Information processing device, information processing method, and information processing program

Abstract

This information processing device derives information representing a reference plane by using data corresponding to a portion of three-dimensional point group data that satisfies a determined condition, the three-dimensional point group data having been obtained by imaging, along the irradiation direction of radiation emitted from a radiation source, the reference plane having a portion overlapping a radiation detector and a subject, and derives, using the derived information representing the reference plane, at least one of a distance between the radiation source and the radiation detector and the angle of incidence of the radiation emitted from the radiation source with respect to the reference plane.

Description

Information processing device, information processing method, and information processing program

[0001] This disclosure relates to an information processing device, an information processing method, and an information processing program.

[0002] Japanese Patent Publication No. 2011-092612 discloses a technique for calculating the distance between a radiation source and a radiation detector based on the distance between markers captured in an image taken by a camera.

[0003] Incidentally, a technique is known in which the distance between a radiation source and a radiation detector is measured using a distance measuring sensor placed near the radiation source. However, with this technique, it can be difficult to accurately derive at least one of the distance between the radiation source and the radiation detector and the angle of incidence of the radiation, for example, when an object is present between the radiation source and the radiation detector.

[0004] This disclosure is made in view of the above circumstances and aims to provide an information processing device, an information processing method, and an information processing program that can accurately derive at least one of the distance between the radiation source and the radiation detector and the angle of incidence of the radiation.

[0005] The first embodiment of the information processing device includes a processor, which uses a portion of three-dimensional point cloud data obtained by photographing a reference plane having a portion that overlaps with the subject and the radiation detector along the direction of irradiation of radiation emitted from a radiation source, and which satisfies a predetermined condition, to derive information representing the reference plane, and uses the derived information representing the reference plane to derive at least one of the distance between the radiation source and the radiation detector and the angle of incidence of the radiation emitted from the radiation source with respect to the reference plane.

[0006] The second embodiment of the information processing device is the same as the first embodiment of the information processing device, wherein the condition is that the illumination direction is within a set range based on the distance between the sensor used to acquire the three-dimensional point cloud data and the surface of the subject.

[0007] The third embodiment of the information processing device is an information processing device of the first or second embodiment, wherein the condition is that the position of the subject's body surface is within a set range in the direction of the first edge and the direction of the second edge intersecting the first edge of the field of view of the camera used to acquire the three-dimensional point cloud data.

[0008] The fourth embodiment of the information processing device is an information processing device according to any one of the first to third embodiments, wherein the condition is that the data is such that the angle made with the direction of radiation irradiation is less than a set angle.

[0009] The fifth embodiment of the information processing apparatus is an information processing apparatus of any one of the first to fourth embodiments, wherein the processor detects an area other than the reference plane, and the condition is that the detected area is located outside the detected area.

[0010] The sixth embodiment of the information processing device is an information processing device in the second embodiment in which the set range is a range set by the distance between the sensor and the body surface of the subject and two different offset values.

[0011] In the seventh embodiment of the information processing apparatus, the processor corrects two offset values ​​using sensor calibration data, as in the sixth embodiment of the information processing apparatus.

[0012] The eighth aspect of the information processing apparatus is an information processing apparatus in the fourth aspect, in which the processor derives a unit normal vector for each data point in the 3D point cloud data, and data points with an angle less than a set angle are data points where the absolute value of the components of the unit normal vector is greater than or equal to a threshold.

[0013] The ninth aspect of the information processing apparatus is the fourth aspect of the information processing apparatus, wherein the condition is that the point cloud data constitutes a shape in which the curvature along the irradiation direction is less than a threshold.

[0014] The tenth embodiment of the information processing apparatus is an information processing apparatus of the third embodiment in which the set range in the direction of the first side and the direction of the second side is a circular range.

[0015] The information processing apparatus of the eleventh embodiment is an information processing apparatus of the third embodiment in which the set range in the direction of the first side and the direction of the second side is a rectangular range.

[0016] The information processing device of the twelfth embodiment is an information processing device of the eleventh embodiment in which the rectangle is a rectangle whose shorter side is the axis direction of the subject's body.

[0017] In the 13th embodiment of the information processing apparatus, the processor detects the head-to-foot direction as the body axis direction based on the keypoint detection result for the optical image obtained by photographing the subject.

[0018] The 14th embodiment of the information processing apparatus is an information processing apparatus of the 5th embodiment in which the 3D point cloud data is data obtained by converting a distance image taken of a reference plane, and the processor detects the region of the subject as a region other than the reference plane by performing clustering on the distance image.

[0019] The information processing apparatus of the 15th embodiment is an information processing apparatus of the 14th embodiment in which the processor further detects the torso region of the subject by using the central position of the radiation irradiated onto the subject.

[0020] The sixteenth embodiment of the information processing apparatus is an information processing apparatus of the fourteenth embodiment in which the processor detects the body axis direction of the subject and further detects the trunk region and the head region of the subject by using the detected body axis direction.

[0021] The 17th embodiment of the information processing device is an information processing device of the 16th embodiment in which the processor detects a region located outside the width direction of the subject with respect to the head region as a pillow region when the boundary position of the region in the body axis direction is located closer to the head than the trunk region.

[0022] In the 18th embodiment of the information processing apparatus, in the fifth embodiment of the information processing apparatus, the processor detects a region of the subject as a region other than the reference plane based on the keypoint detection result of the optical image obtained by photographing the subject.

[0023] An information processing device of the 19th embodiment, in an information processing device of the 18th embodiment, detects the area of ​​a subject by thickening the line segment connecting the joints of keypoints by a set width.

[0024] The 20th embodiment of the information processing apparatus is an information processing apparatus of any one of the first to 19 embodiments in which the processor derives the distance by deriving the intersection point of a straight line along the direction of irradiation and a reference plane in three-dimensional space.

[0025] The 21st embodiment of the information processing device is an information processing device of any one of the first to 20 embodiments, in which the processor derives the angle of incidence using the normal vector of the reference plane and inverse trigonometric functions.

[0026] The 22nd aspect of the information processing apparatus is an information processing apparatus of any one of the first to 21 aspects, in which the processor derives information representing a reference plane by iterative processing using some data and a regression algorithm.

[0027] The 23rd aspect of the information processing method involves an information processing device equipped with a processor, in which the processor derives information representing a reference plane using a portion of three-dimensional point cloud data that satisfies predetermined conditions, obtained by photographing a reference plane having a portion that overlaps with the subject and the radiation detector along the irradiation direction of the radiation emitted from the radiation source, and using the derived information representing the reference plane, performs a process to derive at least one of the distance between the radiation source and the radiation detector and the angle of incidence of the radiation emitted from the radiation source with respect to the reference plane.

[0028] The 24th embodiment of the information processing program causes the processor of an information processing device equipped with a processor to derive information representing a reference plane using a portion of three-dimensional point cloud data that satisfies predetermined conditions, obtained by photographing a reference plane having a portion that overlaps with the subject and the radiation detector along the irradiation direction of the radiation irradiated from the radiation source, and to perform a process to derive at least one of the distance between the radiation source and the radiation detector and the angle of incidence of the radiation irradiated from the radiation source with respect to the reference plane using the derived information representing the reference plane.

[0029] According to this disclosure, at least one of the distance between the radiation source and the radiation detector and the angle of incidence of the radiation can be accurately derived.

[0030] This is a front view showing an example of a radiography system. This is a side view showing an example of a radiography system. This is a diagram illustrating the positional relationship between the radiation source and the ToF camera. This is a front view showing an example of a radiography system. This is a block diagram illustrating an example of the hardware configuration of a console. This is a block diagram illustrating an example of the functional configuration of a console. This is a diagram illustrating the SSD derivation process. This is a diagram illustrating the SSD derivation process related to a modified example. This is a diagram illustrating the first condition. This is a diagram illustrating the first condition. This is a diagram illustrating the second condition. This is a diagram showing an example of the detection result of the area of ​​the subject. This is a diagram illustrating the reference plane derivation process. This is a diagram illustrating the reference plane derivation process. This is a diagram illustrating the effect of re-deriving a plane. This is a diagram illustrating the incident angle derivation process. This is a diagram illustrating the incident angle derivation process. This is a flowchart illustrating an example of the SID and incident angle derivation process.

[0031] Hereinafter, with reference to the drawings, examples of embodiments for carrying out the technology of this disclosure will be described in detail.

[0032] First, the configuration of the radiography system 1 will be described with reference to Figures 1 and 2. As shown in Figures 1 and 2, the radiography system 1 is a system that takes a radiographic image of a subject H by irradiating the subject H with radiation R such as X-rays, and is operated by a user such as a radiologic technologist. The radiography system 1 includes a radiation source 10, an imaging device 13, a radiation detector 14, a patient table 15, and a console 16. Figure 1 is a front view when looking from the head side to the foot side of the subject H, and Figure 2 is a side view when looking from the right side to the left side of Figure 1. The console 16 is an example of an information processing device relating to the disclosed technology.

[0033] The radiation source 10 includes a radiation tube 11 that emits radiation R and a field limiter (also called a collimator) 12 that limits the irradiation field of radiation R. The radiation tube 11 is provided with, for example, a filament, a target, and a grid electrode. A voltage is applied from a voltage generator between the filament, which is the cathode, and the target, which is the anode. This voltage applied between the filament and the target is called the tube voltage. The filament emits thermionic electrons toward the target in accordance with the applied tube voltage. The target emits radiation R through collisions with thermionic electrons emitted from the filament. The grid electrode is placed between the filament and the target. The grid electrode changes the flow rate of thermionic electrons from the filament toward the target in accordance with the voltage applied from the voltage generator. This flow rate of thermionic electrons from the filament toward the target is called the tube current.

[0034] The irradiation field limiter 12 has an entrance aperture through which radiation R from the radiation tube 11 enters, and an exit aperture through which radiation R exits. Four shielding plates are provided near the exit aperture. The shielding plates are made of a material that shields radiation R, such as lead. The shielding plates are arranged on each side of a rectangle, in other words, in a checkered pattern, forming a rectangular irradiation aperture that allows radiation R to pass through. The irradiation field limiter 12 changes the size of the irradiation aperture by changing the position of each shielding plate, thereby changing the irradiation field of radiation R.

[0035] The imaging device 13 is attached to the outside of the irradiation field limiter 12. The imaging device 13 includes an optical camera 17 and a ToF (Time of Flight) camera 18 (see FIG. 3). The optical camera 17 is a camera capable of capturing a color image of RGB (Red, Green, Blue) by detecting the reflected light of the subject H. The optical camera 17 includes an optical system and an imaging element such as a CCD (Charge Coupled Device) image sensor. The optical camera 17 outputs an image (hereinafter referred to as an "optical image") obtained by capturing the subject H to the console 16. In the example of FIG. 1, the imaging device 13 is shown attached to the outside of the irradiation field limiter 12, but the imaging device 13 may be attached to the ceiling or the like. In this case, it is only necessary to acquire the positional relationship between the imaging device 13 and the radiation source 10.

[0036] The ToF camera 18 is a camera capable of capturing a distance image representing the distance to the imaging target by a ToF type sensor. The ToF camera 18 outputs the distance image obtained by capturing the reference plane P to the console 16. Details of the reference plane P will be described later. In the present embodiment, the distance in the distance image represents the distance between the focal point of the ToF camera 18 and the imaging target.

[0037] The radiation detector 14 is portable and acquires a radiation image of the subject H by detecting the radiation R that has passed through the subject H. The radiation detector 14 outputs the radiation image of the subject H to the console 16. The radiation detector 14 is used in a state where it is placed at a position corresponding to the imaging site between the surface on which the subject H of the hospital bed 15 lies and the subject H. In the example of FIG. 1, the hospital bed 15 is shown in the lying position of a hospital bed in a hospital room, but it may be in a semi-recumbent position, or the hospital bed 15 may be a lying position imaging table in which the radiation detector 14 provided in the imaging room is accommodated.

[0038] The console 16 has functions such as allowing the user to confirm and input irradiation conditions of the radiation R, and performing image processing on the radiation image obtained by the radiation detector 14. Examples of the console 16 include a computer such as a personal computer or a server computer.

[0039] The reference plane P according to this embodiment is a plane that serves as a reference when measuring the body thickness of the subject H, and is a plane having a portion overlapping the subject H and the radiation detector 14 along the irradiation direction D of the radiation R irradiated from the radiation source 10. That is, the reference plane P is a plane located on the path along which the radiation R is irradiated. Further, the reference plane P is a plane of the bed 15 with which at least a part of the subject H is in contact, and is a plane parallel to the detection surface of the radiation R of the radiation detector 14. The parallelism here means parallelism within a range including an allowable error. In the example of FIG. 1, the plane of the bed 15 with which the subject H lying on it is in contact is the reference plane P. Note that the reference plane P may be a part of the plane of the bed 15. When the bed 15 is a bed, the reference plane P is the plane of the bed with which the subject H is in contact, and when the bed 15 is a lying-position imaging table in which the radiation detector 14 is housed, the reference plane P is the plane of the top plate of the lying-position imaging table with which the subject H is in contact. The above-described distance image also includes distances to objects other than the reference plane P, such as a pillow, a bed, and the subject H.

[0040] In FIG. 1, the case where the subject H is in a lying position is illustrated, but the subject H may be in a standing position. In this case, a standing-position imaging table is used instead of the bed 15. The reference plane P in this case is the plane of the standing-position imaging table in which the radiation detector 14 provided in the imaging room is housed and with which the subject H is in contact. Also, the radiation detector 14 in this case does not have to be portable.

[0041] In Figure 1, arrow Y1 represents the distance between the focal point of the ToF camera 18 and the surface of the bed 15. Arrow Y2 represents the distance between the radiation source 10 and the surface of the bed 15. Hereafter, the distance between the radiation source 10 and the surface of the bed 15 will also be referred to as "SID (Source to Image receptor Distance)". In this embodiment, since the height difference between the surface of the bed 15 and the detection surface of the radiation detector 14 is relatively small, the distance between the radiation source 10 and the surface of the bed 15 will be treated as the distance between the radiation source 10 and the radiation detector 14. In this embodiment, the case where the distance between the focal point of the radiation source 10 and a preset position such as the center of the detection surface of the radiation detector 14 is applied as the SID will be explained as an example. In Figure 1, arrow Y3 represents the distance between the radiation source 10 and the body surface of the subject H. Hereafter, the distance between the radiation source 10 and the body surface of the subject H will also be referred to as "SSD (Source to Surface Distance)".

[0042] Next, with reference to Figure 3, the positional relationship between the focal point of the radiation source 10 and the focal point of the ToF camera 18 will be explained. In this embodiment, a three-dimensional Cartesian coordinate system using the X, Y, and Z axes is used. Below, the direction of the first side of the field of view of the ToF camera 18 (the direction of the long side in this embodiment) will be described as the X-axis direction. The direction of the second side that intersects (orthogonal in this embodiment) the first side of the field of view of the ToF camera 18 (the direction of the short side in this embodiment) will be described as the Y-axis direction. The irradiation direction D of the radiation R will be described as the Z-axis direction. The irradiation direction D of the radiation R means the axis direction of the radiation R that is irradiated in a cone-beam shape. The axis of the radiation R means the axis connecting the focal point of the radiation source 10 and a predetermined position such as the center of the radiation detector 14. The X-axis direction and the Y-axis direction correspond to the width direction and body axis direction of the subject H when the ToF camera 18 and the subject H are facing each other.

[0043] As shown in Figure 3, the distance between the focus of the radiation source 10 and the focus of the ToF camera 18 along the X-axis is represented as Δx. The distance between the focus of the radiation source 10 and the focus of the ToF camera 18 along the Y-axis is represented as Δy. The distance between the focus of the radiation source 10 and the focus of the ToF camera 18 along the Z-axis is represented as Δz. In this embodiment, the focus of the ToF camera 18 is set as the origin in the three-dimensional Cartesian coordinate system. That is, in the example in Figure 3, the coordinates of the focus of the radiation source 10 are represented as (-Δx, Δy, -Δz).

[0044] In such a three-dimensional space, the reference plane P can be expressed by equation (1), which is the equation of a plane. Note that while Figure 1 shows an example where the direction of radiation R D is perpendicular to the reference plane P, as shown in Figure 4, the reference plane P can also be expressed by equation (1) when radiation R is incident obliquely on the reference plane P. ax + by + cz + d = 0 ... (1)

[0045] Referring to Figure 5, the hardware configuration of the console 16 according to this embodiment will be described. As shown in Figure 5, the console 16 includes a CPU (Central Processing Unit) 31, a memory 32 as a temporary storage area, and a non-volatile storage unit 33. The console 16 also includes a display 34 such as a liquid crystal display, input devices 35 such as a keyboard and mouse, and a network interface 36. The CPU 31, memory 32, storage unit 33, display 34, input devices 35, and network interface 36 are connected to a bus 37. The CPU 31 is an example of a processor. The console 16 transmits and receives data to and from the radiation source 10, radiation detector 14, optical camera 17, and ToF camera 18 via the network interface 36.

[0046] The storage unit 33 is implemented by an HDD (Hard Disk Drive), SSD (Solid State Drive), or flash memory, etc. The storage unit 33, as a storage medium, stores the information processing program 40. The CPU 31 reads the information processing program 40 from the storage unit 33, expands it into the memory 32, and executes the expanded information processing program 40.

[0047] Next, referring to FIG. 6, the functional configuration of the console 16 will be described. As shown in FIG. 6, the console 16 includes an acquisition unit 50, a conversion unit 52, a first derivation unit 54, a second derivation unit 56, a third derivation unit 58, and a display control unit 60. When the CPU 31 executes the information processing program 40, it functions as the acquisition unit 50, the conversion unit 52, the first derivation unit 54, the second derivation unit 56, the third derivation unit 58, and the display control unit 60.

[0048] The acquisition unit 50 acquires the distance image obtained by photographing the reference plane P by the ToF camera 18 via the network I / F 36. The conversion unit 52 converts the distance image acquired by the acquisition unit 50 into three-dimensional point cloud data by a known conversion algorithm. At this time, the conversion unit 52 sets the focal point of the ToF camera 18 at the origin and sets the irradiation direction D of the radiation R in the Z-axis direction. The three-dimensional point cloud data obtained by the conversion unit 52 is an example of the three-dimensional point cloud data obtained by photographing the reference plane according to the disclosed technology.

[0049] The first derivation unit 54 derives the SSD by searching the three-dimensional point cloud data obtained by the conversion unit 52 along the irradiation direction D. Specifically, as shown in FIG. 7, the coordinates (X Ｓｕｒｆａｃｅ , <`0000007`, <`0000008`, <`0000005`, <`0000006`, <`0000004`, <`0000010`, Y ｘｒａｙ , Z ｘｒａｙ ) of the focal point of the radiation source 10 are searched for the presence or absence of points included in the three-dimensional point cloud data along the Z-axis direction, and the coordinates (X <`000000`4, Y <`000000`5, Z <`000000`6) of the point closest to the focal point of the radiation source 10 among the detected points are derived. Point B1 in FIG. 7 represents the focal point of the radiation source 10, and point B2 represents the point closest to the focal point of the radiation source 10 among the points detected by the first derivation unit 54. Then, the first derivation unit 54 derives the SSD by subtracting Z <`000000`7 from Z <`000000`8 according to the following formula (2). SSD = Z <`000000`9 - Z <`000001``0`... (2)

[0050] For example, as shown in Figure 8, the first derivation unit 54 detects point B2 by searching within a set range of the X-Y plane along the irradiation direction D from point B1, which is the focal point of the radiation source 10. In Figure 8, the set range of the X-Y plane is defined as having a side length in the X-axis direction of X. ｒａｎｇｅ And the length of the side in the Y-axis direction is Y ｒａｎｇｅ This shows an example of a rectangular area. In this case, the set area may also be a circular area. Furthermore, the first derivation unit 54 sequentially searches for multiple points, such as the second and third closest points, not just the point closest to the focal point of the radiation source 10, and calculates the average of the Z coordinates of the multiple points that were searched. Ｓｕｒｆａｃｅ This is also acceptable. This helps to suppress variations in search results due to noise contained in the distance image.

[0051] The second derivation unit 56 derives information representing the reference plane P using a portion of the three-dimensional point cloud data obtained by the conversion unit 52 that satisfies predetermined conditions (hereinafter referred to as "partial point cloud data"). In this embodiment, the first derivation unit 54 derives equation (1) above as information representing the reference plane P. Examples of the predetermined conditions include the following:

[0052] As shown in Figure 9, the first condition is that the location is within a set range R1 in the direction of the first edge and the second edge intersecting (orthogonal in this embodiment) the first edge of the field of view of the ToF camera 18 used to acquire the 3D point cloud data, with the position of the body surface of the subject H as the reference. In this embodiment, range R1 is a rectangular area centered on the center of the radiation field of R. In this embodiment, the direction of the first edge is the direction of the long side of the field of view of the ToF camera 18, which is the X-axis direction. In this embodiment, the direction of the second edge is the direction of the short side of the field of view of the ToF camera 18, which is the Y-axis direction. Range R1 may be, for example, a range that includes the surface of the bed 15, which is the area outside the subject H, or it may be a range inside the surface of the bed 15. As a result, the second derivation unit 56 can derive information representing the reference plane P using point cloud data that includes points representing the surface of the bed 15 but does not include points outside the bed 15, such as the floor.

[0053] In Figure 9, the case where the range R1 is a square is illustrated, but as shown in Figure 10, the range R1 may be a rectangle with the shorter side being the body axis direction of the subject H. The second derivation unit 56 may exclude data corresponding to the pillow from the 3D point cloud data by making the range R1 a rectangle with the shorter side being the body axis direction. In this embodiment, the body axis direction of the subject H corresponds to the direction of the second side. The second derivation unit 56 may detect keypoints by inputting the optical image obtained by photographing the subject H with the optical camera 17 into a posture detection model obtained by machine learning such as deep learning. Alternatively, the second derivation unit 56 may detect the head-to-foot direction of the subject H as the body axis direction based on the keypoint detection results for the optical image. For example, the posture detection model can obtain keypoints of the head such as both eyes, both ears, and nose, and keypoints of the trunk such as both shoulders and both hips. The second derivation unit 56 can detect the direction from the head to the feet of the subject H as the body axis direction from the positional relationship of these keypoints. Furthermore, the second output unit 56 may rotate the range R1 to match the rotation angle of the illumination field limiter 12, that is, the rotation angle of the field of view of the ToF camera 18. Also, the range R1 may be a circular range.

[0054] As shown in Figure 11, the second condition is that the distance between the ToF camera 18, which is an example of a sensor used to acquire 3D point cloud data, and the body surface of the subject H is within a set range R2 in the illumination direction D relative to the SSD. The range R2 may be set by the SSD and two different offset values.

[0055] For example, the second derivation unit 56 may use the range of the Z coordinates shown by the following equation (3) as the range R2. ＳｕｒｆａｃｅThis is the Z-coordinate of the body surface of subject H used in the derivation of the SSD. offset1 is the first offset value. The first offset value is set, for example, according to the lower limit of the body thickness of the part being photographed. The second offset value is a value greater than the first offset value and is set according to the upper limit of the body thickness of the part being photographed. This narrows down the 3D point cloud data used in the derivation of the reference plane P to data near the bed 15. The second derivation unit 56 may set the first and second offset values ​​according to information about the subject H, such as race, age, sex, and physique. Ｓｕｒｆａｃｅ +offset1<z<Z Ｓｕｒｆａｃｅ +offset2...(3)

[0056] Furthermore, the distance measured by the ToF camera 18 may contain errors compared to the actual distance. For example, when installing the radiation source 10, the installer may pre-measure the relationship between the actual distance and the distance measured by the ToF camera 18 while varying the distance between the ToF camera 18 and the object being photographed. In this case, the console 16 may store data representing the relationship between the actual distance and the distance measured by the ToF camera 18 as calibration data for the ToF camera 18 in the storage unit 33. In this case, the second derivation unit 56 may correct the two offset values ​​mentioned above using the calibration data for the ToF camera 18.

[0057] The third condition is that the angle between the radiation R and the irradiation direction D is less than a set angle. The reference plane P is often roughly perpendicular to the irradiation direction D. Therefore, points that are estimated not to be on the reference plane P can be excluded from the 3D point cloud data according to the third condition. For example, the second derivation unit 56 derives a unit normal vector for each data point in the 3D point cloud data, and uses data from the 3D point cloud data where the absolute value of the unit normal vector component is greater than or equal to a threshold to derive information representing the reference plane P as data that satisfies the third condition.

[0058] Specifically, the second derivation unit 56 randomly acquires a set number of neighboring point cloud data points located within a set distance from the 3D point cloud data. Next, the second derivation unit 56 derives a covariance matrix from the acquired point cloud data. The second derivation unit 56 also derives eigenvalues ​​from the covariance matrix. In this embodiment, since the target is a 3D space, the second derivation unit 56 derives three eigenvalues. The second derivation unit 56 also derives the unit eigenvector with the smallest calculated eigenvalue. Since the surface formed by the set number of point cloud data points is considered to be closest to the plane spanned by the two axes with the maximum variance, the eigenvector corresponding to the smallest eigenvalue corresponds to the unit normal vector of the 3D point cloud data. Then, the second derivation unit 56 identifies the data from the 3D point cloud data that satisfies the third condition if the absolute value of the Z component of the unit eigenvector, i.e., the unit normal vector, is greater than or equal to a threshold. This makes it possible to exclude data corresponding to the edge of the pillow, the arm region of subject H, and the handrail of the bed 15 from the 3D point cloud data.

[0059] Furthermore, the second derivation unit 56 may derive information representing the reference plane P using data that satisfies the condition that the data satisfies the third condition, namely, point cloud data that constitutes a shape with curvature along the irradiation direction D less than a threshold. This makes it possible to exclude data corresponding to the edge of the pillow, as well as the edges of the arms and torso of the subject H, etc., from the three-dimensional point cloud data. In this case, for example, the second derivation unit 56 derives the curvature of the shape formed by nearby point cloud data located within a set distance.

[0060] The second derivation unit 56 detects an area other than the reference plane P, and the fourth condition is that the area is located outside the area detected by the second derivation unit 56. The second derivation unit 56 detects the area of ​​the subject H as an area other than the reference plane P by performing clustering on the distance image taken of the reference plane P. For example, if the part to be photographed is the torso, the area including the position on the distance image corresponding to the center position of the radiation R is estimated to be the torso, and the area extending from the torso area along the body axis is estimated to be the head area. In this case, the area to the side of the head area is estimated to be the pillow area.

[0061] Therefore, during clustering, the second derivation unit 56 detects the torso region of the subject H by using the center position of the radiation R irradiated onto the subject H. Furthermore, during clustering, the second derivation unit 56 may also detect the torso region and the head region of the subject H by using the detected body axis direction. In this case, the second derivation unit 56 may detect a region located outside the head region in the width direction of the subject H, i.e., in the X-axis direction, as the pillow region if the body axis boundary position of that region is located closer to the head than the torso region. As shown in Figure 12, the clustering by the second derivation unit 56 as described above can detect the head region, the torso region of the subject H, and the pillow region. As a result, the second derivation unit 56 can exclude points located in these regions that do not correspond to the reference plane P from the 3D point cloud data.

[0062] Furthermore, the second derivation unit 56 may detect the region of the subject H as a region other than the reference plane P based on the detection results of the key points described above. In this case, the second derivation unit 56 may detect the region of the subject H by thickening the line segments connecting the joints of the key points by a set width. For example, when the second derivation unit 56 estimates the region of the arm of the subject H, it generates line segments connecting the shoulder, elbow, and wrist in that order from the detection results of the positions of the left and right shoulders, elbows, and wrists as key points, and estimates the region of the arm by thickening these line segments by a set width. Then, the second derivation unit 56 extracts the point cloud data of the distance image corresponding to the region of the arm in the optical image from the 3D point cloud data and derives information representing the reference plane P.

[0063] The partial point cloud data may satisfy one of the four conditions described above for 3D point cloud data, or it may satisfy two or more conditions.

[0064] In this embodiment, the second derivation unit 56 derives information representing the reference plane P through iterative processing using partial point cloud data and a regression algorithm. A specific example of the derivation process performed by the second derivation unit 56 will be described below.

[0065] The second derivation unit 56 uses RANSAC (RANDOM SAmple Consensus), a robust algorithm, as the regression algorithm. As an example, as shown in Figure 13, the second derivation unit 56 first randomly selects three points from the partial point cloud data and derives information representing a plane passing through the three selected points. In the example in Figure 13, point C represents the three points. The second derivation unit 56 derives equation (1), which represents the plane passing through the three points, as information representing the plane.

[0066] Next, as shown in Figure 14, the second derivation unit 56 derives the number of points in the partial point cloud data whose distance from the derived plane is within a threshold δ. For each of several different sets of three points, the second derivation unit 56 derives information representing the plane and the number of points whose distance from that plane is within a threshold δ. Then, the second derivation unit 56 selects the plane with the largest number of derived points.

[0067] Furthermore, the second derivation unit 56 uses points whose distance from the selected plane is within a threshold δ to re-derive the plane according to an optimization algorithm such as the least squares method, and uses the information representing the derived plane as the information representing the reference plane P. As an example, as shown in Figure 15, by re-deriving the plane, the second derivation unit 56 can make the reference plane P a plane that is closer to the target plane than the plane selected based on the number of points whose distance is within a threshold δ.

[0068] The third derivation unit 58 uses the information representing the reference plane P derived by the second derivation unit 56 to derive the SID and the incident angle of the radiation R irradiated from the radiation source 10 with respect to the reference plane P. In this embodiment, the case in which the third derivation unit 58 derives both the SID and the incident angle is described as an example, but the third derivation unit 58 may derive only one of the SID or the incident angle.

[0069] Specifically, the third derivation unit 58 derives the SID by deriving the intersection point A (see Figure 1) of a straight line along the irradiation direction D in three-dimensional space and the reference plane P. As mentioned above, the coordinates of the focal point of the radiation source 10 are known coordinates (-Δx, Δy, -Δz), and equation (1) is derived by the second derivation unit 56 as information representing the reference plane P. Therefore, the third derivation unit 58 can derive the SID from the three-dimensional positional relationship between the focal point of the radiation source 10 and the intersection point A.

[0070] Furthermore, the third derivation unit 58 derives the incident angle using the normal vector of the reference plane P and inverse trigonometric functions. Referring to Figures 16 and 17, a specific example of the incident angle derivation process performed by the third derivation unit 58 will be explained. As shown in Figures 16 and 17, the normal vector n, which is the unit normal vector of the reference plane P represented by equation (1), is (a, b, c), and the unit vector x of the radiation irradiation direction D is (0, 0, 1). The relationship between the normal vector n and the unit vector x, and the angle θ between the normal vector n and the unit vector x, is expressed by the following equation (4). θ corresponds to the incident angle in three-dimensional space. x・n=c=cosθ・・・(4)

[0071] As shown in Figure 17, if ψ is the angle between the vector obtained by projecting the normal vector n onto the Y-Z plane and the direction of illumination D, then tanψ is expressed by the following equation (5). Therefore, ψ is expressed by the following equation (6). ψ corresponds to the angle of incidence in the Y-Z plane. tanψ = b / c ... (5) ψ = atan(b / c) ... (6)

[0072] Similarly, if φ is the angle between the vector obtained by projecting the normal vector n onto the Z-X plane and the direction of illumination D, then tanφ is expressed by equation (7) below. Therefore, φ is expressed by equation (8) below. φ corresponds to the angle of incidence in the Z-X plane. tanφ = a / c ... (7) φ = atan(a / c) ... (8)

[0073] The third derivation unit 58 in this embodiment derives ψ and φ as the angle of incidence of radiation R with respect to the reference plane P, using the normal vector n(a, b, c) and atan as an example of an inverse trigonometric function, according to equations (6) and (8).

[0074] Furthermore, the third derivation unit 58 may derive θ as the angle of incidence of radiation R with respect to the reference plane P, using the normal vector n(a, b, c) and acos as an example of an inverse trigonometric function, according to equation (9) below: θ = acos(c) ... (9)

[0075] The display control unit 60 controls the display of the SID and incident angle derived by the third derivation unit 58. For example, the display control unit 60 controls the display of the SID and incident angle derived by the third derivation unit 58 on a display provided on the outer surface of the irradiation field limiter 12. As a result, the user can grasp the SID and incident angle, thereby supporting the positioning work of the radiation source 10 and subject H performed by the user. The display control unit 60 may also control the display of the SID and incident angle on the display 34, on a display installed on a display cart with casters, or on multiple displays.

[0076] Next, the operation of the console 16 will be explained with reference to Figure 18. The CPU 31 executes the information processing program 40, which performs the SID and incidence angle derivation process shown in Figure 18. The SID and incidence angle derivation process is performed, for example, when the user inputs an instruction to start execution, or when the position of the radiation source 10 is changed.

[0077] In step S10 of Figure 18, the acquisition unit 50 acquires a distance image obtained by photographing the reference plane P with the ToF camera 18 via the network I / F 36. In step S12, the conversion unit 52 converts the distance image acquired in step S10 into 3D point cloud data. In step S14, the first derivation unit 54 derives the SSD by searching the 3D point cloud data obtained by the processing in step S12 along the irradiation direction D.

[0078] In step S16, the second derivation unit 56 derives information representing the reference plane P using a portion of the three-dimensional point cloud data obtained in step S12 that satisfies predetermined conditions. In step S18, the third derivation unit 58 derives the SID and the incident angle of the radiation R irradiated from the radiation source 10 with respect to the reference plane P using the information representing the reference plane P derived in step S16. In step S20, the display control unit 60 controls the display of the SID and incident angle derived in step S18. When the processing in step S20 is completed, the derivation process of the SID and incident angle is completed.

[0079] As explained above, according to this embodiment, point cloud data that is not considered to correspond to the reference plane P is excluded from the 3D point cloud data before the reference plane P is derived. Therefore, the reference plane P can be derived with high accuracy, and as a result, the SID and the incident angle of radiation can be derived with high accuracy. Furthermore, according to this embodiment, the computational cost when deriving the reference plane P can be reduced compared to when all 3D point cloud data is used.

[0080] In the above embodiment, the case in which a distance image is captured by a ToF camera 18 has been described, but the disclosed technology is not limited to this embodiment. For example, the distance image may be captured by a stereo camera.

[0081] Furthermore, in the above embodiment, the second derivation unit 56 may derive a plane passing through the three selected points as a candidate reference plane P, and if the angle of incidence of radiation R with respect to the candidate reference plane P exceeds a threshold, the candidate reference plane P may be excluded from processing. Specifically, in this case, the second derivation unit 56 may omit the process of deriving the number of points whose distance is within the threshold δ for the candidate reference plane P. This reduces the computation cost. The second derivation unit 56 can derive the angle of incidence in this case in the same way as the third derivation unit 58.

[0082] Furthermore, in the above embodiment, the second derivation unit 56 may derive the distance from the focal point of the ToF camera 18 to the reference plane P, and then derive the thickness of the subject H by subtracting SSD from the derived distance.

[0083] Furthermore, in the above embodiment, if the reference plane P and the detection surface of the radiation detector 14 are separated along the Z-axis direction and their positional relationship is known, the third derivation unit 58 may derive SID by adding the distance between the reference plane P and the detection surface of the radiation detector 14 along the Z-axis direction to the distance between the focal point of the radiation source 10 and the intersection point A.

[0084] Furthermore, at least one of the functional units of the console 16 in the above embodiment may be provided by other devices such as the control unit of the radiation source 10, the control unit of the imaging device 13, and the control unit of the radiation detector 14.

[0085] Furthermore, in the above embodiments, each process is executed on any computer. The computer may execute these processes using a processor as hardware, a program as software, or a combination thereof. In this case, the processor is configured to work in cooperation with the program to execute the various processes in the above embodiments, and can function as a unit or means in the above embodiments. The execution order of the processes by the processor is not limited to the order described and may be changed as appropriate. The computer may be a general-purpose computer, a computer designed for a specific purpose, a workstation, or any other system capable of executing each process.

[0086] A processor may consist of one or more hardware components, and the type of hardware is not limited. For example, a processor may consist of programmable logic devices such as a CPU (Central Processing Unit), MPU (Micro Processing Unit), FPGA (Field Programmable Gate Array), dedicated circuits for performing specific processing such as an ASIC (Application Specific Integrated Circuit), a GPU (Graphic Processing Unit), or an NPU (Neural Processing Unit). Furthermore, the type of hardware may be a combination of different types of hardware. When multiple hardware components are configured to execute one or more processes of a processor, these components may reside in physically separate devices or in the same device. Furthermore, in any embodiment, the order of the processes performed by the processor is not limited to the order described above and may be changed as appropriate. The hardware components are composed of electrical circuits (circuits) and the like, which are combinations of circuit elements such as semiconductor elements.

[0087] Furthermore, the program may be firmware or software such as microcode. Alternatively, the program may be, for example, a set of program modules, each function of which may be implemented by a processor configured to perform its respective function. The program may be program code or multiple code segments stored on one or more non-temporary computer-readable media (e.g., storage media or other storage). The program may be divided and stored on multiple non-temporary computer-readable media located on physically separate devices. Program code or code segments may represent any combination of procedures, functions, subprograms, routines, subroutines, modules, software packages, classes, or instructions, data structures, or program statements. Program code or code segments may be connected to other code segments or hardware circuits by sending and receiving information, data, arguments, parameters, or memory contents.

[0088] Furthermore, although the above embodiment describes an embodiment in which the information processing program 40 is pre-stored (installed) in the storage unit 33, the invention is not limited to this. The information processing program 40 may be provided in the form of being recorded on a recording medium such as a CD-ROM (Compact Disc Read Only Memory), DVD-ROM (Digital Versatile Disc Read Only Memory), and USB (Universal Serial Bus) memory. The information processing program 40 may also be provided in the form of being downloaded from an external device via a network. The information processing program 40 can also be provided as a program product. A program product includes any form of product for providing a program. For example, a program product includes a program provided via a network such as the Internet, and a non-temporary computer-readable recording medium such as a CD-ROM or DVD on which the program is stored.

[0089] The following further notes are disclosed regarding the above embodiments. (Note 1) An information processing device comprising a processor, wherein the processor derives information representing the reference plane using a portion of three-dimensional point cloud data that satisfies predetermined conditions, obtained by photographing a reference plane having a portion that overlaps with the subject and the radiation detector along the irradiation direction of the radiation irradiated from the radiation source, and uses the derived information representing the reference plane to derive at least one of the distance between the radiation source and the radiation detector and the angle of incidence of the radiation irradiated from the radiation source with respect to the reference plane.

[0090] (Note 2) The information processing apparatus according to Note 1, wherein the condition is that the above condition is within a set range in the irradiation direction based on the distance between the sensor used to acquire the three-dimensional point cloud data and the body surface of the subject.

[0091] (Note 3) The information processing device according to Note 1 or Note 2, wherein the condition is within a set range in the direction of the first edge and the direction of the second edge intersecting the first edge of the angle of view of the camera used to acquire the three-dimensional point cloud data, with respect to the position of the body surface of the subject.

[0092] (Note 4) The information processing device according to any one of Notes 1 to 3, wherein the condition is that the angle made with the direction of radiation irradiation is less than the set angle.

[0093] (Note 5) The information processing apparatus according to any one of Notes 1 to 4, wherein the processor detects an area other than the reference plane, and the condition is that the area is located outside the detected area.

[0094] (Note 6) The information processing device according to Note 2, wherein the set range is set by the distance between the sensor and the body surface of the subject and two different offset values.

[0095] (Note 7) The processor is an information processing device according to Note 6, which corrects the two offset values ​​using the calibration data of the sensor.

[0096] (Note 8) The information processing apparatus described in Note 4, wherein the processor derives a unit normal vector for each data point in the three-dimensional point cloud data, and data points with an angle less than the set angle are data points where the absolute value of the components of the unit normal vector is greater than or equal to a threshold.

[0097] (Note 9) The information processing device according to Note 4, wherein the condition is that the point cloud data constitutes a shape with curvature along the irradiation direction less than a threshold.

[0098] (Note 10) The information processing apparatus described in Note 3, wherein the set range in the direction of the first side and the direction of the second side is a circular range.

[0099] (Note 11) The information processing apparatus described in Note 3, wherein the set range in the direction of the first side and the direction of the second side is a rectangular range.

[0100] (Note 12) The information processing device according to Note 11, wherein the rectangle is a rectangle whose shorter side is the axis direction of the subject's body.

[0101] (Note 13) The information processing apparatus according to Note 12, wherein the processor detects the head-to-foot direction as the body axis direction based on the keypoint detection result for the optical image obtained by photographing the subject.

[0102] (Note 14) The three-dimensional point cloud data is data obtained by converting a distance image taken of the reference plane, and the processor detects the region of the subject as a region other than the reference plane by performing clustering on the distance image, as described in Note 5.

[0103] (Note 15) The information processing apparatus according to Note 14, wherein the processor further detects the region of the torso of the subject by using the central position of the radiation irradiated onto the subject.

[0104] (Note 16) The information processing device according to Note 14, wherein the processor detects the body axis direction of the subject, and further detects the trunk region and the head region of the subject by using the detected body axis direction.

[0105] (Note 17) The information processing device according to Note 16, wherein the processor detects a region located outside the width direction of the subject with respect to the head region as a pillow region when the boundary position of the region in the body axis direction is located closer to the head than the trunk region.

[0106] (Note 18) The information processing apparatus according to Note 5, wherein the processor detects the region of the subject as a region other than the reference plane based on the detection result of keypoints in an optical image obtained by photographing the subject.

[0107] (Note 19) The information processing device according to Note 18, wherein the processor detects the area of ​​the subject by thickening the line segment connecting the joints of the keypoint by a set width.

[0108] (Note 20) The information processing apparatus according to any one of Notes 1 to 19, wherein the processor derives the distance by deriving the intersection point of a straight line along the irradiation direction and the reference plane in three-dimensional space.

[0109] (Note 21) The information processing apparatus according to any one of Notes 1 to 20, wherein the processor derives the incident angle using the normal vector of the reference plane and an inverse trigonometric function.

[0110] (Note 22) The information processing apparatus according to any one of Notes 1 to 21, wherein the processor derives information representing the reference plane by iterative processing using the aforementioned data and a regression algorithm.

[0111] (Note 23) An information processing method comprising an information processing device equipped with a processor, wherein the processor derives information representing the reference plane using a portion of three-dimensional point cloud data that satisfies predetermined conditions, obtained by photographing a reference plane having a portion that overlaps with the subject and the radiation detector along the irradiation direction of the radiation irradiated from the radiation source, and using the derived information representing the reference plane, derives at least one of the distance between the radiation source and the radiation detector and the angle of incidence of the radiation irradiated from the radiation source with respect to the reference plane.

[0112] (Note 24) An information processing program for causing the processor of an information processing device equipped with a processor to perform the following processes: derive information representing the reference plane using a portion of three-dimensional point cloud data that satisfies a defined condition, obtained by photographing a reference plane having a portion that overlaps with the subject and the radiation detector along the irradiation direction of the radiation irradiated from the radiation source; and derive at least one of the distance between the radiation source and the radiation detector and the angle of incidence of the radiation irradiated from the radiation source with respect to the reference plane using the derived information representing the reference plane.

[0113] The disclosure of Japanese Patent Application No. 2025-017422, filed on 5 February 2025, is incorporated herein by reference in its entirety. Furthermore, all documents, patent applications, and technical standards described herein are incorporated herein by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually indicated as being incorporated by reference.

Claims

1. An information processing device comprising a processor, wherein the processor derives information representing the reference plane using a portion of three-dimensional point cloud data that satisfies predetermined conditions, obtained by photographing a reference plane having a portion that overlaps with the subject and the radiation detector along the irradiation direction of the radiation irradiated from the radiation source, and uses the derived information representing the reference plane to derive at least one of the distance between the radiation source and the radiation detector and the angle of incidence of the radiation irradiated from the radiation source with respect to the reference plane.

2. The information processing apparatus according to claim 1, wherein the condition is that the above condition is within a set range in the irradiation direction based on the distance between the sensor used to acquire the three-dimensional point cloud data and the body surface of the subject.

3. The information processing apparatus according to claim 1 or 2, wherein the condition is that the condition is within a set range in the direction of the first edge and the direction of the second edge intersecting the first edge of the field of view of the camera used to acquire the three-dimensional point cloud data, with respect to the position of the body surface of the subject.

4. The information processing apparatus according to claim 1 or 2, wherein the condition is that the angle made with the irradiation direction of the radiation is less than a set angle.

5. The information processing apparatus according to claim 1 or 2, wherein the processor detects an area other than the reference plane, and the condition is that the area is located outside the detected area.

6. The information processing apparatus according to claim 2, wherein the set range is a range set by the distance between the sensor and the body surface of the subject and two different offset values.

7. The information processing apparatus according to claim 6, wherein the processor corrects the two offset values ​​using the calibration data of the sensor.

8. The information processing apparatus according to claim 4, wherein the processor derives a unit normal vector for each data point in the three-dimensional point cloud data, and data points with an angle less than the set angle are data points where the absolute value of the components of the unit normal vector is greater than or equal to a threshold.

9. The information processing apparatus according to claim 4, wherein the condition is that the point cloud data constitutes a shape with curvature along the irradiation direction less than a threshold.

10. The information processing apparatus according to claim 3, wherein the set range in the direction of the first side and the direction of the second side is a circular range.

11. The information processing apparatus according to claim 3, wherein the set range in the direction of the first side and the direction of the second side is a rectangular range.

12. The information processing apparatus according to claim 11, wherein the rectangle is a rectangle whose shorter side is the axis direction of the subject's body.

13. The information processing apparatus according to claim 12, wherein the processor detects the head-to-foot direction as the body axis direction based on the keypoint detection result for the optical image obtained by photographing the subject.

14. The information processing apparatus according to claim 5, wherein the three-dimensional point cloud data is data obtained by converting a distance image taken of the reference plane, and the processor detects the region of the subject as a region other than the reference plane by performing clustering on the distance image.

15. The information processing apparatus according to claim 14, wherein the processor further detects the region of the torso of the subject by using the central position of the radiation irradiated onto the subject.

16. The information processing apparatus according to claim 14, wherein the processor detects the body axis direction of the subject, and further detects the trunk region and the head region of the subject by using the detected body axis direction.

17. The information processing apparatus according to claim 16, wherein the processor detects a region located outside the width direction of the subject with respect to the head region as a pillow region when the boundary position of the region in the body axis direction is located closer to the head than the trunk region.

18. The information processing apparatus according to claim 5, wherein the processor detects a region of the subject as a region other than the reference plane based on the detection result of keypoints in an optical image obtained by photographing the subject.

19. The information processing apparatus according to claim 18, wherein the processor detects the area of ​​the subject by thickening the line segment connecting the joints of the keypoints by a set width.

20. The information processing apparatus according to claim 1, wherein the processor derives the distance by deriving the intersection point of a straight line along the irradiation direction and the reference plane in three-dimensional space.

21. The information processing apparatus according to claim 1, wherein the processor derives the incident angle using the normal vector of the reference plane and an inverse trigonometric function.

22. The information processing apparatus according to claim 1, wherein the processor derives information representing the reference plane by iterative processing using the aforementioned data and a regression algorithm.

23. An information processing method comprising an information processing device equipped with a processor, wherein the processor derives information representing the reference plane using a portion of three-dimensional point cloud data that satisfies predetermined conditions, obtained by photographing a reference plane having a portion that overlaps with the subject and the radiation detector along the irradiation direction of the radiation irradiated from the radiation source, and using the derived information representing the reference plane, derives at least one of the distance between the radiation source and the radiation detector and the angle of incidence of the radiation irradiated from the radiation source with respect to the reference plane.

24. An information processing program for causing the processor of an information processing device equipped with a processor to perform the following processes: derive information representing the reference plane using a portion of three-dimensional point cloud data obtained by photographing a reference plane having a portion that overlaps with the subject and the radiation detector along the irradiation direction of the radiation irradiated from the radiation source, provided that the data satisfies a predetermined condition; and use the derived information representing the reference plane to derive at least one of the distance between the radiation source and the radiation detector and the angle of incidence of the radiation irradiated from the radiation source with respect to the reference plane.

Images