Radiation generator, method for operating a radiation generator, and operating program for a radiation generator

Abstract

The present invention provides a radiation generator capable of obtaining map data that reflects the position and / or orientation of structures in the surrounding environment while suppressing the effects of external disturbances, a method for operating the radiation generator, and an operating program for the radiation generator. [Solution] The radiation generating device has a main body with a radiation source that emits radiation toward the patient, and is capable of autonomous movement on a cart with wheels. The self-position estimation and map data creation unit estimates its own position from images for driving control captured by a camera and creates map data of the surrounding environment. The self-position estimation and map data creation unit is installed on the top plate of the supine imaging table and the holder of the standing imaging table, and acquires top plate position detection data and holder position detection data from top plate position detection sensors and holder position detection sensors that detect the positions of the top plate and holder. The self-position estimation and map data creation unit updates the map data using the top plate position detection data and holder position detection data.

Description

TECHNICAL FIELD

[0001] The technology of the present disclosure relates to a radiation generating apparatus, an operation method for the radiation generating apparatus, and an operation program for the radiation generating apparatus. BACKGROUND ART

[0002] In medical practice, a mobile radiation generating apparatus is used, in which a main body having a radiation source that emits radiation toward an object can travel via a traveling mechanism equipped with wheels. For example, the mobile radiation generating apparatus is used for so-called round imaging, in which imaging of a patient as an object is performed while traveling through patient rooms. Mobile radiation generating apparatuses capable of autonomous traveling based on SLAM (Simultaneous Localization and Mapping) technology are also under investigation.

[0003] Patent Literature 1 discloses an apparatus capable of autonomous traveling based on SLAM technology, although the apparatus is a CT (Computed Tomography) apparatus and not a radiation generating apparatus. In Patent Literature 1, map data of the surrounding environment is created in accordance with output data from a LiDAR (Light Detection and Ranging) sensor mounted on the apparatus. Furthermore, in order to supplement information on blind spots or transparent structures that are difficult for the LiDAR sensor to detect, an optical sensor and a radar sensor are mounted on the apparatus, and the map data is updated using output data from these sensors. PRIOR ART DOCUMENTS PATENT DOCUMENTS

[0004] Patent Literature 1 European Patent Publication No. 4330913 Specification SUMMARY OF INVENTION PROBLEM TO BE SOLVED BY THE INVENTION

[0005] In Patent Document 1, optical sensors and radar sensors were susceptible to disturbances such as the movements of operators like radiological technologists, which could prevent the acquisition of accurate output data.

[0006] One embodiment of the technology of this disclosure provides a radiation generator, a method for operating the radiation generator, and an operating program for the radiation generator that can obtain map data reflecting the position and / or orientation of structures in the surrounding environment while suppressing the effects of disturbances. [Means for solving the problem]

[0007] The radiation generating device of this disclosure is a radiation generating device in which a main body having a radiation source that emits radiation toward an object is capable of autonomous movement by a driving mechanism having wheels, and comprises an environmental information sensor and a processor, the processor creating map data of the surrounding environment while estimating its own position from the output data of the environmental information sensor, acquiring detection data from a detection sensor provided on a structure in the surrounding environment that detects the position and / or orientation of movable components of the structure, and updating the map data using the detection data.

[0008] The processor preferably estimates the amount of travel of the travel mechanism based on the change in position of the same structure included in past output data and current output data.

[0009] The processor preferably recognizes moving components based on the detected data and estimates the amount of travel by excluding the components recognized as moving.

[0010] Map data is preferably 3D data.

[0011] It is preferable that the spatial resolution of the map data is set lower in the vertical direction than in the horizontal direction.

[0012] The processor preferably changes the travel speed of the travel mechanism according to the distance to the structure.

[0013] The processor preferably increases the degree of reduction in travel speed as it approaches the structure.

[0014] The environmental information sensor is a distance sensor, and the output data includes distance data. Preferably, the processor calibrates the distance data using the displacement of the components based on the detected data.

[0015] The structure preferably includes a supine imaging table, and the component is preferably the top plate of the supine imaging table.

[0016] The structure preferably includes a standing imaging table, and the component is preferably a holder for the standing imaging table.

[0017] The method for operating a radiation generator according to this disclosure is a method for operating a radiation generator in which a main body having a radiation source that emits radiation toward an object is capable of autonomous movement by a wheeled driving mechanism, and includes: creating map data of the surrounding environment while estimating its own position from output data of an environmental information sensor; acquiring detection data from a detection sensor provided on a structure in the surrounding environment that detects the position and / or orientation of movable components of the structure; and updating the map data using the detection data.

[0018] The operating program for the radiation generator of this disclosure is an operating program for a radiation generator in which a main body having a radiation source that emits radiation toward a subject is capable of autonomous movement by a wheeled driving mechanism, and causes a computer to perform the following processes: creating map data of the surrounding environment while estimating its own position from output data of an environmental information sensor; acquiring detection data from a detection sensor provided on a structure in the surrounding environment that detects the position and / or orientation of movable components of the structure; and updating the map data using the detection data. [Effects of the Invention]

[0019] According to the technology of the present disclosure, it is possible to provide a radiation generator, an operation method for a radiation generator, and an operation program for a radiation generator that can obtain map data reflecting structures in the surrounding environment while suppressing the influence of disturbance. [Brief Description of the Drawings]

[0020] [Figure 1] Fig. 1 is a diagram showing a state of supine-position radiography using a radiation imaging system. [Figure 2] Fig. 2 is a diagram showing a state of standing-position radiography using a radiation imaging system. [Figure 3] Fig. 3 is a block diagram showing the electrical configuration of the radiation generator. [Figure 4] Fig. 4 is a diagram showing how the radiation generator travels from a standby position toward a first target position in supine-position radiography and stops at the first target position. [Figure 5] Fig. 5 is a diagram showing how the radiation generator travels from a standby position toward a second target position in standing-position radiography and stops at the second target position. [Figure 6] Fig. 6 is a diagram showing processing of each processing unit of a processor related to travel control. [Figure 7] Fig. 7 is a diagram showing processing of each processing unit of a processor related to travel control. [Figure 8] Fig. 8 is a diagram showing how the travel distance of a carriage unit is estimated based on the amount of change in the position of the same feature point. [Figure 9] Fig. 9 is a diagram showing how moving components are recognized based on detection data, and the travel distance is estimated by excluding components recognized as moving. [Figure 10] Fig. 10 is a diagram showing how map data is updated using top board position detection data and holder position detection data. [Figure 11] Fig. 11 is a diagram showing the spatial resolution in the horizontal direction and the height direction of map data. [Figure 12] Fig. 12 is a diagram showing processing of each processing unit of a processor related to alignment control. [Figure 13] Fig. 13 is a diagram showing processing in the learning phase of a cassette contour extraction model. [Figure 14] This diagram shows the alignment control process. [Figure 15] This is a flowchart showing the shooting procedure. [Figure 16] This figure shows that when an electronic cassette is detected from the alignment control image, the system switches from driving control to alignment control. [Figure 17] This figure shows that when a patient is detected from the position control image, the system switches from driving control to position control. [Figure 18] This figure shows a method for extracting a patient's joint points from an image used for alignment control. [Figure 19] This table shows how the travel speed of the bogie is changed according to the distance to the structure. [Figure 20] This diagram shows a supine imaging table with a rotatable top. [Figure 21] This figure shows a method for calibrating distance sensor distance data using the displacement of components based on detected data. [Modes for carrying out the invention]

[0021] [First Embodiment] As an example, as shown in Figures 1 and 2, the radiography system 10 comprises a radiation generator 11 and an electronic cassette 12. The radiation generator 11 has a main body 14 having a radiation source 13 that emits radiation R toward the patient P, mounted on a trolley 16 having wheels 15. The radiation R is, for example, X-rays. The radiation generator 11 is equipped with a battery and can be powered by the battery. The radiation generator 11 can also be powered by a commercial power supply via a power cord. The trolley 16 is an example of a "traveling mechanism" related to the technology of this disclosure. The patient P is an example of a "subject" related to the technology of this disclosure.

[0022] The radiation generator 11 is movable within the imaging room RM (see also Figures 5 and 6). The radiation generator 11 is also used for so-called mobile imaging, where images are taken of patients P while moving around the patient rooms. For this reason, the radiation generator 11 is also called a mobile imaging unit. Alternatively, the radiation generator 11 can be used for imaging in the emergency room. Furthermore, the radiation generator 11 can be brought into the operating room and used during surgery.

[0023] The electronic cassette 12 has a configuration in which a sensor panel, which acts as a radiation detector for detecting radiation R, is built into a portable housing. The electronic cassette 12 is powered by a battery and communicates wirelessly with the radiation generator 11. The sensor panel has a detection surface 17 in which multiple pixels are arranged in a matrix and generate a signal charge in response to radiation R or visible light converted from radiation R by a scintillator. The electronic cassette 12 detects the radiation R that is irradiated from the radiation source 13 and passes through the patient P, and outputs a radiation image 18 of the patient P.

[0024] Since the electronic cassette 12 is portable and wireless, it can be used for so-called free radiography, as shown in Figure 1, by placing it on the top plate 20 (under the patient P) of the supine radiography table (bed) 19 installed in the radiography room RM and performing radiography. More specifically, Figure 1 shows the electronic cassette 12 being inserted between the top plate 20 and the patient P to take an image of the patient P lying supine on the top plate 20 (see also Figure 5). In addition, as shown in Figure 2, the electronic cassette 12 can also be used by housing it in the holder 22 of the standing radiography table 21 installed in the radiography room RM (see also Figure 6).

[0025] The top plate 20 of the supine imaging table 19 is movable in the forward, backward, left, and right directions. The top plate 20 is also movable up and down. The top plate 20 is equipped with a top plate position detection sensor 23. The top plate position detection sensor 23 detects the horizontal and height positions of the top plate 20. The top plate position detection sensor 23 is, for example, a rotary encoder, a potentiometer, a gyro sensor, or a combination of several such sensors. The supine imaging table 19 is an example of a "structure" related to the technology of this disclosure. The top plate 20 is an example of a "component" related to the technology of this disclosure. The top plate position detection sensor 23 is an example of a "detection sensor" related to the technology of this disclosure.

[0026] The holder 22 of the standing imaging table 21 is vertically movable relative to the support column 24. The support column 24 is provided with a holder position detection sensor 25. The holder position detection sensor 25 detects the height position of the holder 22. The holder position detection sensor 25 is, for example, a rotary encoder, a potentiometer, a gyro sensor, or a combination of several such sensors. The standing imaging table 21 is an example of a "structure" related to the technology of this disclosure. The holder 22 is an example of a "component" related to the technology of this disclosure. The holder position detection sensor 25 is an example of a "detection sensor" related to the technology of this disclosure.

[0027] The main body 14 is rectangular in shape and is erected in the center of the trolley 16. The main body 14 is divided into a front section 26 and a rear section 27. The front section 26 can move up and down relative to the rear section 27. The rear section 27 is fixed to the trolley 16.

[0028] The base end of the arm 28 is attached to the front part 26. More specifically, the arm 28 is divided into a first part to which the base end is attached to the front part 26, and a second part to which the base end is attached to the first part. The radiation source 13 is attached to the tip of the second part, which is the free end opposite the base end.

[0029] The first part is vertically movable relative to the front part 26 and is also vertically foldable relative to the front part 26. The second part is vertically foldable relative to the first part. Furthermore, the second part is extendable and retractable. The radiation source 13 is rotatable, i.e., pivotable, relative to the second part. By displacing the arm 28 and the radiation source 13 relative to the arm 28, the height, horizontal position, and orientation (direction) of the radiation source 13 can be adjusted.

[0030] An operation panel 29 is provided on the upper surface of the rear section 27. The operation panel 29 is, for example, composed of a touch panel display and has the function of displaying information in addition to operation functions. The operation panel 29 is operated by an operator OP, such as a radiological technologist. The operator OP sets the radiation irradiation conditions R through the operation panel 29. The operator OP also checks the radiation image 18 through the operation panel 29.

[0031] Furthermore, an irradiation switch (not shown) is provided at the rear 27. The irradiation switch is used by the operator OP to instruct the start of radiation irradiation. An extension cable is connected to the irradiation switch, and it can be detached from the rear 27 for use. The irradiation switch can only be operated after the travel control and positioning control, which will be described later, have been completed.

[0032] The radiation source 13 includes a radiation tube 30 and an irradiation field limiter 31. The radiation tube 30 generates radiation R. The radiation tube 30 is equipped with a filament, a target, a grid electrode, etc. (all not shown). A tube voltage is applied between the filament, which is the cathode, and the target, which is the anode, from a voltage generator (not shown) built into the rear section 27. The filament emits thermionic electrons toward the target in accordance with the applied tube voltage. The target emits radiation R in response to collisions with thermionic electrons emitted from the filament. The grid electrode is placed between the filament and the target and 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. The tube voltage and tube current are set in the radiation source 13 as irradiation conditions, along with the irradiation time.

[0033] The irradiation field limiter 31, also called a collimator, limits the irradiation field of radiation R generated from the radiation tube 30. The irradiation field limiter 31 has a configuration in which, for example, four shielding plates made of lead or the like are placed on each side of a rectangle to shield the radiation R, and a rectangular exit aperture that transmits radiation is formed in the center. The irradiation field limiter 31 changes the size of the exit aperture by changing the position of each shielding plate, thereby changing the irradiation field of radiation R.

[0034] A camera 32 is attached to the radiation source 13. The camera 32 is used to assist in aligning the radiation source 13 with the electronic cassette 12. The camera 32 incorporates a visible light-sensitive image sensor, such as a CMOS (Complementary Metal Oxide Semiconductor) image sensor or a CCD (Charge-coupled Device) image sensor. The camera 32 captures moving images at a predetermined frame rate.

[0035] There are a total of four wheels 15, located on the front, rear, left, and right sides of the bogie 16. In other words, the bogie 16 is a four-wheeled type. Each wheel 15 is a swivel type, for example, that rotates around a pivot axis that extends in the vertical direction (also called the vertical direction) perpendicular to the axis of rotation when the vehicle is moving. The main body 14 moves autonomously using this bogie 16.

[0036] Here, autonomous driving refers to autonomously driving toward a set target location while recognizing the surrounding environment using environmental information sensors and estimating its own position. The target locations are pre-set, for example, the position beside the supine imaging table 19 shown in Figure 1, and the position facing the standing imaging table 21 shown in Figure 2. The radiation generating device 11 uses SLAM technology to achieve autonomous driving.

[0037] A camera 33 is provided on the trolley section 16. The camera 33 is used to support the autonomous movement of the main body section 14. The camera 33 photographs the area in front of the trolley section 16. Like the camera 32, the camera 33 incorporates a visible light-sensitive image sensor and captures moving images at a predetermined frame rate. The camera 33 is an example of an "environmental information sensor" related to the technology disclosed herein. The main body section 14, and by extension the radiation generator 11, can be driven not only autonomously but also manually by an operator OP.

[0038] Figure 3 is a block diagram showing an example of the electrical configuration of the radiation generator 11. The processor 40 comprehensively controls the entire radiation generator 11. The processor 40 consists, for example, of a CPU (Central Processing Unit) and memory such as RAM (Random Access Memory), and functions as various processing units by loading and executing various programs into memory. Specifically, the processor 40 performs driving control, positioning control, and imaging control. Driving control is the control related to the autonomous driving of the main unit 14. Positioning control is the control related to the positioning of the radiation source 13 and the electronic cassette 12. Imaging control includes the control of radiation R irradiation by the radiation source 13 and the output control of the radiation image 18 by the electronic cassette 12.

[0039] The storage 41 consists of non-volatile memory such as a hard disk drive or a solid-state drive. The storage 41 stores an operating program 42 and control data 43. The operating program 42 is an example of an "operating program for a radiation generator" related to the technology of this disclosure. The control data 43 includes data for driving control, data for alignment control, and data for imaging control. The data for driving control includes map data 73 (see Figure 7) created by SLAM technology. The data for alignment control includes a cassette contour extraction model 86 (see Figure 12) for extracting the contour OLC (see Figure 12) of the electronic cassette 12. The data for imaging control includes an irradiation condition table in which the irradiation conditions for each imaging area are registered.

[0040] The processor 40 is connected to a top plate position detection sensor 23, a holder position detection sensor 25, an operation panel 29, a communication interface 50, a travel actuator 51, a positioning actuator 52, and a radiation source position detection sensor 53. The processor 40 receives top plate position detection data 74 (see Figure 7) from the top plate position detection sensor 23 and holder position detection data 75 (see Figure 7) from the holder position detection sensor 25. The processor 40 uses the top plate position detection data 74 and the holder position detection data 75 to update the surrounding environment map data 73 (see Figure 7).

[0041] The processor 40 controls the display of various screens on the control panel 29. The processor 40 also receives various operation instructions from the operator OP through the control panel 29 and executes various controls according to those instructions. The communication I / F 50 is, for example, a wireless communication I / F and is responsible for wireless communication with the electronic cassette 12.

[0042] The travel actuator 51 includes a motor for rotating the wheels 15 under the control of the processor 40, and a motor for turning them. The alignment actuator 52 includes a motor for raising and lowering the front section 26, a motor for raising and lowering the arm 28, a motor for folding the second section of the arm 28, a motor for extending and retracting the second section, and a motor for rotating the radiation source 13 relative to the second section under the control of the processor 40.

[0043] The radiation source position detection sensor 53 measures the vertical direction and amount of vertical movement of the arm 28 relative to the front part 26, the bending direction and amount of bending of the second part of the arm 28 relative to the first part, the extension and retraction direction and amount of extension and retraction of the second part, and the rotation direction and amount of rotation of the radiation source 13 relative to the second part. The radiation source position detection sensor 53 is, for example, a rotary encoder, a potentiometer, or a gyro sensor, or a combination of several of these sensors. The radiation source position detection sensor 53 outputs the measured values ​​to the processor 40. The processor 40 derives the position and orientation of the radiation source 13 based on the measured values ​​of the radiation source position detection sensor 53.

[0044] As an example, as shown in Figures 4 and 5, a standby position HP for the radiation generator 11 is provided in a corner of the imaging room RM. At the standby position HP, the battery is charged, imaging orders are received from the Radiology Information System (RIS), and irradiation conditions are set. The standby position HP occupies an area that is the same size as the radiation generator 11, or slightly larger.

[0045] In the imaging room RM, a first target position TP1 (see Figure 4) for the radiation generator 11 during supine imaging and a second target position TP2 (see Figure 5) for the radiation generator 11 during upright imaging are set. The first target position TP1 is located beside the supine imaging table 19, more specifically, directly opposite one of the longer sides of the supine imaging table 19. The second target position TP2 is directly opposite the upright imaging table 21, separated from the upright imaging table 21 by the SID distance required for upright imaging. The first target position TP1 and the second target position TP2 occupy the same area as the radiation generator 11, or slightly larger, than the radiation generator 11, similar to the standby position HP. In the following, the first target position TP1 and the second target position TP2 may be collectively referred to as target position TP.

[0046] During supine imaging, the radiation generator 11 travels from the standby position HP towards the first target position TP1 and stops when its own position PS (see Figure 7) reaches the first target position TP1 (see Figure 4). During upright imaging, the radiation generator 11 travels from the standby position HP towards the second target position TP2 and stops when its own position PS reaches the second target position TP2 (see Figure 5). Although not shown in the illustration, the radiation generator 11 may travel from the first target position TP1 towards the second target position TP2 in order to perform upright imaging after supine imaging. Also, the radiation generator 11 may travel from the second target position TP2 towards the first target position TP1 in order to perform supine imaging after upright imaging. Alternatively, it may be determined that the self-position PS has reached the target position TP when the self-position PS coincides with the center of the target position TP, or it may be determined that the self-position PS has reached the target position TP when the self-position PS falls within a range that includes the center of the target position TP with a certain margin.

[0047] Autonomous driving to the target position TP is initiated, for example, by an instruction from an operator OP via the control panel 29. Alternatively, the system may be configured to allow the start of autonomous driving to be initiated via a remote controller, in addition to or instead of the control panel 29.

[0048] The processor 40 performs driving control using SLAM technology. Specifically, as shown in Figures 6 and 7, the processor 40 functions as an image acquisition unit 60, a feature point extraction unit 61, a self-position estimation / map data creation unit 62, and a driving control unit 63.

[0049] The image acquisition unit 60 sequentially acquires images 70 of the surrounding environment of the radiation generator 11 (hereinafter referred to as "driving control images") captured by the camera 33 at a predetermined frame rate. After performing preprocessing such as noise reduction and distortion correction on the driving control images 70, the image acquisition unit 60 outputs the driving control images 70 to the feature point extraction unit 61. The driving control images 70 are an example of "output data" related to the technology of this disclosure.

[0050] The feature point extraction unit 61 uses algorithms such as ORB (Oriented FAST (Features from Accelerated Segment Test) and Rotated BRIEF (Binary Robust Independent Elementary Features)) or SURF (Speeded-Up Robust Features) to extract corners and other features of structures in the surrounding environment captured in the driving control image 70 as feature points FP. The feature point extraction unit 61 outputs the feature point extraction results 71 to the self-position estimation / map data creation unit 62. Although not shown in the diagram, the feature point extraction unit 61 also stores the feature point extraction results 71 in the storage 41. The feature point extraction results 71 are pairs of the coordinates of each feature point FP and their feature vectors.

[0051] The self-position estimation and map data creation unit 62 estimates the self-position PS of the radiation generator 11 and creates map data 73 of the surrounding environment. The self-position estimation and map data creation unit 62 receives the feature point extraction results 71 from the feature point extraction unit 61. The self-position estimation and map data creation unit 62 also receives feature point extraction results 71P for multiple past frames (hereinafter referred to as feature point extraction results (past)) and map data 73P for multiple past frames (hereinafter referred to as map data (past)). Furthermore, the self-position estimation and map data creation unit 62 receives top plate position detection data 74 from the top plate position detection sensor 23 and holder position detection data 75 from the holder position detection sensor 25. The top plate position detection data 74 and holder position detection data 75 are examples of "detection data" related to the technology of this disclosure.

[0052] The feature point extraction results (past data) 71P and map data (past data) 73P are stored in storage 41 as data for driving control of the control data 43. The feature point extraction results 71 and map data 73 that make up the feature point extraction results (past data) 71P and map data (past data) 73P consist of, for example, tens to hundreds of frames. The feature point extraction results (past data) 71P and map data (past data) 73P include feature point extraction results 71 and map data 73 that are considered to play an important role in estimating the self-position PS and creating the map data 73. Such feature point extraction results 71 and map data 73 are, for example, feature point extraction results 71 and map data 73 obtained every certain distance traveled. Also, for example, feature point extraction results 71 and map data 73 obtained when there has been a large change in viewpoint from the previous frame. Furthermore, for example, feature point extraction results 71 and map data 73 obtained when more than a set amount of new feature points FP have been extracted.

[0053] The self-localization and map data creation unit 62 compares the feature points FP from the feature point extraction result 71 from the feature point extraction unit 61 with the feature points FP from the feature point extraction result (past data) 71P. In this process, the self-localization and map data creation unit 62 refers to the feature vector of each feature point FP. More specifically, it recognizes feature points FP from the feature point extraction result 71 from the feature point extraction unit 61 and feature points FP from the feature point extraction result (past data) 71P where the distance (Euclidean distance, etc.) of the feature vectors is less than a threshold as the same feature point FP. Based on this comparison result of feature points FP and the map data (past data) 73P, the self-localization and map data creation unit 62 estimates the self-localization PS.

[0054] As an example, as shown in Figure 8, the self-position estimation and map data creation unit 62 estimates the travel distance of the trolley unit 16 based on the change in the position of the same feature point FP in the feature point extraction result 71PF of the previous frame and the feature point extraction result 71CF of the current frame. Then, it estimates the self-position PS by referring to this estimated travel distance of the trolley unit 16. The feature point extraction result 71PF of the previous frame is an example of "past output data" related to the technology of this disclosure. The feature point extraction result 71CF of the current frame is an example of "current output data" related to the technology of this disclosure.

[0055] However, as shown in Figure 9 as an example, the self-position estimation and map data creation unit 62 recognizes moving components based on the detected data and estimates the travel distance of the trolley unit 16 by excluding feature points FP related to the components recognized as moving. Figure 9 shows an example in which the top plate 20 is recognized as being rising based on top plate position detection data 74, and the travel distance of the trolley unit 16 is estimated by excluding feature points FP related to the top plate 20. The determination of whether or not a feature point FP is a feature point FP related to a component is made, for example, based on an image obtained by inputting the travel control image 70 into a semantic segmentation model, which distinguishes each structure and each component shown in the travel control image 70.

[0056] The self-localization and map data creation unit 62 creates new map data 73 (updates map data 73) based on the feature point extraction results 71 from the feature point extraction unit 61, the feature point extraction results (past) 71P, the map data (past) 73P, and the estimation results 72. In this case, as shown in Figure 10 as an example, the self-localization and map data creation unit 62 further updates the map data 73 using the top plate position detection data 74 and the holder position detection data 75. More specifically, the self-localization and map data creation unit 62 changes the horizontal and height positions of the top plate 20 in the map data 73 according to the horizontal and height positions of the top plate 20 represented by the top plate position detection data 74. In addition, the self-localization and map data creation unit 62 changes the height position of the holder 22 in the map data 73 according to the height position of the holder 22 represented by the holder position detection data 75. Figure 10 illustrates the case where the height positions of the top plate 20 and the holder 22 are changed. In this way, the self-position estimation and map data creation unit 62 performs the estimation of the self-position PS and the creation of map data 73 in parallel during the autonomous driving process of the radiation generator 11. However, the self-position estimation and map data creation unit 62 does not estimate the self-position PS until a sufficient amount of feature point extraction results (past data) 71P have been accumulated, and only creates (updates) the map data 73.

[0057] The self-position estimation and map data creation unit 62 outputs the self-position estimation result 72 of the self-position PS to the driving control unit 63. The driving control unit 63 controls the driving of the driving actuator 51 to set the self-position PS to the target position TP. Although not shown in the diagram, the self-position estimation and map data creation unit 62 also stores the map data 73 in the storage 41.

[0058] Map data 73 specifically consists of 3D data of the radiography room RM, including structures such as the supine radiography table 19 and the standing radiography table 21. Additionally, the standby position HP and target position TP are registered in map data 73.

[0059] As an example, as shown in Figure 11, the map data 73 has a lower spatial resolution in the vertical direction than in the horizontal direction. For example, the horizontal spatial resolution is 5 cm, while the vertical spatial resolution is 10 times higher, at 50 cm.

[0060] When the travel actuator 51 is driven under the control of the travel control unit 63 and the self-position PS of the radiation generator 11 becomes the target position TP, the processor 40 switches from travel control to positioning control.

[0061] As an example, as shown in Figure 12, the processor 40 functions as a cassette contour extraction unit 80, a source position and orientation derivation unit 81, and an alignment control unit 82.

[0062] The cassette contour extraction unit 80 receives sequential images 85 (hereinafter referred to as alignment control images) 85, which include the patient P and the electronic cassette 12, captured by the camera 32 at a predetermined frame rate. When the self-position PS of the radiation generator 11 is set to the target position TP, the alignment control image 85 shows the patient P and the electronic cassette 12. In Figure 12, supine imaging is shown as an example, so the supine imaging table 19 is also shown in the alignment control image 85. It is assumed that the alignment of the patient P and the electronic cassette 12 has been completed by the operator OP before the travel control and alignment control.

[0063] The cassette contour extraction unit 80 extracts the contour OLC of the electronic cassette 12 from the alignment control image 85 using the cassette contour extraction model 86. The contour OLC reveals the center CC (see Figure 14) of the detection surface 17 of the electronic cassette 12, and the orientation of the electronic cassette 12 relative to the radiation source 13. The cassette contour extraction unit 80 outputs the cassette contour extraction result 87 to the alignment control unit 82. Here, the contour OLC is exemplified as a contour that follows all sides of the electronic cassette 12, but it is not limited to this. The four corners of the electronic cassette 12 may also be extracted as contours.

[0064] The cassette contour extraction model 86 is a pre-trained model, for example, composed of a convolutional neural network, which is trained to output a cassette contour extraction result 87 when an alignment control image 85 is input. The cassette contour extraction model 86 is stored in the storage 41 as alignment control data for the control data 43.

[0065] As an example, as shown in Figure 13, in the learning phase, the cassette contour extraction model 86 is given training data 90. The training data 90 consists of a training alignment control image 85L and ground truth data 87CA. The training alignment control image 85L shows an electronic cassette 12. The ground truth data 87CA is data annotated with the contour OLC of the electronic cassette 12 shown in the training alignment control image 85L, and is, so to speak, data for checking the answer.

[0066] The cassette contour extraction model 86 receives a training alignment control image 85L as input. The cassette contour extraction model 86 outputs a training cassette contour extraction result 87L in response to the input of the training alignment control image 85L. Based on this training cassette contour extraction result 87L and the ground truth data 87CA, a loss calculation is performed on the cassette contour extraction model 86 using a loss function. Then, based on the result of the loss calculation, various coefficients of the cassette contour extraction model 86 (such as the coefficients of the convolutional layer filters) are updated, and the cassette contour extraction model 86 is updated according to the update settings.

[0067] In the learning phase of the cassette contour extraction model 86, the above series of processes—inputting the learning alignment control image 85L into the cassette contour extraction model 86, outputting the learning cassette contour extraction result 87L from the cassette contour extraction model 86, loss calculation, update settings, and updating the cassette contour extraction model 86—are repeated while the learning data 90 is exchanged. The repetition of the above series of processes ends when the extraction accuracy of the learning cassette contour extraction result 87L reaches a predetermined set level. The cassette contour extraction model 86, whose extraction accuracy has reached the set level, is stored in the storage 41 and used by the cassette contour extraction unit 80. Alternatively, learning may be terminated after the above series of processes has been repeated a set number of times, regardless of the extraction accuracy of the learning cassette contour extraction result 87L.

[0068] The radiation source position and orientation derivation unit 81 derives the position and orientation of the radiation source 13 based on the measurement values ​​of the radiation source position detection sensor 53. The radiation source position and orientation derivation unit 81 outputs the derivation result 88 to the alignment control unit 82. The derivation result 88 consists of the coordinates in the alignment control image 85 of the irradiation center RC (see Figure 14) of the radiation R and the coordinates in the alignment control image 85 of the rectangular frame F (see Figure 14) that indicates the orientation of the radiation source 13. Since the positional relationship between the radiation source 13 and the camera 32 is known, the coordinates in the alignment control image 85 of the irradiation center RC and frame F can be easily calculated from the measurement values ​​of the radiation source position detection sensor 53.

[0069] The alignment control unit 82 controls the drive of the alignment actuator 52 to align the radiation source 13 and the electronic cassette 12 directly. More specifically, as shown in Figure 14 as an example, if the orientation of the radiation source 13 is tilted relative to the electronic cassette 12, the alignment control unit 82 rotates the radiation source 13 to correct the tilt. Also, if the irradiation center RC of the radiation R and the center CC of the detection surface 17 of the electronic cassette 12 are misaligned, the alignment control unit 82 moves the radiation source 13 to correct the misalignment. Note that the tilt may be corrected after correcting the misalignment between the irradiation center RC of the radiation R and the center CC of the detection surface 17 of the electronic cassette 12. Furthermore, Figure 14 illustrates the case where the orientation of the radiation source 13 is tilted around the normal of the detection surface 17, but it is not limited to this. Even if the orientation of the radiation source 13 is tilted around an axis along the long side or short side of the detection surface 17, the alignment control unit 82 rotates the radiation source 13 to correct the tilt.

[0070] Next, the operation of the above configuration will be explained using the flowchart shown in Figure 15 as an example. Before imaging, the radiation generator 11 is waiting at the standby position HP in the imaging room RM.

[0071] A shooting order is transmitted from the radiation information system to the radiation generator 11. The operator OP operates the control panel 29 to set the irradiation conditions according to the shooting order (step ST100).

[0072] The operator OP places the electronic cassette 12 on the top plate 20 of the supine imaging table 19 if it is a supine imaging, or places the electronic cassette 12 in the holder 22 of the standing imaging table 21 if it is an upright imaging. Then, the operator OP aligns the patient P with the electronic cassette 12 (step ST110). Specifically, the center IC of the imaging area of ​​the patient P (see Figure 18) is aligned with the center CC of the detection surface 17 of the electronic cassette 12. In addition, the body axis (head-to-tail axis) of the patient P is made parallel to the long side of the detection surface 17 of the electronic cassette 12.

[0073] In response to instructions from the operator OP, the autonomous movement of the radiation generator 11 from the standby position HP towards the target position TP is initiated (step ST120). During autonomous movement, the processor 40 performs driving control using SLAM technology, as shown in Figures 6 and 7 as an example.

[0074] Specifically, first, the driving control image 70 captured by the camera 33 is acquired by the image acquisition unit 60. The driving control image 70 is output from the image acquisition unit 60 to the feature point extraction unit 61. The feature point extraction unit 61 then extracts feature points FP of structures present in the surrounding environment captured in the driving control image 70. The feature point extraction result 71 is output from the feature point extraction unit 61 to the self-position estimation / map data creation unit 62. The feature point extraction result 71 is also stored in the storage 41.

[0075] In the self-position estimation / map data creation unit 62, the self-position PS of the radiation generator 11 is estimated based on the feature point extraction results 71, the feature point extraction results (past) 71P, and the map data (past) 73P, and the estimation result 72 is output. At this time, as shown in Figure 8, the self-position estimation / map data creation unit 62 estimates the travel distance of the trolley unit 16 based on the change in the position of the same feature point FP in the feature point extraction results 71PF of the previous frame and the feature point extraction results 71CF of the current frame, and the self-position PS is estimated by referring to the travel distance as well. However, the travel distance estimation is performed excluding feature points FP related to the top plate 20 or holder 22 that are recognized as being in motion. The self-position PS estimation result 72 is output from the self-position estimation / map data creation unit 62 to the travel control unit 63.

[0076] Furthermore, in the self-position estimation and map data creation unit 62, map data 73 is created based on the feature point extraction results 71, the feature point extraction results (past) 71P, the map data (past) 73P, and the estimation results 72. At this time, as shown in Figure 10, the self-position estimation and map data creation unit 62 further updates the map data 73 using the top plate position detection data 74 and the holder position detection data 75. The map data 73 is stored in the storage 41.

[0077] Under the control of the travel control unit 63, the drive of the travel actuator 51 is controlled to set the self-position PS to the target position TP. In this way, travel control is performed. Travel control continues until the self-position PS of the radiation generator 11 reaches the target position TP (NO in step ST130).

[0078] When the self-position PS of the radiation generator 11 becomes the target position TP (YES in step ST130), the control switches from travel control to alignment control (step ST140).

[0079] As shown in Figure 12, the alignment control image 85 captured by the camera 32 is input to the cassette contour extraction unit 80. The cassette contour extraction unit 80 uses the cassette contour extraction model 86 to extract the OLC contour of the electronic cassette 12 from the alignment control image 85. The cassette contour extraction result 87 is output from the cassette contour extraction unit 80 to the alignment control unit 82.

[0080] The radiation source position and orientation derivation unit 81 derives the position and orientation of the radiation source 13 based on the measurement values ​​of the radiation source position detection sensor 53. The derivation results 88 of the position and orientation of the radiation source 13 are output from the radiation source position and orientation derivation unit 81 to the alignment control unit 82.

[0081] Under the control of the alignment control unit 82, the drive of the alignment actuator 52 is controlled to align the radiation source 13 and the electronic cassette 12 directly. Alignment control is performed in this manner. Alignment control continues until the radiation source 13 and the electronic cassette 12 are aligned directly (NO in step ST150).

[0082] When the radiation source 13 and the electronic cassette 12 are facing each other directly (YES in step ST150), the irradiation switch can be operated. The operator OP operates the irradiation switch to instruct the start of radiation irradiation. As a result, radiation R is irradiated from the radiation source 13, and the radiation image 18 is output from the electronic cassette 12 (step ST160).

[0083] As described above, the radiation generator 11 has a main body 14 having a radiation source 13 that emits radiation R toward the patient P, and is capable of autonomous movement on a trolley 16 having wheels 15. The self-position estimation and map data creation unit 62 estimates its own position PS from the driving control image 70 captured by the camera 33 and creates map data 73 of the surrounding environment. The self-position estimation and map data creation unit 62 acquires top plate position detection data 74 and holder position detection data 75 from top plate position detection sensors 23 and holder position detection sensors 25, which are installed on the top plate 20 of the supine imaging table 19 and the holder 22 of the standing imaging table 21 and detect the positions of the top plate 20 and the holder 22. The self-position estimation and map data creation unit 62 updates the map data 73 using the top plate position detection data 74 and holder position detection data 75.

[0084] Therefore, unlike the optical sensor and radar sensor described in Patent Document 1, the positions of the top plate 20 and holder 22 can be accurately determined without being affected by disturbances such as the movement of the operator OP. Thus, it is possible to obtain map data 73 that reflects the positions of the top plate 20 and holder 22 while suppressing the effects of disturbances. The target position TP can be reached via an appropriate travel path, such as the shortest path or a path that avoids obstacles. Furthermore, the map data 73 can be updated more easily than updating the map data according to the output data of the LiDAR sensor using the output data of the optical sensor and radar sensor, as in Patent Document 1. Moreover, it is possible to obtain map data 73 that reflects the real-time positions of components such as the top plate 20.

[0085] As shown in Figure 8, the self-position estimation and map data creation unit 62 estimates the travel distance of the trolley unit 16 based on the change in position of the same feature point FP included in the feature point extraction result 71PF of the previous frame and the feature point extraction result 71CF of the current frame. This contributes to improving the accuracy of the estimation of the self-position PS.

[0086] As shown in Figure 9, the self-position estimation and map data creation unit 62 recognizes the moving top plate 20 and holder 22 based on the top plate position detection data 74 and holder position detection data 75, and estimates the distance traveled by excluding the feature points FP related to the top plate 20 and holder 22 that are recognized as moving. Feature points FP related to the moving top plate 20 and holder 22 become noise in the estimation of the distance traveled. Therefore, by excluding the feature points FP related to the moving top plate 20 and holder 22 from the estimation of the distance traveled, it is possible to further improve the accuracy of the self-position PS estimation.

[0087] As shown in Figure 7, the map data 73 is three-dimensional data. Therefore, it is possible to create map data 73 that includes structures with height and whose height position changes, such as the supine imaging table 19 and the standing imaging table 21. It is possible to reach the target position TP using an appropriate travel path, such as the shortest path or a path that avoids obstacles.

[0088] As shown in Figure 11, the map data 73 has a lower spatial resolution in the vertical direction than in the horizontal direction. Therefore, it is possible to include more information about the surrounding environment in the horizontal direction, which is important for autonomous driving, in the map data 73. In addition, the processing load for creating the map data 73 can be reduced.

[0089] As shown in Figures 1 and 2, the structure includes a supine imaging table 19 and a standing imaging table 21, and the components are the top plate 20 of the supine imaging table 19 and the holder 22 of the standing imaging table 21. The supine imaging table 19 and the standing imaging table 21 are almost always installed in an imaging room (RM). Furthermore, the top plate 20 and the holder 22 are usually movable. Therefore, by defining the structure as the supine imaging table 19 and the standing imaging table 21, and the components as the top plate 20 and the holder 22, it is possible to create map data 73 that is compatible with the configuration of a typical imaging room (RM). [Second Embodiment] In the first embodiment described above, the control switches from driving control to alignment control when the self-position PS becomes the target position TP, but this is not limited to this. For example, the control may be as shown in Figures 16 and 17.

[0090] As shown in Figure 16, the processor 40 of this embodiment functions as a cassette detection unit 95 in addition to the processing units of the first embodiment described above. The cassette detection unit 95 receives an alignment control image 85 as input. The cassette detection unit 95 uses image recognition technology to detect whether or not an electronic cassette 12 is visible in the alignment control image 85.

[0091] In this embodiment, the processor 40 switches from driving control to alignment control when the cassette detection unit 95 detects that the electronic cassette 12 is visible in the alignment control image 85. The cassette detection unit 95 may also use a trained model, such as the cassette contour extraction model 86 of the first embodiment, to detect whether or not the electronic cassette 12 is visible in the alignment control image 85.

[0092] Alternatively, as shown in Figure 17, the processor 40 of this embodiment functions as a patient detection unit 100 in addition to the processing units of the first embodiment described above. The patient detection unit 100 receives an image 85 for alignment control. The patient detection unit 100 uses image recognition technology to detect whether or not patient P is visible in the image 85 for alignment control. When the patient detection unit 100 detects that patient P is visible in the image 85 for alignment control, the processor 40 of this embodiment switches from driving control to alignment control. Switching from driving control to alignment control when the electronic cassette 12 or patient P is detected in the image 85 for alignment control allows for a timely transition from driving control to alignment control, just as when the self-position PS becomes the target position TP, enabling smooth execution of alignment control.

[0093] [Third Embodiment] In the first embodiment described above, the contour OLC of the electronic cassette 12 is extracted from the alignment control image 85, but this is not limited to this. For example, it may be as shown in Figure 18.

[0094] In Figure 18, the processor 40 of this embodiment functions as an articular point extraction unit 105 instead of the cassette contour extraction unit 80 of the first embodiment. The articular point extraction unit 105 receives sequential input of alignment control images 85. The articular point extraction unit 105 extracts the articular points J of patient P from the alignment control images 85 using the articular point extraction model 106. Figure 18 illustrates the case where both shoulder joints, both elbow joints, and both hip joints are extracted as articular points J. The lines connecting each articular point J indicate the center IC of the imaging area of ​​patient P and the posture of patient P relative to the radiation source 13. The articular point extraction model 106 is a trained model composed of, for example, a convolutional neural network, similar to the cassette contour extraction model 86. The articular point extraction unit 105 outputs the articular point extraction result 107 to the alignment control unit 82.

[0095] If the orientation of the radiation source 13 is tilted relative to the patient P, the alignment control unit 82 controls the drive of the alignment actuator 52 to rotate the radiation source 13 to correct the tilt. Also, if the irradiation center RC of the radiation R and the center IC of the imaging area of ​​the patient P are misaligned, the alignment control unit 82 controls the drive of the alignment actuator 52 to move the radiation source 13 to correct the misalignment. By extracting the joint points J of the patient P from the alignment control image 85 in this way, it is possible to contribute to more accurate alignment of the radiation source 13 and the electronic cassette 12.

[0096] As an example, as shown in Table 110 in FIG. 19, the travel control unit 63 may change the travel speed of the truck unit 16 in accordance with the estimated self-position PS and the distance to a structure obtained from map data 73. Specifically, the shorter the distance to the structure is, the higher the degree of reduction in travel speed the travel control unit 63 implements. That is, when the distance D to the structure is equal to or greater than the first distance threshold DTH1 (DTH1 ≤ D), the travel control unit 63 causes the truck unit 16 to travel at a travel speed V1. When the distance D to the structure is equal to or greater than the second distance threshold DTH2 and less than the first distance threshold DTH1 (DTH2 ≤ D < DTH1), the travel control unit 63 causes the truck unit 16 to travel at a travel speed V2 slower than the travel speed V1. When the distance D to the structure is less than the second distance threshold DTH2 (D < DTH2), the travel control unit 63 causes the truck unit 16 to travel at a travel speed V3 slower than the travel speed V2. The first distance threshold DTH1 is, for example, 1 m, and the second distance threshold DTH2 is, for example, 30 cm. The travel speed V2 is, for example, 1 / 2 of the travel speed V1, and the travel speed V3 is, for example, 1 / 10 of the travel speed V1.

[0097] By doing this, when the radiation generator 11 is likely to collide with a structure, it is possible to secure a time margin for retracting the structure or changing the travel route to avoid the structure. Further, even if the radiation generator 11 collides with a structure, the impact can be reduced. The safety of an operator OP and a patient P can also be ensured.

[0098] As an example, as shown in Figure 20, the top plate 20 of the supine imaging table 19 may rotate around a vertical axis. In this case, in addition to the top plate position detection sensor 23, the top plate 20 is provided with a top plate attitude detection sensor 115. The top plate attitude detection sensor 115 detects the attitude (rotational position) of the top plate 20. The top plate attitude detection sensor 115 is, for example, a rotary encoder, a potentiometer, a gyro sensor, or a combination of several of these sensors. The top plate attitude detection sensor 115 is an example of a "detection sensor" related to the technology of this disclosure. In this case, the self-position estimation / map data creation unit 62 updates the map data 73 using the top plate attitude detection data 116 from the top plate attitude detection sensor 115 in addition to the top plate position detection data 74, etc.

[0099] While a supine imaging table 19 and a standing imaging table 21 are given as examples of structures, and a top plate 20 and a holder 22 are given as examples of components, the invention is not limited to these. A patient support device may also be used as a structure. A patient support device is a support column with a handle for the patient P to hold onto when performing standing imaging. In this case, the component is a handle that rotates between a storage position and a usage position with the support column as the axis of rotation. Alternatively, a ceiling-mounted display may also be used as a structure. A ceiling-mounted display consists of an arm suspended from the ceiling and a display attached to the end of the arm. In this case, the component is a display whose position and orientation can be changed by the arm.

[0100] Furthermore, the map data 73 may be updated only when a component is in a position and / or orientation that could interfere with autonomous driving. The map data 73 may also be updated periodically, i.e., when a predetermined time (e.g., 30 seconds to 1 minute) has elapsed since the last update. For example, a camera may be installed to photograph the entire imaging room RM, and the amount of movement of potential disturbances such as the operator OP, patient P, the top of the supine imaging table 19, and the holder 22 of the standing imaging table 21 may be detected from the image of the camera. Then, the map data 73 may be updated when the detected amount of movement exceeds a predetermined amount.

[0101] The alignment control image 85 may be displayed on the operation panel 29. In this case, the contour OLC of the electronic cassette 12, the center CC of the detection surface 17 of the electronic cassette 12, the irradiation center RC of the radiation R, etc., may be superimposed on the alignment control image 85 as markers.

[0102] The bogie section 16 may be eliminated, and the wheels 15 may be directly attached to the lower part of the main body section 14.

[0103] Camera 32 may be attached to the arm 28 instead of the radiation source 13. Similarly, camera 33 may be provided on the main body 14 instead of the trolley 16. There may be multiple cameras 32 and 33.

[0104] Cameras 32 and 33 may be combined into a single camera. In this case, the camera orientation is set so that it can capture the driving control image 70 when driving control is being performed, and so that it can capture the alignment control image 85 when alignment control is being performed.

[0105] The environmental information sensor is not limited to the example camera 33. LiDAR (Light Detection and Ranging) sensors, TOF (Time-of-Flight) sensors, etc., may also be used. In addition, an IMU (Inertial Measurement Unit) combining an accelerometer and a gyroscope, an ultrasonic sensor, a radar sensor, a magnetic sensor, etc., may also be used.

[0106] Let's consider a case where the environmental information sensor is replaced with a distance sensor 120, such as a LiDAR sensor or a TOF (Time-of-Flight) sensor, instead of the camera 33. In this case, as an example shown in Figure 21, the radiation generator 11 may have a calibration mode that calibrates the distance data 121 output from the distance sensor 120 using the displacement of the components based on the detection data. In calibration mode, the components are moved by a specified amount, and the displacement of the components is measured by the detection sensor and the distance sensor 120.

[0107] Figure 21 illustrates the case where the top plate 20 is lowered by a specified amount of 100 mm. It illustrates the case where the amount the top plate 20 is lowered by 100 mm based on the top plate position detection data 74 output from the top plate position detection sensor 23, and the amount the top plate 20 is lowered by 98 mm based on the distance data 121 output from the distance sensor 120. In this case, the self-position estimation / map data creation unit 62 calibrates the distance data 121 by multiplying it by a calibration coefficient of 100 / 98. The displacement amount of the components based on the detection data is more accurate than the displacement amount of the components based on the distance data 121. Therefore, more accurate distance data 121 can be obtained, and as a result, more accurate self-position estimation and map data 73 can be created. Markers may be provided on the components to facilitate the measurement of the displacement amount of the components by the distance sensor 120.

[0108] The radiation image detection device is not limited to the example electronic cassette 12. A CR (Computed Radiography) cassette may also be used. Furthermore, the subject is not limited to the example patient P. A diseased animal such as a dog or cat may also be used.

[0109] In each of the above embodiments, the processing of each processing unit, such as the image acquisition unit 60, the feature point extraction unit 61, the self-position estimation / map data creation unit 62, the driving control unit 63, the cassette contour extraction unit 80, the radiation source position / attitude derivation unit 81, the alignment control unit 82, the cassette detection unit 95, the patient detection unit 100, and the joint point extraction unit 105, is performed on any computer. Furthermore, any computer may perform these processing using a processor as hardware, a program as software, or a combination thereof. In that case, the processor is configured to work in cooperation with the program to perform the various processing in each of the above embodiments, and can function as each unit or means in each of the above embodiments. Also, the execution order of the processing by the processor is not limited to the order described and may be changed as appropriate. Any computer may be a general-purpose computer, a computer designed for a specific application, a workstation, or any other system capable of performing each processing.

[0110] 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 the example CPU, or programmable logic devices such as an MPU (Micro Processing Unit), FPGA (Field Programmable Gate Array), dedicated circuits for executing specific processes 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 multiple hardware components may reside in physically separate devices or in the same device. Also, in any embodiment, the order of each process performed by the processor is not limited to the order described above and may be changed as appropriate. The hardware is composed of electrical circuits (circuitry) that combine circuit elements such as semiconductor elements.

[0111] 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 devices). The program may be divided and stored on multiple non-temporary computer-readable media located in 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.

[0112] From the above description, the technology described in the following supplementary information can be understood.

[0113] [Additional note 1] A radiation generating device having a main body that emits radiation towards a subject, and a driving mechanism having wheels that enables autonomous movement, Environmental information sensors and, Processor and Equipped with, The aforementioned processor, While creating a map of the surrounding environment from the output data of the aforementioned environmental information sensor, the system estimates its own position from the map data. The detection data is acquired from a detection sensor provided on a structure in the surrounding environment, which detects the position and / or orientation of movable components of the structure. The map data is updated using the detection data. Radiation generating device. [Additional note 2] The aforementioned processor, A radiation generating device according to Appendix 1, which estimates the amount of travel of the travel mechanism based on the amount of change in the position of the same structure included in past output data and current output data. [Additional note 3] The aforementioned processor, Based on the detection data, the moving component is recognized. A radiation generating device according to Appendix 2, which estimates the amount of travel by excluding the components that are recognized as being in motion. [Additional note 4] The aforementioned map data is 3D data, and the radiation generating device is as described in any one of the appendices 1 to 3. [Additional note 5] The aforementioned map data is a radiation generating device as described in Appendix 4, in which the spatial resolution in the height direction is set lower than that in the horizontal direction. [Additional note 6] The aforementioned processor, A radiation generating device according to any one of the appendix items 1 to 5, which changes the travel speed of the travel mechanism according to the distance to the aforementioned structure. [Additional note 7] The aforementioned processor, The radiation generating device described in Appendix 6, wherein the closer the distance to the aforementioned structure, the greater the degree of reduction in the travel speed. [Additional note 8] The environmental information sensor is a distance sensor, and the output data includes distance data. The aforementioned processor, A radiation generating device according to any one of the appendix items 1 to 7, wherein the distance data is calibrated using the displacement amount of the component based on the detection data. [Additional note 9] The radiation generating device according to any one of Appendix 1 to Appendix 8, wherein the structure includes a supine imaging table, and the component is the top plate of the supine imaging table. [Additional Note 10] The radiation generating device according to any one of the appendices 1 to 9, wherein the structure includes a standing imaging table, and the component is a holder for the standing imaging table.

[0114] The technology of this disclosure can be appropriately combined with the various embodiments and / or variations described above. Furthermore, it is understood that various configurations can be adopted without departing from the spirit of the invention, and the invention is not limited to the embodiments described above. In addition, the technology of this disclosure extends not only to programs, but also to storage media for non-temporarily storing programs, and to computer program products containing programs.

[0115] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0116] In this specification, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0117] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference. [Explanation of symbols]

[0118] 10. Radiography System 11. Radiation Generating Devices 12 Electronic Cassettes 13 Radiation source 14 Main body 15 wheels 16 Bogie section 17 Detection surface 18. Radiological images 19 Recumbent photography stand 20 top plate 21 Standing position X-ray table 22 holder 23 Top panel position detection sensor 24 Posts 25. Holder position detection sensor 26 Front 27 Rear 28 Arms 29. Control Panel 30 Radiation tubes 31 Irradiation field limiter 32, 33 Camera 40 processors 41 Storage 42 Operating Program 43 Control data 50 Communication I / F 51. Drive actuator 52 Alignment Actuator 53. Source position detection sensor 60 Image acquisition unit 61 Feature Point Extraction Unit 62 Self-Position Estimation and Map Data Creation Section 63 Driving control unit 70 Images for driving control 71 Feature point extraction results 71CF Feature point extraction results for the current frame 71P Feature point extraction results (past data) 71PF Feature point extraction results from the previous frame 72 Estimation results 73 Map data 73P Map data (past data) 74 Top panel position detection data 75 Holder position detection data 80 Cassette contour extraction section 81 Source position / attitude derivation part 82 Alignment Control Unit 85 Image for alignment control 85L Image for learning and alignment control 86 Cassette Contour Extraction Model 87 Cassette contour extraction results 87CA Correct Answer Data 87L Training Cassette Contour Extraction Results 88 Derivation results 90 training data 95 Cassette detection unit 100 Patient detection unit 105 Articular point extraction area 106 Joint Point Extraction Model 107 Joint point extraction results 110 table 115 Tabletop orientation detection sensor 116 Top plate orientation detection data 120 Distance Sensor 121 Distance Data The center of the detection surface of the CC electronic cassette. D Distance from the structure DTH1, DTH2: First distance threshold, Second distance threshold Frame F FP characteristic points HP standby position Center of the IC imaging area J joint point Outline of an OLC electronic cassette OP Operator PS Self-position R radiation RC radiation irradiation center RM Photo Studio ST100, ST110, ST120, ST130, ST140, ST150, ST160 Step TP target position TP1, TP2 1st target position, 2nd target position V1, V2, V3 Traveling speed

Claims

1. A radiation generating device having a main body that emits radiation towards a subject, and a driving mechanism having wheels that enables autonomous movement, Environmental information sensors and, Processor and Equipped with, The aforementioned processor, While creating a map of the surrounding environment from the output data of the aforementioned environmental information sensor, the system estimates its own position from the map data. The detection data is acquired from a detection sensor that is provided on a structure in the surrounding environment and detects the position and / or orientation of movable components of the structure. The map data is updated using the detection data. Radiation generating device.

2. The aforementioned processor, The radiation generating device according to claim 1, which estimates the amount of travel of the travel mechanism based on the amount of change in the position of the same structure included in past output data and current output data.

3. The aforementioned processor, Based on the detection data, the moving component is recognized. The radiation generating device according to claim 2, which estimates the amount of travel by excluding the components that are recognized as being in motion.

4. The radiation generating device according to claim 1, wherein the map data is three-dimensional data.

5. The radiation generating device according to claim 4, wherein the spatial resolution of the map data is set to be lower in the height direction than in the horizontal direction.

6. The aforementioned processor, The radiation generating device according to claim 1, wherein the travel speed of the travel mechanism is changed according to the distance to the structure.

7. The aforementioned processor, The radiation generating device according to claim 6, wherein the degree of reduction in the travel speed increases as the distance to the structure decreases.

8. The environmental information sensor is a distance sensor, and the output data includes distance data. The aforementioned processor, The radiation generating device according to claim 1, wherein the distance data is calibrated using the amount of displacement of the component based on the detection data.

9. The radiation generating device according to claim 1, wherein the structure includes a supine imaging table, and the component is the top plate of the supine imaging table.

10. The radiation generating device according to claim 1, wherein the structure includes a standing imaging table, and the component is a holder for the standing imaging table.

11. A method for operating a radiation generating device in which a main body having a radiation source that emits radiation toward a subject is capable of autonomous movement by a driving mechanism having wheels, To create a map of the surrounding environment while estimating one's own position from the output data of an environmental information sensor. To acquire detection data from a detection sensor provided on a structure in the surrounding environment, which detects the position and / or orientation of movable components of the structure, and, Updating the map data using the aforementioned detection data, A method for operating a radiation generating device, including the device itself.

12. An operating program for a radiation generator in which a main body having a radiation source that emits radiation toward a subject is capable of autonomous movement by a driving mechanism having wheels, To create a map of the surrounding environment while estimating one's own position from the output data of an environmental information sensor. To acquire detection data from a detection sensor provided on a structure in the surrounding environment, which detects the position and / or orientation of movable components of the structure, and, Updating the map data using the aforementioned detection data, An operating program for a radiation generator that causes a computer to perform a process including [specific details].

Images