic X-ray image I _T40 created by the CT _T40 used in the treatment plan (Step 201). Next, two-dimensional similarity information for I _T00 -I _{T30 and} I _T50 -I _T90 for I _T40 is calculated (Step 202). Finally, a weighting coefficient within the ROI is assigned using information obtained by accumulating each two-dimensional similarity information (Step 203).

By using the weighted ROI created by the above process from the first day of treatment, the weighting coefficient for the area where the edge of soft tissue that moves due to breathing or heartbeat, such as the diaphragm or heart wall, included in the ROI overlaps with the bone to be positioned can be set small in advance, preventing the occurrence of alignment failures when aligning the edges of the soft tissue in the fluoroscopic X-ray image and the pseudo-fluoroscopic X-ray image during automatic alignment calculation.

In each of the above-described embodiments, the pseudo-fluoroscopic X-ray image and the fluoroscopic X-ray image are used as is when calculating the two-dimensional similarity information, but in order to emphasize the contour of the bone to be positioned and calculate the two-dimensional similarity information, the two-dimensional similarity information may be calculated using images that have been subjected to edge collaboration processing on both images.

The edge enhancement process is not particularly limited, but for example, a differential filter may be applied to each of the vertical and horizontal directions as a general edge enhancement filter. A Prewitte filter or a Sobel filter may be used as a differential filter. Alternatively, edge enhancement may be performed using other filters. Furthermore, edge enhancement may be performed by combining multiple filters.

In this embodiment, a particle beam therapy system is exemplified, but it may be a broader radiation therapy system. For example, the configuration and method for patient positioning exemplified in this embodiment are broadly applicable to radiation therapy systems. As an example, the accelerator 1 described in this embodiment may be an electron beam accelerator intended for therapeutic X-rays, or a particle beam accelerator such as a proton or carbon beam accelerator.

Furthermore, the above-mentioned configurations, functions, processing units, processing means, etc. may be realized in part or in whole in hardware, for example by designing them as integrated circuits. Furthermore, the above-mentioned configurations, functions, etc. may be realized in software by a processor interpreting and executing a program that realizes each function. Information on the programs, tables, files, etc. that realize each function can be stored in a memory, a recording device such as a hard disk or SSD, or a recording medium such as an IC card, SD card, or DVD.

In addition, the control lines and information lines shown are those that are considered necessary for the explanation, and do not necessarily show all the control lines and information lines on the product. In reality, almost all components are interconnected.

A...particle beam therapy system, 1...accelerator, 2...beam transport device, 3...gantry, 4...irradiation nozzle, 5A...planar detector, 5B...planar detector, 6A...X-ray tube, 6B...X-ray tube, 7...bed, 8...robot arm, 9...patient, 10...treatment planning device, 11...communication device, 12...data server, 14...fluoroscopic X-ray image capture device, 15...bed control device, 20...patient positioning device, 21...pseudo-fluoroscopic X-ray image creation unit, 22...ROI drawing unit, 23...two-dimensional similarity information calculation unit, 24...ROI weighting processing unit, 25...image Matching unit, 26... image display unit, 50... ROI, 51... 2D similarity information, 52... square area including ROI, 53... spine, 100... pseudo-fluoroscopic X-ray image with ROI superimposed, 101... pseudo-fluoroscopic X-ray image with 2D similarity image superimposed, 102... image of enlarged square area 52 in image 101, 200... pseudo-fluoroscopic X-ray image with ROI drawn, 201... pseudo-fluoroscopic X-ray image with ROI weighted based on 2D similarity information, 202... pseudo-fluoroscopic X-ray image with ROI with updated weighting coefficient N days after treatment

Claims (8)

A positioning device for controlling the position of a bed on which a patient is placed,
creating a pseudo-fluoroscopic X-ray image by projecting a three-dimensional image of the patient onto a predetermined surface, the pseudo -fluoroscopic X-ray image being used for creating a treatment plan for irradiating the patient with radiation in radiation therapy;
obtaining a first region of interest in the pseudo-fluoroscopic x-ray image that includes a region of a target structure of the patient;
calculating two-dimensional similarity information between the fluoroscopic X-ray image and the pseudo fluoroscopic X-ray image based on the fluoroscopic X-ray image of the patient taken by X-ray fluoroscopy in the first region of interest and the pseudo fluoroscopic X-ray image;
applying a weighting factor to the first region of interest based on the two-dimensional similarity information to generate a second region of interest in the pseudo-fluoroscopic X-ray image ;
a positioning device which calculates an amount of movement for moving the bed so that a position of the target structure in the fluoroscopic X-ray image coincides with a position of the target structure in the pseudo fluoroscopic X-ray image only in a portion included in the second region of interest. The positioning device according to claim 1, characterized in that the positioning device uses, as the two-dimensional similarity information, information obtained by converting a similarity index used when calculating the similarity between the fluoroscopic X-ray image and the pseudo fluoroscopic X-ray image into two dimensions. The positioning device according to claim 2, characterized in that the positioning device uses a normalized cross-correlation coefficient as the similarity index. The positioning device according to claim 1, characterized in that the positioning device displays the second region of interest, and a comparison result between the second region of interest, the position of the target structure in the fluoroscopic X-ray image, and the position of the target structure in the pseudo-fluoroscopic X-ray image. The positioning device according to claim 1, characterized in that the positioning device calculates the two-dimensional similarity information using the pseudo fluoroscopic X-ray image and an image obtained by edge-enhancing the fluoroscopic X-ray image. The positioning device according to claim 1, characterized in that the positioning device creates a plurality of the pseudo fluoroscopic X-ray images using the three-dimensional images corresponding to a plurality of respiratory phases of the patient, and calculates the two-dimensional similarity information using the pseudo fluoroscopic X-ray image corresponding to any one of the respiratory phases and the pseudo fluoroscopic X-ray images corresponding to the other respiratory phases. A particle beam therapy apparatus having a positioning device for controlling the position of a bed on which a patient is placed,
The positioning device includes:
creating a pseudo-fluoroscopic X-ray image by projecting a three-dimensional image of the patient onto a predetermined surface, the pseudo -fluoroscopic X-ray image being used for creating a treatment plan for irradiating the patient with radiation in radiation therapy;
obtaining a first region of interest in the pseudo-fluoroscopic x-ray image that includes a region of a target structure of the patient;
calculating two-dimensional similarity information between the fluoroscopic X-ray image and the pseudo fluoroscopic X-ray image based on the fluoroscopic X-ray image of the patient taken by X-ray fluoroscopy in the first region of interest and the pseudo fluoroscopic X-ray image;
applying a weighting factor to the first region of interest based on the two-dimensional similarity information to generate a second region of interest in the pseudo-fluoroscopic X-ray image ;
a particle beam therapy device which calculates an amount of movement for moving the bed so that the position of the target structure in the fluoroscopic X-ray image coincides with the position of the target structure in the pseudo fluoroscopic X-ray image only in the portion included in the second region of interest. A positioning method for use in a positioning device that controls the position of a bed on which a patient is placed, comprising:
creating a pseudo-fluoroscopic X-ray image by projecting a three-dimensional image of the patient onto a predetermined surface, the pseudo -fluoroscopic X-ray image being used for creating a treatment plan for irradiating the patient with radiation in radiation therapy;
obtaining a first region of interest in the pseudo-fluoroscopic x-ray image that includes a region of a target structure of the patient;
calculating two-dimensional similarity information between the fluoroscopic X-ray image and the pseudo fluoroscopic X-ray image based on the fluoroscopic X-ray image of the patient taken by X-ray fluoroscopy in the first region of interest and the pseudo fluoroscopic X-ray image;
applying a weighting factor to the first region of interest based on the two-dimensional similarity information to generate a second region of interest in the pseudo-fluoroscopic X-ray image ;
a positioning method comprising: calculating an amount of movement for moving the bed so that a position of the target structure in the fluoroscopic X-ray image coincides with a position of the target structure in the pseudo fluoroscopic X-ray image only in a portion included in the second region of interest.