Medical radiation detection apparatus and radiation detection posture ai recognition and training method

Abstract

Provided is a radiation detection object posture AI recognition training and adjustment method, comprising: placing a detection object on a detection bed; setting a distance between the detection bed and a radiation source or an imager; photographing the detection object, and marking an obtained posture picture to give a marked posture picture; and performing AI training by using the marked posture picture to give a radiation detection object posture AI recognition feature data set. The posture picture is recognized using an AI recognition algorithm, so as to give information about whether the posture of the detection object is correct. Also provided is a medical radiation detection apparatus, comprising a support frame A, a rotating arm, a control module, a rotating apparatus, a detection object position detection apparatus, and a detection position adjustment apparatus. The detection object position detection apparatus is configured for detecting the position of the detection object. The detection position adjustment apparatus adjusts the height of the detection object according to the position of the detection object, or adjusts the position of the radiation source or the imager.

Description

Medical radiation detection device, AI recognition and training method for radiation detection posture Technical Field

[0001] This application belongs to the field of medical instrument technology, and specifically relates to X-ray-based medical imaging detection devices, AI recognition training method for radiation detection object posture, and AI recognition method for radiation detection posture. Background Technology

[0002] Explanation of the name:

[0003] AI: Artificial Intelligence, abbreviated as AI, is a new technical science that studies and develops theories, methods, technologies, and application systems to simulate, extend, and expand human intelligence. Image recognition technology is an important field of artificial intelligence. It refers to the technology of recognizing objects in images to identify targets and objects of various patterns.

[0004] CT: Computed Tomography is a medical imaging technique. It uses X-ray beams to perform tomographic scanning of the human body and uses computer processing to produce detailed images of the body's internal structures. CT scans can also be considered as 3D imaging of the object being examined.

[0005] DR: The full name is Digital Radiography. In this application, DR refers to medical DR, which is a relatively advanced examination method in radiology and can be used for the diagnosis, differential diagnosis, and prognostic assessment of diseases affecting multiple organ systems. DR examination can also be considered as performing 2D imaging of the object being examined.

[0006] DD motor: also called a direct drive motor. It features high torque at low speeds, simple structure, low mechanical loss, low noise, and low maintenance.

[0007] In hospitals, different detection equipment is used in different scenarios; for example, DR equipment is used to take pictures of the chest, and high-resolution CT equipment is used to take pictures of blood vessels in the head.

[0008] In the field of pet healthcare, CT or DR equipment is expensive and takes up a lot of space. Many pet hospitals do not have the physical space or funds to install both CT and DR equipment at the same time.

[0009] With the increasing prevalence of community hospitals, many of them also hope to install CT and DR equipment, but they also face issues such as installation location limitations and costs.

[0010] If CT and DR equipment are integrated, how to reduce the size and cost of the integrated equipment is one of the technical problems that this application needs to solve.

[0011] When performing radiation detection using medical radiation detection devices, the target object needs to be in different postures and positions depending on the medical testing purpose. This places special requirements on the operators of the detection equipment, resulting in high operating costs.

[0012] In CT or DR examinations, different doses of radiation are applied to objects of varying thicknesses to precisely control image quality. Excessive radiation can affect the health of the object being examined and cause overexposure, which is detrimental to image interpretation. Insufficient radiation can result in a unclear image. Therefore, it is necessary to accurately measure the thickness of the object being examined.

[0013] Confirming the correct posture of the object under test and accurately measuring the thickness of the test area are technical challenges in the field of radiotherapy, and are also among the technical problems that this application aims to solve. Summary of the Invention

[0014] In this application, the inventors propose a medical radiological detection device with a transmitter and an imager respectively mounted at both ends of a rotating arm. The rotation space of the rotating arm is open, and the transmitter and imager rotate around a detection bed. The detection bed can be moved up and down to adjust the height of the object being detected. It can perform both 2D and 3D imaging of the object, is suitable for various scenarios, is very convenient to use, and provides a better user experience. The medical radiological detection device can detect the height of the object, obtain its posture and position information, and adjust the detection position based on the posture and position information to obtain high-quality images.

[0015] In this application, the inventors propose an AI recognition training method and an AI recognition method for the posture of a radiation detection object. Through AI training, an AI recognition feature dataset of the posture of a radiation detection object is obtained, thereby enabling the machine to recognize the posture of the radiation detection object and automatically adjust the radiation detection position. The position recognition and position adjustment during the detection process are automatically recognized and completed by the machine, which improves the detection efficiency and can quickly find the geometric center of different objects to be tested as the central axis of the imaging, ensuring the quality of the imaging.

[0016] The solution to the above-mentioned technical problems provided in this application is a medical radiological detection device comprising: a support frame A, a rotating arm, a control module, a rotating device, a detection object position detection device, and a detection position adjustment device; the support frame A is used to support other modules or devices; one end of the rotating arm is mechanically connected to the rotating device; the other end of the rotating device is mechanically connected to the support frame A; the rotating arm can rotate relative to the support frame A under the drive of the rotating device; the control module is disposed on the support frame A or the rotating arm, and the control module is electrically connected to the rotating device; the control module is electrically connected to the emission source; the control module is electrically connected to the imager; the control module is electrically connected to the detection object position detection device; and the control module is electrically connected to the detection position adjustment device; one end of the rotating arm includes an emission source, and during the rotation of the rotating arm, the emission source rotates around the detection object, and the emission source directs the radiation beam towards... The aforementioned detection object; the other end of the rotating arm includes an imager. During the rotation of the rotating arm, the imager rotates around the detection object and can detect the radiation signal emitted by the aforementioned emission source; the control module is used to control the rotating arm to rotate to a set angle, control the emission source to emit radiation signals, control the imager to detect radiation signals, and perform 2D imaging of the detection object, the aforementioned 2D imaging being used for medical diagnosis; the control module is used to control the rotation of the rotating arm, control the emission source to emit radiation signals, control the imager to detect radiation signals, and perform 3D imaging of the detection object during the rotation, the aforementioned 3D imaging being used for medical diagnosis; the detection object position detection device is used to detect the position of the detection object; including any one of the following features: TC1: the detection position adjustment device adjusts the height of the detection object according to the position of the detection object; TC2: the detection position adjustment device adjusts the position of the radiation source or imager according to the position of the detection object.

[0017] The aforementioned medical radiation detection device also includes a detection bed, which is used to support the object being detected; the detection bed can move up and down to adjust the height of the object being detected.

[0018] The aforementioned medical radiological detection device includes any one or more of the following features: Feature TA1: The object position detection device includes a weighing module, which includes pressure sensors placed on the detection bed; Feature TA2: The object position detection device includes a weighing module, which includes at least three pressure sensors supporting the detection bed surface; Feature TA3: The object position detection device includes a dual-camera ranging device, which is installed at one end of the rotating arm; Feature TA4: The object position detection device includes a laser ranging device, which is installed at one end of the rotating arm; Feature TA5: The object position detection device includes an ultrasonic ranging device, which is installed at one end of the rotating arm; Feature TA6: The object position detection device includes a camera device, the detection position of the object covers a standard reference object or a standard reference image; the object position detection device calculates the height information of the object based on the image obtained by the camera device and the number of pixels occupied by the standard image. The detection device includes a camera device, which calculates the object's posture information using AI based on the images obtained by the camera device, and provides information on whether the posture is correct. The detection device also includes a dual-camera ranging device, which obtains the object's height information. The detection position adjustment device includes a lifting column, which connects the detection bed to the lifting column, allowing the detection bed to move up and down. The rotating device includes a rotating motor and a rotating shaft, which drive the rotating arm to rotate. The rotating shaft's bushing and shaft are connected to the support frame A or the rotating arm, respectively. The rotating device also includes a DD motor, with one end connected to the support frame A and the other end connected to the rotating arm. The part of the detection bed that receives radiation is made of carbon fiber.

[0019] The radiation source is a cone-beam X-ray source, and the imager is an amorphous silicon flat panel imager; it includes any one or more of the following features: Feature TB1: The X-ray source includes a kV X-ray source; Feature TB2: The X-ray source emits X-rays with an energy of 100kV; Feature TB3: The X-ray source includes a linear accelerator; Feature TB4: The detection pixel of the amorphous silicon flat panel imager is greater than 3072×3072; Feature TB5: The size of the amorphous silicon flat panel imager is greater than 43cm×43cm.

[0020] The aforementioned medical radiological detection device also includes a support frame B and a rotating connection device. The support frame B is mechanically coupled to the support frame A through the rotating connection device. The support frame B supports the aforementioned detection bed and can drive the detection bed to rotate around the rotating connection device.

[0021] The aforementioned medical radiological detection device also includes a support frame C, the bottom of which is connected to at least three casters. The support frame C supports the aforementioned detection bed and can drive the detection bed to rotate. At least one caster has a locking mechanism.

[0022] The control module can control the rotating arm to rotate to a set angle, control the radiation source to emit radiation signals, control the imager to detect radiation signal values, and perform dynamic 2D imaging of the object being detected.

[0023] The solution proposed in this application to address the aforementioned technical problems is a method for training AI-based posture recognition of radiation detection objects. This method includes placing the detection object on a detection bed; setting the distance between the detection bed and the radiation source or imager; taking a picture of the detection object to obtain a posture image; labeling the posture image to obtain a posture label image; and using the posture label image for AI training to obtain a dataset of AI-based posture recognition features for radiation detection objects.

[0024] The above-mentioned AI training method for radiation detection object posture recognition includes any one or more of the following features: Feature TD1: The detection object is photographed using a dual-camera ranging device; a posture image is obtained; the obtained posture image includes the depth information of the detection object; Feature TD2: The detection object is photographed using a camera; a posture image is obtained; the detection position of the detection object covers a standard reference object or a standard reference image; Feature TD3: The detection object is any one or more of cats, dogs, and humans of different sizes; Feature TD4: Different distances are set between the detection bed and the radiation source or imager; the detection object is photographed to obtain corresponding posture images at different distances; Feature TD5: The above-mentioned radiation detection object posture AI recognition feature dataset is obtained by training at least 100 sets of different posture identification images; Feature TD6: The posture image is used to identify whether the detection posture of the detection object is correct or incorrect, and a detection posture identification image is obtained. The detection posture identification image is used for AI training to obtain a radiation detection object posture AI recognition feature dataset that includes detection posture recognition information.

[0025] The solution to the above-mentioned technical problem in this application can also be a radiation detection posture AI recognition method, which involves placing the object to be detected on a detection bed; taking a picture of the object to obtain a posture image; and using an AI recognition algorithm to identify the posture image to obtain information on whether the posture of the object to be detected is correct.

[0026] The aforementioned AI-based posture recognition method for radiation detection also includes an AI recognition algorithm that provides a posture adjustment suggestion diagram.

[0027] The above-mentioned AI-based posture recognition method for radiation detection uses the aforementioned AI-based posture recognition feature dataset for radiation detection objects.

[0028] The solution to the above-mentioned technical problems in this application is an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method described above.

[0029] The solution to the above-mentioned technical problem in this application can also be a readable storage medium storing a computer program that, when executed by a processor, implements the method described above; and storing data thereon, which is the above-mentioned AI recognition feature dataset of the pose of the radiation detection object.

[0030] The solution to the above-mentioned technical problem is a data storage device, which includes the above-mentioned radiation detection object posture AI recognition feature dataset, or includes program code that runs the above-mentioned method.

[0031] The technical advantages of the above-mentioned technical solution include: the rotation space of the rotating arm is open, the structure is simple, the cost is lower, and the user experience is better.

[0032] The technical advantages of the above-mentioned technical solution include: it can perform both 2D and 3D imaging, is applicable to a variety of scenarios, is very convenient to use, and provides a better user experience.

[0033] The technical effects of the above-mentioned technical solution include: it can also perform 2D dynamic imaging, further expanding the application scenarios, making it very convenient to use and providing a better user experience.

[0034] The technical effects of the above-mentioned technical solution include: the detection bed can move up and down to adjust the height of the detection object, making the imaging accuracy of the equipment better and enabling the adjustment of the imaging center position.

[0035] The technical effects of the above-mentioned technical solution include: the setting of the weighing module facilitates the acquisition of the weight of the object to be tested, and the displacement of the testing bed can be determined based on the weight.

[0036] The technical effects of the above-mentioned technical solution include: the weighing module includes at least 3 pressure sensors, and can also obtain the center of gravity position of the object being tested, enabling more accurate displacement of the testing bed.

[0037] The technical effects of the above-mentioned technical solution include: the detection object position detection device can detect the position of the detection object, can more accurately perform the displacement of the detection bed, and can more accurately control the positional relationship between the imaging center position of the detection object and the imaging central axis of the rotating arm.

[0038] The technical effects of the above-mentioned technical solution include: the dual-camera ranging device can accurately acquire relative position information and accurately detect bed displacement.

[0039] The technical effects of the above-mentioned technical solution include: the laser ranging device can obtain relative position information more accurately and can detect bed displacement more accurately.

[0040] The technical advantages of the above-mentioned technical solution include: multiple ranging methods, convenient use and configuration, and the ability to set different ranging devices for different occasions, making it more flexible and diverse.

[0041] The technical effects of the above-mentioned technical solution include: obtaining the height information and center position information of the detected object through AI algorithm calculation, improving the efficiency of information acquisition and accelerating position adjustment.

[0042] The technical effects of the above-mentioned technical solution include: the lifting column can stably and reliably drive the testing bed to move up and down.

[0043] The technical effects of the above-mentioned technical solutions include: the cone-beam X-ray source and the amorphous silicon flat panel imager ensure image quality for both 2D and 3D imaging.

[0044] The technical effects of the above-mentioned technical solution include: the setting of the support frame B facilitates the rotation of the detection bed, changes the position of the bed, and allows for convenient and flexible adjustment of the position of the detection object relative to the imaging optical path.

[0045] The technical benefits of the above-mentioned technical solution include: the casters facilitate movement.

[0046] The technical effects of the above solution include: using pose-labeled images for AI training to obtain a dataset of AI-recognized features for the pose of radiation detection objects; accumulating this dataset for easy updates and iterations; and classifying different detection targets to obtain corresponding AI-recognized feature datasets, which can further improve the accuracy of AI-recognized object location.

[0047] The technical effects of the above-mentioned technical solution include: the detection object can be any one or more of cats, dogs, and humans of different sizes, and it can automatically match the corresponding feature data set according to different detection objects, thus expanding the application scope of the device and making it suitable for a variety of application scenarios.

[0048] The technical effects of the above-mentioned technical solution include: the detection of the object's posture, whether correct or incorrect, can be obtained through AI recognition, and prompts can be provided when necessary, thereby improving the reliability of the detection.

[0049] The technical effects of the above-mentioned technical solution include: using AI recognition algorithms to identify the above-mentioned posture images and obtain the posture information of the object to be detected; adjusting the relative position of the detection bed and the radiation source or imager according to the position information, automatically identifying and completing the position adjustment, which greatly improves the detection efficiency. Attached Figure Description

[0050] Figure 1 is a flowchart illustrating the training method for AI-based pose recognition of radiation detection objects;

[0051] Figure 2 is a flowchart illustrating the AI-based posture recognition method for radiation detection.

[0052] Figure 3 is a side view of a medical radiation detection device;

[0053] Figure 4 is a side view of a medical radiation detection device;

[0054] Figure 5 is a side view of a medical radiation detection device;

[0055] Figure 6 is a three-dimensional schematic diagram of a medical radiation detection device;

[0056] Figure 7 is a side view of a medical radiation detection device in one state;

[0057] Figure 8 is a side view of the medical radiation detection device in another state;

[0058] Figure 9 is a three-dimensional schematic diagram of a medical radiation detection device;

[0059] Figure 10 is a three-dimensional schematic diagram of a medical radiation detection device;

[0060] Figure 11 is a three-dimensional schematic diagram of a medical radiation detection device;

[0061] Figure 12 is a three-dimensional schematic diagram of a medical radiation detection device;

[0062] Figure 13 is a three-dimensional schematic diagram of a medical radiation detection device. Detailed Implementation

[0063] The contents of this application will be further described in detail below with reference to the accompanying drawings. It should be noted that the following description is of preferred embodiments of the present invention and does not constitute any limitation on the present invention. The description of the preferred embodiments of the present invention is merely an explanation of the general principles of the invention. The designations "first," "second," "A," and "B" used in this invention are for ease of explanation only and do not represent a temporal or spatial order. The combinations of letters and numbers "TA," "TB," and "H" used in this invention are for ease of explanation only, and their specific meanings are determined by the specific terms they represent.

[0064] As shown in Figure 1, an embodiment of a radiation detection object posture AI recognition training method includes the following steps: placing the detection object on a detection bed; setting the distance between the detection bed and the radiation source or imager; taking a picture of the detection object to obtain a posture image; labeling the posture image to obtain a posture label image; and using the posture label image for AI training to obtain a radiation detection object posture AI recognition feature dataset.

[0065] In some embodiments of the AI-based posture recognition training method for radiation detection objects, a dual-camera ranging device is used to capture images of the detection object; a posture image is obtained; the obtained posture image includes the depth information of the detection object.

[0066] In some embodiments of the AI-based posture recognition training method for radiation detection objects, a camera is used to photograph the detection object; a posture image is obtained; and the detection position of the detection object covers a standard reference object or a standard reference image.

[0067] In some embodiments of AI-based posture recognition training methods for radiation detection objects, the detection objects are any one or more of cats, dogs, and humans of different sizes.

[0068] In some embodiments of AI-based posture recognition training methods for radiation detection objects, different distances are set between the detection bed and the radiation source or imager; images of the detection object are taken to obtain corresponding posture images at different distances.

[0069] In some embodiments of the radiation detection object pose AI recognition training method, the aforementioned radiation detection object pose AI recognition feature dataset is obtained by training with at least 100 sets of different pose identification images.

[0070] In some embodiments of the radiation detection object pose AI recognition training method, the pose image is labeled with information indicating whether the detection pose of the object is correct or incorrect, and a detection pose label image is obtained. The detection pose label image is then used for AI training to obtain a radiation detection object pose AI recognition feature dataset that includes the detection pose recognition information.

[0071] As shown in Figure 2, in an embodiment of a radiation detection posture AI recognition method, the object to be detected is placed on a detection bed; an initial distance is set between the detection bed and the radiation source or imager; the object to be detected is photographed to obtain a posture image; the posture image is recognized using an AI recognition algorithm to obtain the posture information of the object to be detected; and the relative position of the detection bed and the radiation source or imager is adjusted according to the position information. The AI ​​recognition algorithm uses the aforementioned radiation detection object posture AI recognition feature dataset.

[0072] AI recognition algorithms are used to identify the pose of the object being detected, and to obtain information on whether the pose of the object being detected is correct.

[0073] A computing processing device includes program code that runs the above-described AI recognition training method for radiation detection object posture or AI recognition method for radiation detection posture.

[0074] A data storage device includes the aforementioned AI-based feature dataset for recognizing the posture of a radiation detection object, or a memory includes program code that runs the aforementioned AI-based training method for recognizing the posture of a radiation detection object or the AI-based posture recognition method for radiation detection.

[0075] As shown in Figures 3 to 8, an embodiment of a medical radiological detection device includes: a support frame A, a rotating arm, a control module, a rotation device, a detection object position detection device, and a detection position adjustment device; one end of the rotating arm is mechanically connected to the rotation device; the other end of the rotation device is mechanically connected to the support frame A; the rotating arm can rotate relative to the support frame A under the drive of the rotation device; one end of the rotating arm includes a radiation source, which rotates around the detection bed during the rotation of the rotating arm, and directs the radiation beam towards the detection bed; the other end of the rotating arm includes an imager, which rotates around the detection bed during the rotation of the rotating arm, and can detect the radiation signal emitted by the radiation source; the control module is used to control the rotating arm to rotate to a set angle, control the radiation source to emit radiation signals, control the imager to detect radiation signals, and perform 2D imaging of the detection object, the 2D imaging being used for medical diagnosis; the control module is also used to control the rotation of the rotating arm, control the radiation source to emit radiation signals, control the imager to detect radiation signals, and perform 3D imaging of the detection object during the rotation, the 3D imaging being used for medical diagnosis.

[0076] In other embodiments, the control module can control the rotating arm to rotate to a set angle, control the radiation source to emit radiation signals, control the imager to detect radiation signal values, and perform dynamic 2D imaging of the object being detected.

[0077] As shown in Figures 3 to 8, in an embodiment of a medical radiation detection device, a detection object position detection device is used to detect the position of the detection object; a detection position adjustment device adjusts the height of the detection object according to its position.

[0078] In other embodiments, the detection position adjustment device adjusts the position of the radiation source or imager according to the position of the object being detected.

[0079] As shown in Figures 3 to 8, in an embodiment of the medical radiological detection device, a detection bed is also included. The detection bed is used to support the object to be detected. The detection bed can be moved up and down to adjust the height of the object to be detected.

[0080] As shown in Figures 3 to 8, in an embodiment of the medical radiological detection device, the radiation source is a cone-beam X-ray source, and the imager is an amorphous silicon flat panel imager. The X-ray source includes a kV X-ray source; the X-ray source emits X-rays with an energy of 100kV; the X-ray source includes a linear accelerator; the detection pixel size of the amorphous silicon flat panel imager is greater than 3072×3072; and the size of the amorphous silicon flat panel imager is greater than 43cm×43cm. The control module receives the image from the amorphous silicon flat panel imager and generates an X-ray tomographic image of the target body based on the received image. The control module also receives the 2D projection image from the amorphous silicon flat panel imager and generates a 3D X-ray tomographic image of the target body based on the received 2D projection image.

[0081] As shown in Figures 3 and 7, in an embodiment of the medical radiological detection device, the detection position adjustment device includes a lifting support column, the detection bed is connected to the lifting support column, and the lifting support column can drive the detection bed to move up and down.

[0082] As shown in Figures 3, 6, and 7, in an embodiment of the medical radiological detection device, a support frame B and a rotating connection device are also included. The support frame B is mechanically coupled to the support frame A through the rotating connection device. The support frame B supports the aforementioned detection bed and can drive the detection bed to rotate around the rotating connection device. The support frame B supports the detection bed to rotate to be parallel to the rotating device; the support frame B supports the detection bed to rotate to be perpendicular to the rotating device; and the support frame B supports the detection bed to rotate to an angle intersecting with the rotating device.

[0083] As shown in Figures 3 to 7, in embodiments of the medical radiological detection device, a support frame C is further included. The bottom of the support frame C is connected to at least three casters. The support frame C supports the aforementioned detection bed and can drive the detection bed to rotate. The casters are omnidirectional support wheels, serving both support and steering functions.

[0084] In some embodiments of medical radiation detection devices, the object location detection device includes a camera device, and the object detection location covers a standard reference object or standard reference image; the object location detection device calculates the height information and center position information of the object through AI algorithms based on the image obtained by the camera device; with the center position information, the thickness information of the object is obtained, and the radiation intensity value is determined according to different parts, such as more bone in the head and more soft tissue in the abdomen.

[0085] In some embodiments of medical radiological detection devices, the object location detection device includes a dual-camera ranging device; the object location detection device calculates the height information and center position information of the object to be detected using an AI algorithm based on the images obtained by the dual-camera ranging device.

[0086] The images obtained by the dual cameras contain depth information. After obtaining the distance relative to the cameras, we can know the height of the object being detected minus the initial height of the bed to obtain the thickness information of the object. The initial height of the bed plus half the thickness of the object being detected gives us the center position of the object. By changing the height of the bed, we can align the center position with the center position required for radiation, thus improving the quality of the radiation image.

[0087] In veterinary hospitals, there are all sorts of animals that need to be examined. It is very difficult for humans to judge their height information, but with the training of AI, it is possible to obtain the height information of various animals in various postures with relatively high accuracy.

[0088] In some embodiments of medical radiation detection devices, a dual-camera ranging device is included; the dual-camera ranging device is installed at one end of the aforementioned rotating arm; the dual-camera ranging device is used to obtain the position of the object being detected relative to the emission source or imager.

[0089] A dual-camera ranging device can also directly obtain the distance.

[0090] In some embodiments of medical radiation detection devices, a laser rangefinder is included; the laser rangefinder is mounted at one end of the aforementioned rotating arm and is used to obtain the position of the object being detected relative to the emission source or imager.

[0091] In some embodiments of medical radiation detection devices, an ultrasonic ranging device is included; the ultrasonic ranging device is installed at one end of the aforementioned rotating arm to obtain the position of the object being detected relative to the emission source or imager.

[0092] As shown in Figure 5, in an embodiment of the medical radiological detection device, a weighing module is included, which includes a pressure sensor placed on the detection bed; the weighing module includes at least three pressure sensors, which support the detection bed surface.

[0093] Animals or humans have certain fixed shapes; for example, the head has a certain ellipticity limit. Knowing the weight, the approximate height range of the object being tested can be determined. By measuring the mass, the center point of the object can be determined, which is the thickness of the tested area. Based on the thickness, the radiation dose value can be determined. The bed height can be adjusted based on the center point location; or the relative position of the emission source and the imager can be adjusted.

[0094] Increasing the number of weighing modules allows for the separate measurement of the mass of multiple segments, enabling more effective determination of the center point location of different segments.

[0095] A method for training and adjusting the posture of a radiation detection object using AI involves placing the object on a detection bed; setting the distance between the detection bed and the radiation source or imager; capturing images of the object to obtain posture pictures; labeling these posture pictures to obtain posture-labeled images for AI training, resulting in a dataset of AI-recognized posture features for radiation detection objects. The method further involves setting the initial distance between the detection bed and the radiation source or imager; capturing images of the object to obtain posture pictures; using an AI recognition algorithm to identify the posture pictures and obtain the object's posture information; and adjusting the relative position of the detection bed and the radiation source or imager based on the position information. The medical radiation detection device includes a support frame A, a rotating arm, a control module, a rotation device, an object position detection device, and an object position adjustment device. The object position adjustment device adjusts the height of the object or the position of the radiation source or imager based on the object's position.

[0096] As shown in Figures 9 to 13, in a preferred embodiment, the support frame C includes a support portion C1 and a test bed sliding portion C2; the support portion C1 and the test bed sliding portion C2 are connected by a slide rail; the test bed sliding portion C2 can slide relative to the support portion C1.

[0097] The testing bed C3 is mechanically connected to the sliding part of the testing bed C2, and the lifting support column C4 can drive the testing bed C3 to move up and down.

[0098] The sliding part C2 of the testing bed can slide relative to the support part C1, which makes it easier to detect a wider area during the testing process.

[0099] As shown in Figure 9, the sliding part C2 of the testing bed slides to the end of one end, which is state 1. As shown in Figure 10, the sliding part C2 of the testing bed slides to the end of the other end, exposing the slide rail, which is state 2.

[0100] As shown in Figure 10, a remote control is also included, which can control the sliding part C2 of the testing bed to slide relative to the support part C1 in terms of direction and distance.

[0101] Using a remote control, doctors can easily control the sliding part C2 of the examination bed to drive the relative rotation of the examination bed C3 to the rotating arm, making it convenient for doctors to control the irradiation of different parts.

[0102] As shown in Figure 9, it also includes an emergency braking button, which is electrically connected to the control module and can control the rotation state of the rotating arm.

[0103] The rotating arm has a large momentum during rotation, which poses a certain danger. The rotating arm can be stopped in an emergency by pressing the emergency brake button.

[0104] The emergency brake button can be set to foot control; pressing the button will rotate the arm, and releasing it will stop the rotation, making it convenient for doctors to operate.

[0105] The emergency stop button can be used in conjunction with the remote control. For example, the remote control can be used to move the C3 testing bed, while the emergency stop button can be used to turn the radiation source on and off. This allows for flexible control depending on the situation.

[0106] Figure 11 shows four distance sensors connected to the control module via electrical signals. The distance sensors are installed at both ends of the rotating arm to detect the distance to objects near the arm. When the detected distance to an object is less than a set value, the rotating arm is prevented from rotating. The figure shows distance sensors K1 and K2 on one side.

[0107] Figure 12 shows the installation positions of distance sensor K3 and distance sensor K4 on the other side.

[0108] Distance sensors can take many forms, from simple electromagnetic induction sensors to complex laser rangefinders.

[0109] As shown in Figures 11 and 12, two distance sensors are located on both sides of the transmitter, such as distance sensor K3 and distance sensor K1.

[0110] As shown in Figures 11 and 12, two distance sensors are located on both sides of the imager, such as distance sensor K2 and distance sensor K4.

[0111] As shown in Figures 11 and 12, the device includes two positioning cursor transmitters. One transmitter emits a positioning cursor M1, and the other emits a positioning cursor M2. The positioning cursor transmitters are installed near the emission source and emit positioning patterns, which are used to illuminate the detection area.

[0112] As shown in Figure 9, the system includes a display screen, which is electrically connected to the control module. The display screen is mechanically connected to the support frame A.

[0113] The monitor can also be installed at one end of the testing bed. Considering the operation of getting on and off the bed, the monitor is preferably placed on the support frame A.

[0114] As shown in Figure 13, the detection bed C3 is in a separated state. The detection bed C3 is not mechanically fixed to the support frame A1 and can be separated, making it convenient to change to different positions for detection, such as turning the detection bed C3 to a direction perpendicular to the rotating arm.

[0115] The bottom of the testing bed C3 has rollers with a locking device. After changing position, the rollers are locked, which can provide stable support during the testing process.

[0116] While the present invention has been described and illustrated with reference to preferred embodiments and several alternatives, the invention is not limited to the specific descriptions herein. Other alternatives or equivalent components may also be used to practice the invention.

Claims

1. A medical radiation detection device, characterized in that, include: Support frame A, rotating arm, control module, rotating device, object position detection device, and detection position adjustment device; The rotating arm is mechanically connected to one end of the rotating device; the other end of the rotating device is mechanically connected to the support frame A; the rotating arm can rotate relative to the support frame A under the drive of the rotating device. The control module is mounted on the support frame A or the rotating arm. The control module is electrically connected to the rotating device, the transmitter, and the imager. The control module is electrically connected to the object position detection device; the control module is also electrically connected to the detection position adjustment device. One end of the rotating arm includes a radiation source. During the rotation of the rotating arm, the radiation source rotates around the object being detected and directs the radiation beam toward the object being detected. The other end of the rotating arm includes an imager. During the rotation of the rotating arm, the imager rotates around the object being detected and can detect the radiation signal emitted by the radiation source. The control module is used to control the rotating arm to rotate to a set angle, control the emission source to emit radiation signals, control the imager to detect radiation signals, and perform 2D imaging of the object to be detected. The 2D imaging is used for medical diagnosis. The control module is also used to control the rotating arm to rotate, control the emission source to emit radiation signals, control the imager to detect radiation signals, and perform 3D imaging of the object to be detected during the rotation. The 3D imaging is used for medical diagnosis. The object position detection device is used to detect the position of the object being detected; it includes any one of the following features: TC1: The detection position adjustment device adjusts the height of the detection object according to its position; TC2: The detection position adjustment device adjusts the position of the radiation source or imager according to the position of the object being detected.

2. The medical radiation detection device as described in claim 1, characterized in that, It also includes a testing bed, which is used to support the object being tested; the testing bed can move up and down to adjust the height of the object being tested.

3. The medical radiation detection device as described in claim 2, characterized in that, Includes any one or more of the following features, Feature TA1: The object position detection device includes a weighing module, and the weighing module includes a pressure sensor placed on the detection bed; Feature TA2: The object position detection device includes a weighing module, which includes at least three pressure sensors that support the detection bed surface; Feature TA3: The object position detection device includes a dual-camera ranging device; the dual-camera ranging device is installed at one end of the rotating arm; Feature TA4: The object position detection device includes a laser rangefinder; the laser rangefinder is installed at one end of the rotating arm; Feature TA5: The object position detection device includes an ultrasonic ranging device; the ultrasonic ranging device is installed at one end of the rotating arm. Feature TA6: The object position detection device includes a camera device, and the detection position of the object covers a standard reference object or a standard reference image; the object position detection device calculates the height information and center position information of the object based on the image obtained by the camera device and the number of pixels occupied by the standard image. Feature TA7: The object position detection device includes a camera device. Based on the image obtained by the camera device, the object position detection device uses AI to calculate the pose information of the object to be detected and provides information on whether the pose is correct. Feature TA8: The object position detection device includes a dual-camera ranging device; the object position detection device obtains the height information of the object to be detected based on the dual-camera ranging device; Feature TA9: The detection position adjustment device includes a lifting column, the detection bed is connected to the lifting column, and the lifting column can drive the detection bed to move up and down; Feature TA10: The rotating device includes a rotating motor and a rotating shaft. The rotating motor drives the rotating shaft to rotate, and the rotating shaft is used to drive the rotating arm to rotate. The bushing of the rotating shaft and the rotating shaft are respectively connected to the support frame A or the rotating arm. Feature TA11: The rotating device includes a DD motor; one end of the DD motor is connected to the support frame A, and the other end of the DD motor is connected to the rotating arm; Feature TA12: The part of the detection bed that receives radiation irradiation for imaging is made of carbon fiber material.

4. The medical radiation detection device as described in claim 1, characterized in that, The radiation source is a cone-beam X-ray source, and the imager is an amorphous silicon flat panel imager; Includes any one or more of the following features, Feature TB1: The X-ray source includes a kV X-ray source; Feature TB2: The X-ray source emits X-rays with an energy of 100kV; Feature TB3: The X-ray source includes a linear accelerator; Feature TB4: The detection pixel count of the amorphous silicon flat panel imager is greater than 3072×3072; Feature TB5: The dimensions of the amorphous silicon flat panel imager are greater than 43cm × 43cm.

5. The medical radiation detection device as described in claim 1, characterized in that, It also includes a support frame B and a rotating connection device. The support frame B and the support frame A are mechanically coupled through the rotating connection device. The support frame B supports the test bed and can drive the test bed to rotate around the rotating connection device.

6. The medical radiation detection device as described in claim 1, characterized in that, It also includes a support frame C, the bottom of which is connected to at least three casters. The support frame C supports the testing bed and can drive the testing bed to rotate. At least one caster has a locking mechanism.

7. The medical radiation detection device as described in claim 1, characterized in that, The control module can control the rotating arm to rotate to a set angle, control the radiation source to emit radiation signals, control the imager to detect radiation signal values, and perform dynamic 2D imaging of the object being detected.

8. A method for training AI-based posture recognition of radiation detection objects, characterized in that, include Place the object to be tested on the testing bed; Set the distance between the testing bed and the radiation source or imager; Take a picture of the object being detected to obtain a pose image; Label the posture image to obtain a posture label image; AI was trained using pose-labeled images to obtain a dataset of pose recognition features for radiation detection objects.

9. The method for training AI-based posture recognition of radiation detection objects according to claim 1, characterized in that, Includes any one or more of the following features, Feature TD1: The detection object is photographed using a dual-camera ranging device; a pose image is obtained; the obtained pose image includes the depth information of the detection object; Feature TD2: Take a picture of the object to be detected using a camera; obtain a pose image; the detection position of the object covers a standard reference object or standard reference image; Feature TD3: The detection object is any one or more of cats, dogs, and humans of different sizes; Feature TD4: Set different distances between the detection bed and the radiation source or imager; take pictures of the object being detected to obtain corresponding posture images at different distances; Feature TD5: The radiation detection object pose AI recognition feature dataset is obtained by training at least 100 sets of different pose identification images; Feature TD6: Identify the correct or incorrect pose of the detected object in the pose image, obtain the detected pose identification image, use the detected pose identification image for AI training, and obtain the AI ​​recognition feature dataset of the detected object pose, which includes the detected pose recognition information.

10. A radiation detection posture AI recognition method, characterized in that, Place the object to be tested on the testing bed; Take a picture of the object being detected to obtain a pose image; The posture image is identified using an AI recognition algorithm to obtain information on whether the posture of the detected object is correct.

11. The radiation detection posture AI recognition method according to claim 10, characterized in that, It also includes AI recognition algorithms that provide posture adjustment suggestions.

12. The radiation detection posture AI recognition method according to claim 10, characterized in that, Use the radiation detection object pose AI recognition feature dataset as described in any one of claims 8 to 9.

13. An electronic device, characterized in that, This includes memory, processor, and computer programs stored in memory and capable of running on the processor. When the processor executes the program, it implements the method as described in any one of claims 8 to 12.

14. A readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the method as described in any one of claims 8 to 12; It stores data, which is the AI ​​recognition feature dataset of the radiation detection object posture as described in any one of claims 8 to 9.

15. The medical radiation detection device as described in claim 6, characterized in that, The support frame C includes a support part C1 and a test bed sliding part C2; the support part C1 and the test bed sliding part C2 are connected by a slide rail; the test bed sliding part C2 can slide relative to the support part C1.

16. The medical radiation detection device as described in claim 15, characterized in that, It also includes a remote control, which can control the sliding direction and distance of the sliding part C2 of the testing bed relative to the support part C1.

17. The medical radiation detection device as described in claim 1, characterized in that, It also includes an emergency braking button, which is electrically connected to the control module and can control the rotation state of the rotating arm.

18. The medical radiation detection device as described in claim 1, characterized in that, It also includes at least one distance sensor, which is electrically connected to the control module. The distance sensor is installed at both ends of the rotating arm and is used to detect the distance of objects near the rotating arm. When the distance of the detected object is less than a set value, the rotating arm is prohibited from rotating.

19. The medical radiation detection device as described in claim 18, characterized in that, The distance sensors are located on both sides of the transmitter.

20. The medical radiation detection device as described in claim 18, characterized in that, The distance sensors are located on both sides of the imager.

21. The medical radiation detection device as described in claim 1, characterized in that, It also includes at least one positioning cursor transmitter, which is installed near the emission source and emits a positioning pattern for illuminating the detection area.

22. The medical radiation detection device as described in claim 1 or 2, characterized in that, It also includes a display, which is electrically connected to the control module.

23. The medical radiation detection device as described in claim 22, characterized in that... The display is mechanically connected to the support frame A.

24. The medical radiation detection device as described in claim 22, characterized in that... The display is mechanically connected to one end of the testing bed.

Images