[go: up one dir, main page]
More Web Proxy on the site http://driver.im/

WO2024101477A1 - Système de tomosynthèse numérique en ligne en temps réel - Google Patents

Système de tomosynthèse numérique en ligne en temps réel Download PDF

Info

Publication number
WO2024101477A1
WO2024101477A1 PCT/KR2022/017652 KR2022017652W WO2024101477A1 WO 2024101477 A1 WO2024101477 A1 WO 2024101477A1 KR 2022017652 W KR2022017652 W KR 2022017652W WO 2024101477 A1 WO2024101477 A1 WO 2024101477A1
Authority
WO
WIPO (PCT)
Prior art keywords
subject
projection data
ray
synthesis system
tomographic image
Prior art date
Application number
PCT/KR2022/017652
Other languages
English (en)
Korean (ko)
Inventor
김경우
이희신
김진국
Original Assignee
제이피아이헬스케어 주식회사
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 제이피아이헬스케어 주식회사 filed Critical 제이피아이헬스케어 주식회사
Priority to PCT/KR2022/017652 priority Critical patent/WO2024101477A1/fr
Publication of WO2024101477A1 publication Critical patent/WO2024101477A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N23/00—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00
    • G01N23/02—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by transmitting the radiation through the material
    • G01N23/04—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by transmitting the radiation through the material and forming images of the material
    • G01N23/046—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by transmitting the radiation through the material and forming images of the material using tomography, e.g. computed tomography [CT]
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N23/00—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00
    • G01N23/02—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by transmitting the radiation through the material
    • G01N23/06—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by transmitting the radiation through the material and measuring the absorption
    • G01N23/083—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by transmitting the radiation through the material and measuring the absorption the radiation being X-rays
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T11/00—2D [Two Dimensional] image generation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration

Definitions

  • the present invention relates to a real-time in-line digital tomography synthesis system that can be applied to an industrial in-line environment.
  • a digital tomosynthesis system is a device that reconstructs images in three dimensions using projection image data acquired from multiple angles within a limited angle.
  • the subject is fixed so that it does not move, the X-ray generator and the Reconstruct and output it as a single 3D image.
  • the DTS device is capable of acquiring 3D images at high speed, it has limitations that make it difficult to apply it to an industrial in-line environment because the X-ray generator and X-ray detector must be moved.
  • the present invention uses an in-line method of moving the subject instead of the X-ray generator and the We aim to provide a real-time in-line digital tomography synthesis system that allows
  • the present invention uses an in-line method of moving the subject instead of the X-ray generator and the We aim to provide a real-time in-line digital tomography synthesis system that allows
  • the digital tomographic image synthesis system includes an object moving rail that moves the object in a preset direction and a preset speed, and an X-ray generator fixed to face each other in the first direction of the object moving rail. and a pair of X-ray detectors, an object location confirmation unit that confirms and reports the current location of the object based on an image or sensor, and a control unit that controls at least one of the object movement rail, the pair of X-ray detectors, and the object location confirmation unit.
  • the operating method of the digital tomography synthesis system includes the steps of the control unit setting the object moving rail to a moving state, and when the current location of the object is included within the first imaging area, setting the object moving rail to a stationary state.
  • Step in a standby state in which the object movement rail does not move, acquiring first projection data for the object stationary within the first imaging area based on the X-ray generator and the X-ray detector, the first projection data is acquired. If so, setting the object moving rail to a moving state, the control unit stops the object moving rail if the current position of the object is included within the second imaging area, 2 Obtaining second projection data for a subject stationary within the imaging area, and producing an output image including at least one of a cross-sectional tomography image and a three-dimensional image based on the first projection data and the second projection data. Includes a reconstruction step.
  • the step of acquiring first projection data in the operating method of the digital tomography synthesis system includes acquiring a plurality of first projection data for a subject stationary within a first imaging area;
  • a step of obtaining filtered first projection data by applying a temporal filter to the first projection data of It includes obtaining a plurality of second projection data and applying a temporal filter to the plurality of second projection data to obtain filtered second projection data.
  • At least one of the step of acquiring a plurality of first projection data and the step of acquiring a plurality of second projection data in the operating method of the digital tomographic image synthesis system includes the X-ray generator having a plurality of pulse-type It includes radiating X-rays to an X-ray detection unit and receiving the plurality of pulsed X-rays by the X-ray detection unit to obtain at least one of a plurality of first projection data and a plurality of second projection data.
  • the reconstructing step of the operating method of the digital tomography image synthesis system includes reconstructing at least one of a cross-sectional tomography image and a 3D image based on the filtered first projection data and the filtered second projection data. It includes steps to:
  • the X-ray generator is located at the lower part of the subject moving rail
  • the X-ray detector is located at the upper part of the subject moving rail.
  • the focal size of the X-rays emitted from the X-ray generator and focused on the X-ray detector is 50 micrometers or less.
  • the operating method of the digital tomographic image synthesis system includes the steps and results of applying the output image to a predetermined good/failure judgment machine learning model to obtain result information indicating whether the subject shown in the output image is normal. It further includes a step of outputting information, and the good/failure judgment machine learning model is generated by machine learning the relationship between a plurality of past output images determined to be normal and label information indicating that the plurality of past output images are normal.
  • the step of acquiring result information of the operating method of the digital tomographic image synthesis system includes applying the output image to a predetermined good/failure judgment machine learning model to obtain probability information that the subject shown in the output image is normal. If the step and the normal probability information are greater than or equal to a predetermined first threshold probability, it includes the step of determining the result information as normal.
  • FIG. 1 and 2 are diagrams for explaining a real-time inline digital tomographic image synthesis system according to a first embodiment of the present invention.
  • Figure 3 is a diagram showing an X-ray image for each subject location obtained according to the first embodiment of the present invention.
  • Figure 4 is a diagram for explaining an image reconstruction method according to the first embodiment of the present invention.
  • Figure 5 is a diagram for explaining a real-time in-line digital tomographic image synthesis system according to a second embodiment of the present invention.
  • Figure 6 is a diagram for explaining a real-time in-line digital tomographic image synthesis system according to a third embodiment of the present invention.
  • Figure 7 is a diagram for explaining a real-time in-line digital tomographic image synthesis system according to a fourth embodiment of the present invention.
  • Figure 8 shows the configuration of a digital tomographic image synthesis system according to an embodiment of the present disclosure.
  • Figure 9 shows a method of operating a digital tomographic image synthesis system according to an embodiment of the present disclosure.
  • Figure 10 is a flowchart showing a method of operating a tomographic image synthesis system according to an embodiment of the present disclosure.
  • FIG. 11 is a diagram for explaining the operation of a tomographic image synthesis system according to an embodiment of the present disclosure.
  • Figure 12 is a diagram for explaining the effect of the system according to an embodiment of the present disclosure.
  • block diagrams herein should be understood as representing a conceptual view of an example circuit embodying the principles of the invention.
  • all flow diagrams, state transition diagrams, pseudo-code, etc. are understood to represent various processes that can be substantially represented on a computer-readable medium and are performed by a computer or processor, whether or not the computer or processor is explicitly shown. It has to be.
  • Figures 1 and 2 are diagrams for explaining a real-time in-line digital tomographic image synthesis system according to a first embodiment of the present invention, where Figure 1 is an external view and Figure 2 is a configuration diagram.
  • the digital tomographic image synthesis system of the present invention includes an object movement rail 110, an X-ray generator 120, an X-ray detector 130, an object location confirmation unit 140, and an image reconstruction unit 150.
  • the subject movement rail 110 provides a subject support space and simultaneously moves the subject 200 placed in the subject support space at a constant speed and in a certain direction.
  • the X-ray generator 120 is fixedly installed in the first direction (eg, vertical direction) of the object moving rail 110 to continuously or repeatedly generate X-rays having the same irradiation angle.
  • the irradiation angle may be approximately 95 degrees.
  • the X-ray detection unit 130 is fixedly installed in the first direction of the subject moving rail 110 to face the X-ray generator 121 with the subject 200 in between, and produces It is acquired continuously or repeatedly and output.
  • the subject location confirmation unit 140 is capable of inferring and notifying the current location of the subject based on the relative position value of the subject in the X-ray image.
  • a distance sensor that senses and reports the distance between the starting point of the object moving rail 112 and the current object, and a number of object detection sensors distributed at preset locations to sense and report the presence of the object are used. Therefore, the current location of the subject can be directly measured and notified, but the specific implementation plan does not need to be limited to this.
  • the image reconstruction unit 150 selectively receives the output image of the Obtain X-ray images (i.e., X-ray images for each subject location).
  • the X-ray images for each subject location can be reconstructed into at least one of a tomography image for each cross-section of the subject and a 3D image for the user.
  • Figure 3 is a diagram showing an X-ray image for each subject location obtained according to the first embodiment of the present invention.
  • the present invention can obtain an X-ray image for each location of the subject by moving the subject instead of the X-ray generator and the X-ray detector. It can be seen that the X-ray shooting angle has different effects for each location.
  • Figure 4 is a diagram for explaining an image reconstruction method according to the first embodiment of the present invention. For convenience of explanation, hereinafter, only the case where images of the subject are acquired at a total of three positions will be described.
  • step S1 while the subject is placed on the subject movement rail 110, the subject movement rail 110 may move the subject in a preset direction and at a preset speed.
  • step S1 when the subject moving rail 110 receives the user's shooting input from the input unit or detects that the subject is placed on the subject moving rail 110 from the sensor unit, the subject is moved in a preset direction and in a preset direction. The position can be moved with speed.
  • the subject location confirmation unit 140 confirms and reports the current location of the subject in real time
  • the image reconstruction unit 150 detects the subject's X-rays whenever the subject is located at preset points 1, 2, and 3. Acquire images repeatedly.
  • the X-ray image may be projection image data for all voxels constituting the subject.
  • step S2 the total overlapping image data is calculated by adding up all the X-ray images (i.e., projection image data of each voxel) acquired at each subject location.
  • three complementary overlapping image data are calculated by summing the The image processing operation of normalizing the difference with the image data by dividing it by the total number of
  • step S3 the cross-sectional tomographic image obtained through step S2 is provided to the user as is, or the cross-sectional tomographic image is sequentially stacked in cross-sectional order to synthesize one 3D image and then provide it to the user. do.
  • Figure 5 is a diagram for explaining a real-time in-line digital tomographic image synthesis system according to a second embodiment of the present invention.
  • the digital tomographic image synthesis system of the present invention expands and installs a plurality of pairs of the X-ray generator 120 and the make it possible
  • two pairs of an X-ray generator 120 and an X-ray detector 130 are provided, and the first X-ray generator 121 and the first They are installed to face each other in a direction (for example, a vertical direction), and the second They can be installed facing each other in a horizontal direction (for example).
  • tomographic images of xy cross-section are transmitted through the first X-ray generator 121 and the first It is possible to obtain and provide tomographic images of xz cross-section through the second X-ray generator 122 and the second X-ray detector 132 installed to face each other in the second direction of 110.
  • the number of X-ray generators 120 and X-ray detectors 130 can be easily increased, and the number of cross-sections (e.g., xy cross-section, xz cross-section) that can be detected through them can also be easily increased. Let it happen. As a result, the type and amount of information that can be provided to users through the same system can become more diverse.
  • Figure 6 is a diagram for explaining a real-time in-line digital tomographic image synthesis system according to a third embodiment of the present invention.
  • the digital tomographic image synthesis system of the present invention includes a plurality of pairs of X-ray generators 120 and X-ray detectors 130, and a plurality of X-ray generators. By fixing the pair of 120 and the X-ray detector 130 adjacent to each other in the same direction, the image acquisition angle can be expanded through them.
  • the image acquisition angle of the digital tomography synthesis system is easily increased in proportion to the number of pairs of the X-ray generator 120 and the X-ray detector 130 (A ⁇ B ⁇ C).
  • Figure 7 is a diagram for explaining a real-time in-line digital tomographic image synthesis system according to a fourth embodiment of the present invention.
  • the digital tomographic image synthesis system of the present invention allows the pair of the X-ray generator 120 and the X-ray detector 130 to be arranged in all directions.
  • the X-ray generator 120 may be placed on the upper part of the object moving rail 110 so that the X-rays of the X-ray generator 120 are radiated downward.
  • the X-ray detection unit 130 may be located below the object moving rail.
  • the X-ray generator 120 may be placed below the object moving rail 110 so that the X-rays from the X-ray generator 120 are radiated upward.
  • the X-ray detection unit 130 may be located at the top of the object moving rail.
  • the X-ray generator 120 may be placed on the side so that X-rays are irradiated from the side, and the X-ray detector 130 may be placed against the X-ray generator 120.
  • the focal size of the X-rays emitted from the X-ray generator 120 and focused on the X-ray detector 130 may be 50 micrometers or less. Therefore, the spatial resolution of the output image can be high. Therefore, it is possible to easily determine whether there is a problem inside the subject shown in the output image.
  • Focal spot to object distance may be 5 mm or more and 10 mm or less.
  • the distance between the X-ray generator 120 and the object may be shorter than the distance between the object and the X-ray detector 130.
  • the output image for the subject may be an image taken by enlarging the subject. Therefore, it is possible to easily determine whether the subject is normal based on the subject displayed enlarged in the output image.
  • the output image can be displayed by enlarging only the problematic part of the subject. For example, if the object under test is a battery, the part inside the battery where at least one pin is located may appear in the output image.
  • the output image may include the entire subject.
  • the focal size of the X-ray generator 120 and the X-ray detector 130 may be 50 micrometers or less. Therefore, the spatial resolution of the output image can be high. That is, the subject appearing in the output image is enlarged and can be displayed clearly. Therefore, it is easy to check whether the subject shown in the output image is normal.
  • Figure 8 shows the configuration of a digital tomographic image synthesis system according to an embodiment of the present disclosure.
  • the tomographic image synthesis system 100 includes an X-ray generator 120, a high voltage generator, an ) and a control unit 800.
  • the control unit 800 may control the overall operation of the tomographic image synthesis system 100.
  • the high voltage generator generates a high voltage for generating X-rays and applies it to the X-ray source included in the X-ray generator 120.
  • the X-ray generator 120 may include an X-ray source that generates X-rays by receiving a high voltage generated by the high voltage generator.
  • the X-ray source includes an X-ray tube, and the X-ray tube may be implemented as a two-pole vacuum tube with an anode and a cathode. Additionally, the X-ray generator 120 may include a collimator that guides the path of the X-rays emitted from the X-ray source and adjusts the X-ray irradiation area.
  • the X-ray detection unit 130 detects X-rays radiated from the X-ray generator 120 and passing through the object.
  • the X-ray detection unit 130 may be a digital X-ray detection unit 130.
  • the X-ray detection unit 130 may be implemented using a TFT or a CCD.
  • the X-ray detection unit 130 may be included in the tomography image synthesis system 100 or may be a separate device that can be connected to and separated from the tomography image synthesis system 100.
  • the tomographic image synthesis system 100 may include a sensor unit 810.
  • the sensor unit 810 can acquire various information using at least one sensor.
  • the sensor unit 810 may be provided as a sensor that uses measurement means such as pressure, potential, and optics.
  • the sensor unit 810 may include at least one of a distance measurement sensor or an encoder.
  • the sensor may include a pressure sensor, infrared sensor, LED sensor, touch sensor, etc.
  • the sensor unit 810 may include a distance sensor that senses and reports the distance between the above-described starting point and the current subject, and a plurality of object detection sensors that are distributed at preset locations to sense and report the presence of the subject. .
  • the tomographic image synthesis system 100 may include a communication unit 820.
  • the communication unit 820 may be configured to communicate wired or wirelessly with an internal module or an external device of the tomographic image synthesis system 100.
  • External devices may include external servers and user terminals.
  • the user terminal may include a PC, smartphone, tablet, or wearable device.
  • the communication unit 820 may include a wired/wireless communication module for network connection.
  • wireless communication technology for example, wireless LAN (WLAN) (Wi-Fi), wireless broadband (Wibro), World Interoperability for Microwave Access (Wimax), High Speed Downlink Packet Access (HSDPA), etc. may be used.
  • the network connection unit includes a short-distance communication module and can transmit and receive data with any device/terminal located in a short distance.
  • short range communication technologies include Bluetooth, Radio Frequency Identification (RFID), infrared data association (IrDA), Ultra-Wideband (UWB), and ZigBee. It is not limited to this.
  • the tomographic image synthesis system 100 can communicate with a workstation.
  • the workstation can store or analyze images captured by the tomographic image synthesis system 100.
  • Communication between the workstation and the tomographic synthesis system 100 uses high-speed digital interfaces such as LVDS (Low Voltage Differential Signaling), asynchronous serial communication such as UART (universal asynchronous receiver transmitter), erroneous synchronous serial communication, or CAN (Controller Area).
  • LVDS Low Voltage Differential Signaling
  • UART universal asynchronous receiver transmitter
  • erroneous synchronous serial communication or CAN (Controller Area).
  • Low-latency network protocols such as (Network) can be used, and various communication methods can be used within the range obvious to those skilled in the art.
  • the tomographic image synthesis system 100 may include a memory 830.
  • the control unit 800 may execute commands stored in memory.
  • the memory 830 may be included in the control unit 800 or may be external to the control unit 800.
  • the memory 830 may store various information related to the tomographic image synthesis system 100.
  • the memory 830 may include an operation method of an X-ray source and related information, and may include captured images and user authentication information, but is not limited thereto.
  • the memory 830 may be implemented through a non-volatile storage medium that can continuously store arbitrary data.
  • memory 830 may include, but is not limited to, disks, optical disks, and magneto-optical storage devices, as well as storage devices based on flash memory and/or battery-backed memory.
  • the memory 830 is a main storage device directly accessed by the processor, such as random access memory (RAM) such as dynamic random access memory (DRAM) and static random access memory (SRAM), and the stored data is stored when the power is turned off. It may refer to a volatile storage device in which information is instantaneously erased, but is not limited to these.
  • This memory 830 can be operated by the control unit 800.
  • the tomographic image synthesis system 100 may further include a manipulation unit that provides an interface for operating the tomographic image synthesis system 100.
  • the manipulation unit may include an output unit 840 and an input unit 850.
  • the output unit 840 may output sound and images representing information related to imaging, such as X-ray irradiation, under the control of the control unit 800.
  • the output unit 840 may include a speaker or display.
  • the output unit 840 may output an X-ray image generated by the control unit 800.
  • the output unit 840 may output information necessary for the user to operate the tomographic image synthesis system 100, such as a user interface (UI), user information, or object information.
  • Examples of the output unit 840 may include a speaker, printer, CRT display, LCD display, PDP display, OLED display, FED display, LED display, VFD display, DLP display, FPD display, 3D display, transparent display, etc. , and other various output devices within the scope apparent to those skilled in the art.
  • the workstation and the tomographic image synthesis system 100 may be connected to each other wirelessly or wired, and the workstation may exist in a space physically separated from the tomographic image synthesis system 100.
  • the workstation may include a storage server.
  • the storage server may store X-ray images, information about the object, information about the user, etc.
  • the workstation may include a review device.
  • the review device may receive the X-ray image from the storage server based on the user's command and process the X-ray image.
  • the tomographic image synthesis system 100 can process the X-ray image instead of the workstation.
  • the workstation may include an output unit, an input unit, and a control unit.
  • the output unit and the input unit provide the user with an interface for operating the workstation and the tomographic image synthesis system 100.
  • the control unit of the workstation may control the workstation and the tomographic image synthesis system 100. However, it is not limited to this, and the tomographic image synthesis system 100 can control the workstation and the tomographic image synthesis system 100.
  • the tomographic image synthesis system 100 can be controlled through a workstation and can also be controlled by the control unit 800 included in the tomographic image synthesis system 100. Accordingly, the user may control the tomography image synthesis system 100 through a workstation or control the tomography image synthesis system 100 through the manipulation unit and control unit 800 included in the tomography image synthesis system 100. In other words, the user may control the tomographic image synthesis system 100 remotely through a workstation, or directly control the tomographic image synthesis system 100.
  • the control unit of the workstation and the control unit 800 of the tomographic image synthesis system 100 may be separate, but are not limited to this.
  • the control unit of the workstation and the control unit 800 of the tomographic image synthesis system 100 may be implemented as one integrated control unit, and the integrated control unit may be included in only one of the workstation and the tomographic image synthesis system 100.
  • the control unit 800 may refer to a control unit of a workstation and/or a control unit of the tomographic image synthesis system 100.
  • the output unit and input unit of the workstation and the output unit 840 and input unit 850 of the tomographic image synthesis system 100 may each provide the user with an interface for operating the tomographic image synthesis system 100.
  • the workstation and tomographic image synthesis system 100 may each include an output unit and an input unit, but are not limited thereto.
  • the output unit or input unit may be implemented only in one of the workstation and the tomographic image synthesis system 100.
  • the input unit 850 refers to the input unit of the workstation and/or the input unit of the tomographic image synthesis system 100
  • the output unit 840 refers to the output unit of the workstation and/or the output unit of the tomographic image synthesis system 100. It means wealth.
  • the input unit 850 may receive commands for operating the tomographic image synthesis system 100 and various information regarding X-ray imaging from the user.
  • the control unit 800 may control or manipulate the tomographic image synthesis system 100 based on information input to the input unit 850.
  • the input unit 850 may include a joystick, keyboard, mouse, touch screen, voice recognition device, fingerprint recognition device, iris recognition device, etc., and may include other input devices known to those skilled in the art.
  • the user can input a command for X-ray irradiation through the input unit 850, and the input unit 850 may be provided with a switch for inputting such a command.
  • the control unit 800 controls the positions, shooting timing, and shooting conditions of the X-ray generator 120 and the X-ray detector 130 according to the shooting conditions set by the user.
  • control unit 800 controls the high voltage generator and the Control. Additionally, the control unit 800 adjusts the position of the X-ray detection unit 130 and controls the operation timing of the X-ray detection unit 130 according to predetermined imaging conditions.
  • control unit 800 generates an X-ray image of the object using image data received through the X-ray detection unit 130. Specifically, the control unit 800 can receive image data from the X-ray detection unit 130, remove noise from the image data, and adjust the dynamic range and interleaving to generate an X-ray image of the object. there is.
  • Figure 9 shows a method of operating a digital tomographic image synthesis system according to an embodiment of the present disclosure.
  • the digital tomography synthesis system 100 may include a control unit 800.
  • the control unit 800 may control at least one of an object movement rail, a pair of X-ray detectors, and an object location confirmation unit. Below, the operation of the control unit 800 will be described.
  • the control unit 800 may perform step 910 of setting the object moving rail 110 to a moving state.
  • Step 910 may be included in step S1.
  • the subject movement rail 110 can move the subject in a preset direction and at a preset speed.
  • the state of the object moving rail 110 may include a moving state, a stationary state, and a standby state.
  • the moving state is a state in which the subject is moved in a preset direction at a preset speed.
  • the stationary state is a state in the process of transitioning from a moving state to a waiting state.
  • the standby state is a state in which the subject is stopped.
  • the control unit 800 may perform step 920 of setting the subject moving rail 110 to a stationary state.
  • the current location of the subject may be the same as the location of the subject in picture 411 of FIG. 4 .
  • the first capturing area may be a predetermined area.
  • the first imaging area may be a predetermined area of the X-ray detector 130. However, it is not limited to this, and the first imaging area may be a predetermined area of the object moving rail 110.
  • the first capturing area may include point 1.
  • the controller 800 can control the subject to stop at point 1.
  • the control unit 800 performs a first projection on a subject stationary within the first imaging area based on the X-ray generator 120 and the X-ray detector 130.
  • a step 930 of acquiring data may be performed.
  • the first projection data may correspond to acquired image 1 of FIG. 4.
  • the first projection data may be included in the plurality of X-ray images described above. However, it is not limited to this.
  • the first projection data may not be an image, but may be information acquired by the X-ray detector 130 based on X-rays that pass through the subject and reach the X-ray detector 130.
  • the X-ray generator 120 may radiate X-rays to the subject in all of a moving state, a stationary state, and a standby state. However, the tomographic image synthesis system 100 can reconstruct the output image using only projection data in the standby state. However, it is not limited to this.
  • the X-ray generator 120 can radiate X-rays only to the subject in a standby state. In this way, since the This is because X-rays can corrode instrumentation. More specifically, the more the components of the Additionally, the X-ray detection unit 130 includes a fluorescent plate, and the lifespan of the fluorescent plate may be reduced by X-rays.
  • the control unit 800 may perform step 940 of setting the object movement rail 110 to a moving state.
  • the object moving rail 110 can be moved from the standby state.
  • the subject movement rail 110 can move the subject in a preset direction and at a preset speed.
  • the control unit 800 may perform step 950 of stopping the moving rail of the subject.
  • the current location of the subject may be the same as the location of the subject in picture 412 of FIG. 4 .
  • the current location of the subject may be the same as the location of the subject in figure 413 of FIG. 4.
  • the system 100 may include an Nth imaging area other than the first capturing area and the second capturing area. N may be a natural number.
  • the second capturing area may be a predetermined area.
  • the second imaging area may be a predetermined area of the X-ray detector 130. However, it is not limited to this, and the second imaging area may be a predetermined area of the object moving rail 110.
  • the second capturing area may include point 2.
  • the controller 800 can control the subject to stop at point 2. As already described, the control unit 800 can control the subject to stop at a predetermined point N.
  • N may be a natural number.
  • the control unit 800 may perform step 960 of acquiring second projection data for an object stationary within the second imaging area based on the X-ray generator 120 and the X-ray detector 130.
  • the second projection data may correspond to acquired image 2 of FIG. 4 .
  • the second projection data may be included in the plurality of X-ray images described above. However, it is not limited to this.
  • the second projection data may not be an image, but may be information acquired by the X-ray detector 130 based on X-rays that pass through the subject and reach the X-ray detector 130.
  • the control unit 800 may acquire N-th projection data for a subject stationary within a predetermined N-th imaging area.
  • N may be a natural number.
  • a step 970 of reconstructing an output image including at least one of a cross-sectional tomography image and a 3D image based on the first projection data and the second projection data may be performed.
  • Step 970 may include at least one of step S2 and step S3. Additionally, the control unit 800 may reconstruct the output image based on the Nth projection data.
  • the step 930 of acquiring first projection data may further include the following processes.
  • the first projection data may include one image. Additionally, the first projection data may be data for acquiring one image. However, it is not limited to this.
  • the control unit 800 may perform a step of acquiring a plurality of first projection data for an object stationary within the first capturing area. That is, the first projection data may include data for acquiring a plurality of images. Additionally, the control unit 800 may acquire a plurality of projection data for a stationary subject. The plurality of first projection data may be data captured at different times. The controller 800 may obtain filtered first projection data from which noise has been removed using a plurality of first projection data.
  • the controller 800 may apply a temporal filter to a plurality of first projection data to obtain filtered first projection data.
  • a temporal filter may be a filter for removing noise from an image.
  • the temporal filter may include at least one of an average filter or a median filter.
  • the controller 800 may obtain filtered first projection data by averaging or obtaining a median value of the plurality of first projection data. More specifically, the controller 800 may determine a pixel value at a specific location of the first projection data by filtering at least one of the average value and the median value of pixels at a specific location included in the plurality of first projection data.
  • the pixel at a specific location in the filtered first projection data includes pixels at a specific location included in the plurality of first projection data and the surroundings of the specific location included in the plurality of first projection data. It may also be obtained based on the pixels in .
  • the surrounding pixels may be pixels located at least one of the left, right, upper, lower, upper left, lower left, upper right, and lower right sides of the pixel at a specific location.
  • the control unit 800 may set the weight for a pixel located at a specific location in the plurality of first projection data to be larger than the weight for the pixel located around the specific location in the plurality of first projection data. That is, when acquiring the filtered first projection data, the control unit 800 may increase the influence of a pixel at a specific position of the plurality of first projection data to be greater than that of surrounding pixels.
  • the pixel at a specific position in the filtered first projection data is the average or median of the remaining pixels after excluding the value of the abnormal pixel among the pixels at the specific position included in the plurality of first projection data. It can be one of the following values:
  • the control unit 800 may obtain the standard deviation of pixels at specific positions included in the plurality of first projection data.
  • the control unit 800 may search for abnormal pixels when the standard deviation is greater than a predetermined threshold deviation.
  • the controller 800 may determine the pixel value furthest from the average value of pixels at a specific location included in the plurality of first projection data as the value of the abnormal pixel.
  • the controller 800 may determine one of the average value and the median value of the remaining pixels excluding the abnormal pixel as the value of the pixel at a specific location of the filtered first projection data.
  • the first projection data may include 1-1 projection data, 1-2 projection data, and 1-3 projection data.
  • the control unit 800 may acquire a first pixel at a specific location of the 1-1 projection data, a second pixel at a specific location of the 1-2 projection data, and a third pixel at a specific location of the 1-3 projection data. there is.
  • the control unit 800 applies the average value and median value of the value of the first pixel and the value of the second pixel to the filtered first projection. It can be determined by the value of the pixel at a specific location in the data.
  • the control unit 800 may further perform the following process to obtain the second projection data (step 960).
  • the control unit 800 may perform a step of acquiring a plurality of second projection data for a subject stopped inside the second capturing area. That is, the second projection data may include data for acquiring a plurality of images. Additionally, the control unit 800 may acquire a plurality of projection data for a stationary subject. The plurality of second projection data may be data captured at different times. The controller 800 may obtain filtered second projection data from which noise has been removed using a plurality of second projection data.
  • the controller 800 may apply a temporal filter to a plurality of second projection data to obtain filtered second projection data. Since the process of applying a temporal filter has already been described, redundant explanation will be omitted.
  • At least one of acquiring a plurality of first projection data and acquiring a plurality of second projection data may further include the following steps.
  • the control unit 800 may perform a step in which the X-ray generator 120 radiates a plurality of pulsed X-rays to the X-ray detector 130.
  • the fact that the X-ray generator 120 emits a plurality of pulsed X-rays may mean that the emission and non-emission of X-rays are continuous. More specifically, the plurality of pulsed X-rays may mean that X-rays are emitted for a predetermined first time, are not emitted for a second time, are emitted again for a third time, and are not emitted for a fourth time.
  • the first time and the third time may be the same or different.
  • the second time and the fourth time may be the same or different.
  • the X-ray generator 120 emits pulsed X-rays the lifespan of the system 100 can be extended. This is because when high-energy X-rays are continuously emitted, the components of the system are gradually damaged.
  • the X-ray detector !30 may receive a plurality of pulsed X-rays and obtain at least one of a plurality of first projection data and a plurality of second projection data.
  • the controller 800 may adjust the shape of the pulse so that the X-ray emission time increases over time. That is, the control unit 800 can make the first time shorter than the third time.
  • the plurality of N-th projection data may include low-dose images and high-dose images in the N-th imaging area. However, it is not limited to this, and the control unit 800 may reduce the X-ray emission time as time goes by.
  • control unit 800 may gradually increase or decrease the intensity of the X-ray. Accordingly, the plurality of N-th projection data may include low-dose images and high-dose images in the N-th imaging area.
  • the controller 800 may acquire a plurality of N-th projection data during the time the X-rays are emitted. More specifically, the control unit 800 may acquire at least one of a plurality of first projection data and a plurality of second projection data during the time of emitting X-rays.
  • the control unit 800 may further perform the following process to perform the reconfiguration step 970.
  • the controller 800 may include reconstructing at least one of a cross-sectional tomography image and a 3D image based on the filtered first projection data and the filtered second projection data.
  • Figure 10 is a flowchart showing a method of operating a tomographic image synthesis system according to an embodiment of the present disclosure. Additionally, FIG. 11 is a diagram for explaining the operation of a tomographic image synthesis system according to an embodiment of the present disclosure.
  • the control unit 800 applies the output image to a predetermined good/failure judgment machine learning model to obtain result information indicating whether the subject shown in the output image is normal (step 1010).
  • the adoption judgment machine learning model may be a rule-based model or a machine learning model.
  • a good/failure judgment machine learning model can be created by machine learning the relationship between a plurality of past output images determined to be normal and label information indicating that the plurality of past output images are normal.
  • a machine learning model can be built considering the application field of the learning model, the purpose of learning, or the computer performance of the device.
  • the machine learning model may be, for example, a boosting-based ML algorithm, a tree-based algorithm, or a model based on a neural network.
  • AdaBoost AdaBoost
  • GBM Gradient Boosting Machine
  • XGBoost Extra gradient boost
  • LightBoost Decision tree
  • Random forest Deep Neural Network
  • DNN Deep Neural Network
  • RNN Recurrent Neural Network
  • Long Short -Models such as Term Memory models (LSTM)
  • BRDNN Bidirectional Recurrent Deep Neural Network
  • CNN Convolutional Neural Networks
  • a machine learning model may include at least one of supervised learning and unsupervised learning.
  • the control unit 800 may perform step 1020 of outputting result information.
  • FIG. 10 will be described in more detail together with FIG. 11.
  • the system 100 may perform learning based on a plurality of past output images 1111 and a plurality of past result information 1112.
  • the system 100 may acquire a plurality of past output images 1111.
  • a plurality of past output images 1111 can be obtained through S1 to S3 in FIG. 4. Additionally, a plurality of past output images 1111 can be obtained through the steps in FIG. 9.
  • a plurality of past result information 1112 may correspond one-to-one to a plurality of past output images 1111.
  • the plurality of past result information 1112 may indicate whether the subject shown in the plurality of past output images 1111 is normal.
  • Multiple past result information 1112 can be determined by a person. However, it is not limited to this and may be automatically determined by a machine learning model.
  • the plurality of past result information 1112 may be a ground truth value.
  • the system 100 may acquire a pass-through judgment machine learning model 1120 based on a plurality of past output images 1111 and a plurality of past result information 1112.
  • the quality judgment machine learning model 1120 may be a set of weights learned by a machine learning model.
  • the system 100 can obtain a pass/fail judgment machine learning model 1120 using only a plurality of past output images 1111 in which the plurality of past result information 1112 indicates normality. That is, the system 100 may not use the plurality of past output images 1111 in which the plurality of past result information 1112 indicates abnormality. In this case, the quality judgment machine learning model 1120 may not learn the plurality of past output images 1111 in which the plurality of past result information 1112 indicates abnormality.
  • the output of the quality judgment machine learning model 1120 may represent probability information that the subject included in the output image 1131 is normal. The system 100 may determine that the higher the probability, the higher the probability that the subject included in the output image 1131 is normal.
  • the system 100 may acquire a pass/fail judgment machine learning model 1120 using only a plurality of past output images 1111 in which the plurality of past result information 1112 indicates abnormality. . That is, the system 100 may not use the plurality of past output images 1111 in which the plurality of past result information 1112 indicates normality. In this case, the quality judgment machine learning model 1120 may not learn the plurality of past output images 1111 in which the plurality of past result information 1112 indicates normality.
  • the output of the quality judgment machine learning model 1120 may represent probability information that the subject included in the output image 1131 is abnormal. The system 100 may determine that the higher the probability information, the higher the probability that the subject included in the output image 1131 is abnormal.
  • the system 100 acquires a pass/fail judgment machine learning model 1120 using a plurality of past output images 1111 in which a plurality of past result information 1112 indicates abnormality or normality. You can. The closer the result information 1140 of the quality judgment machine learning model 1120 is to a value indicating abnormality, the higher the probability of it being abnormal, and the closer it is to the value indicating normality, the higher the probability of being normal.
  • the system 100 may obtain at least one of a probability of being abnormal and a probability of being normal. If the probability of abnormality is high, the system 100 may determine that the subject included in the output image 1131 is abnormal. Additionally, the system 100 may determine that the subject included in the output image 1131 is normal when the probability of it being normal is high.
  • the system 100 may store the obtained pass/fail judgment machine learning model 1120 in memory or transmit it to another system.
  • the system 100 can obtain result information 1140 by applying the output image 1131 to the pass/failure judgment machine learning model 1120.
  • the system 100 may determine whether the subject included in the output image 1131 is normal based on the result information 1140.
  • the system 100 may further include the following process to perform the step 1010 of obtaining result information.
  • the system 100 may apply the output image 1131 to a predetermined good/failure judgment machine learning model 1120 to obtain probability information that the subject shown in the output image 1131 is normal.
  • the control unit 800 may perform a step of determining the result information to be normal. Additionally, if the normal probability information is greater than or equal to the second critical probability and less than the first critical probability, the control unit 800 may determine the result information to be undetermined. The control unit 800 can output result information to allow the user to determine whether the subject shown in the output image 1131 is normal. If the probability information of normality is less than a predetermined second threshold probability, the control unit 800 may perform a step of determining the result information to be abnormal.
  • the first threshold probability may be greater than or equal to the second threshold probability.
  • Figure 12 is a diagram for explaining the effect of the system according to an embodiment of the present disclosure.
  • FIG. 12 may show a reconstructed 3D image. Additionally, the voxel (pixel) size of FIG. 12 may be 47.50 micrometers.
  • Area 1210 is an image captured while the subject is moving on the subject movement rail 110
  • area 1220 is an image captured only in a standby state.
  • the movement speed cannot be increased to obtain a clear image.
  • the pins 1211 included in the object are not distinguished from each other.
  • a 3D image of relatively high quality can be confirmed.
  • the pins of the image at the bottom of area 1220 are clearly distinguished.
  • the system 100 can increase the moving speed when the object moving rail 110 is in a moving state. This is because the moving speed of the subject does not affect the quality of the image. Therefore, the system 100 of the present disclosure can quickly perform inspection while quickly moving the subject.
  • the system 100 of the present disclosure can quickly determine whether a subject is normal or not using a machine learning model, and can determine normality with minimal human assistance.
  • the system 100 of the present disclosure may be included in an inspection device that inspects whether a finished product is normal.
  • the inspection device can be used to inspect batteries, electronic products, circuits, and integrated circuits. If inspection efficiency does not keep up with production efficiency, the overall production speed of the product will inevitably slow down unless defective products are eventually shipped. To solve this problem, many inspection devices can be introduced, but this has the problem of cost. Therefore, a system that quickly performs inspection is required, and the system of the present disclosure can prevent delays in inspection by quickly inspecting finished products.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Immunology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Pathology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Radiology & Medical Imaging (AREA)
  • Artificial Intelligence (AREA)
  • Geometry (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Graphics (AREA)
  • Toxicology (AREA)
  • Mathematical Physics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Pulmonology (AREA)
  • Computing Systems (AREA)
  • Apparatus For Radiation Diagnosis (AREA)

Abstract

La présente divulgation concerne un procédé de fonctionnement pour un système de tomosynthèse numérique qui comprend les étapes au cours desquelles : une unité de commande règle un rail de déplacement de sujet pour qu'il soit dans un état de déplacement ; le rail de déplacement de sujet est réglé pour être dans un état arrêté si la position actuelle d'un sujet se trouve à l'intérieur d'une première zone d'imagerie ; des premières données de projection du sujet arrêté à l'intérieur de la première zone d'imagerie sont acquises sur la base d'une unité de génération de rayons X et d'une unité de détection de rayons X dans un état de veille dans lequel le rail de déplacement de sujet ne se déplace pas ; le rail de déplacement de sujet est réglé pour être dans l'état de déplacement si les premières données de projection sont acquises ; l'unité de commande arrête le rail de déplacement de sujet si la position actuelle du sujet se trouve à l'intérieur d'une seconde zone d'imagerie ; des secondes données de projection du sujet arrêté à l'intérieur de la seconde zone d'imagerie sont acquises sur la base de l'unité de génération de rayons X et de l'unité de détection de rayons X ; et une image de sortie comprenant une image de tomographie spécifique à une section transversale et/ou une image tridimensionnelle est reconstruite sur la base des premières données de projection et des secondes données de projection.
PCT/KR2022/017652 2022-11-10 2022-11-10 Système de tomosynthèse numérique en ligne en temps réel WO2024101477A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/KR2022/017652 WO2024101477A1 (fr) 2022-11-10 2022-11-10 Système de tomosynthèse numérique en ligne en temps réel

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/KR2022/017652 WO2024101477A1 (fr) 2022-11-10 2022-11-10 Système de tomosynthèse numérique en ligne en temps réel

Publications (1)

Publication Number Publication Date
WO2024101477A1 true WO2024101477A1 (fr) 2024-05-16

Family

ID=91032957

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2022/017652 WO2024101477A1 (fr) 2022-11-10 2022-11-10 Système de tomosynthèse numérique en ligne en temps réel

Country Status (1)

Country Link
WO (1) WO2024101477A1 (fr)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014517319A (ja) * 2011-06-14 2014-07-17 アナロジック コーポレイション セキュリティースキャナー
JP2017142217A (ja) * 2016-02-12 2017-08-17 国立大学法人京都工芸繊維大学 撮影装置、及び、撮影方法
JP2018124084A (ja) * 2017-01-30 2018-08-09 日本信号株式会社 Ｘ線検査装置及び手荷物検査装置
JP2021500571A (ja) * 2017-10-27 2021-01-07 ティアマ 複数の製造物体のインライン寸法制御のための方法及び設備
KR20210015180A (ko) * 2019-08-01 2021-02-10 주식회사 에스에프에이 2차 전지용 오버행 검사장치 및 이를 구비하는 2차 전지 제조 시스템

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014517319A (ja) * 2011-06-14 2014-07-17 アナロジック コーポレイション セキュリティースキャナー
JP2017142217A (ja) * 2016-02-12 2017-08-17 国立大学法人京都工芸繊維大学 撮影装置、及び、撮影方法
JP2018124084A (ja) * 2017-01-30 2018-08-09 日本信号株式会社 Ｘ線検査装置及び手荷物検査装置
JP2021500571A (ja) * 2017-10-27 2021-01-07 ティアマ 複数の製造物体のインライン寸法制御のための方法及び設備
KR20210015180A (ko) * 2019-08-01 2021-02-10 주식회사 에스에프에이 2차 전지용 오버행 검사장치 및 이를 구비하는 2차 전지 제조 시스템

Similar Documents

Publication Publication Date Title
WO2017039259A1 (fr) Appareil et procédé pour le diagnostic d'un équipement électrique à l'aide d'une caméra thermique
WO2017111464A1 (fr) Appareil d'imagerie à rayons x, procédé de commande associé et détecteur de rayons x
WO2021230534A1 (fr) Appareil de prédiction de lésion orbitaire et périorbitaire et procédé de prédiction associé
WO2019164382A1 (fr) Procédé d'inspection de l'état de montage d'un composant, appareil d'inspection de carte de circuit imprimé et support d'enregistrement lisible par ordinateur
WO2018105963A1 (fr) Appareil radiographique et procédé radiographique l'utilisant
JP2011146472A (ja) 太陽光発電システム
CN109381209A (zh) 放射线摄影系统及其工作方法
WO2024101477A1 (fr) Système de tomosynthèse numérique en ligne en temps réel
WO2018186584A1 (fr) Appareil de diagnostic facial tridimensionnel
WO2020204366A2 (fr) Procédé fournissant un guide de balayage et dispositif de traitement d'image associé
WO2017007086A1 (fr) Dispositif de diagnostic de cancer du sein
KR20200087308A (ko) 태양광 발전장치의 접속반 화재 경보 시스템
WO2020141888A1 (fr) Dispositif de gestion de l'environnement de ferme d'élevage
KR20170124231A (ko) 디스플레이 장치의 표시부 정상상태를 무인 검사하는 방법 및 시스템
EP3166497A1 (fr) Appareil d'imagerie médicale et procédé de balayage de celui-ci
WO2021040345A1 (fr) Dispositif électronique et procédé de commande de dispositif électronique
US20210345872A1 (en) Vision screening systems and methods
US11250593B2 (en) System and method for detecting and correcting defective image output from radiation-damaged video cameras
WO2021246643A1 (fr) Procédé et appareil de prédiction de dent manquante dans une image buccale
KR101665327B1 (ko) 전산단층촬영장치 및 방법
WO2018122300A1 (fr) Appareil et procédés de détection et d'évaluation de défaillances dans un système d'affichage
WO2018043878A1 (fr) Dispositif d'imagerie par résonance magnétique
WO2023101374A1 (fr) Procédé, dispositif et système de test d'ensemble non destructif basés sur l'intelligence artificielle pour objet
WO2024205338A1 (fr) Dispositif et procédé de prédiction de risque de maladie cardiaque basée sur l'intelligence artificielle
CN111247594B (zh) 自动化的不相容像素掩蔽

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22965249

Country of ref document: EP

Kind code of ref document: A1