JP6325939B2 - Radiation imaging apparatus and gain image creation method - Google Patents

Description

  The present invention relates to a radiation imaging apparatus, and more particularly to creation of a gain image used for gain correction of a radiation image captured by the radiation imaging apparatus.

  In recent years, as a detector for observing the distribution of X-rays transmitted through the human body used in medical X-ray imaging apparatuses and the like, a solid-state imaging device called a flat panel detector (hereinafter referred to as FPD) has been used. The use of large image sensors is becoming common. In such an X-ray imaging apparatus using an FPD, it is necessary to synchronize the X-ray irradiation timing between the FPD (or a personal computer (PC) or a relay apparatus connected to the FPD) and the X-ray generator. It becomes. For this reason, communication is performed between the FPD and the X-ray generation apparatus, and the FPD is configured to transfer the acquired X-ray image data to an image processing apparatus such as a PC for image processing or storage. Yes.

  Generally, an FPD, an image processing apparatus, and an X-ray generation apparatus that emits X-rays are connected using a cable or the like via an interface necessary for communication with each other. In addition, for the purpose of improving the ease of installation and the freedom of handling and photographing, there are many cases in which connection is made by a wireless interface such as a wireless LAN.

  Furthermore, by enabling the FPD itself to detect the start of X-ray irradiation from the X-ray generator, there is no need for synchronous communication between the FPD and the X-ray generator, which further improves installation and operability. A line imaging device or the like is also realized. Patent Documents 1 and 2 describe such an X-ray imaging apparatus, in which each row of the FPD is sequentially selected to perform detection scanning for switching between an on state and an off state, and a change in current flowing in the apparatus is detected. By detecting this, the start of X-ray irradiation is detected.

  According to the detection method of the start of X-ray irradiation described in Patent Documents 1 and 2, detection for detecting the start of irradiation by the X-ray imaging apparatus detecting the start of irradiation after the X-ray irradiation is actually started. A time difference (detection delay) occurs until scanning is stopped. In the pixels in the row where the detection scan is executed in the time difference range, a part of the electric charge generated by the X-ray irradiation flows out. As a result, a linear or wedge-shaped defect ( Hereinafter, this phenomenon may be referred to as detection delay artifact). As a representative example, an example of a wedge-shaped detection delay artifact is shown in FIG. FIG. 5A shows an example of an image in the case where imaging is executed by automatically detecting the start of X-ray irradiation. In FIG. 5A, a row (X-ray detection row) in which the X-ray imaging apparatus detects the start of irradiation from a row (X-ray irradiation start row 502) in which detection scanning is performed at the time when X-ray irradiation is actually started. 501) indicates that the pixel value is low for each row. FIG. 5B shows the y-axis profile of FIG. A pixel value indicated by a dotted line in FIG. 5B indicates a pixel value that can be originally obtained when there is no detection delay artifact.

  In Patent Documents 1 and 2, as one of countermeasures against detection delay artifacts, correction of detection delay artifacts generated in an X-ray image is performed using the value of the current used to detect the start of X-ray irradiation. Has proposed.

JP2011-247605 JP2011-249891A

  On the other hand, gain correction processing is performed on the acquired X-ray image in the image processing after acquiring the X-ray image. This is due to variations in the gain of each pixel caused by unevenness due to the film formation process of the X-ray detection sensor, the generation of pixels with different characteristics that occur specifically, aging and burn-in that occur with use, etc. This is for correction. As correction data used in gain correction, a gain image obtained by performing imaging by irradiating X-rays without placing a subject is used. The gain correction is performed by dividing the X-ray image after the offset correction (here, the correction process for removing or reducing the dark current component) by the gain image, and adding an appropriate value such as an average value of the entire gain image to the obtained X-ray image. This is done by multiplying the various coefficients.

  However, when a gain image is captured by the above-described method of detecting the start of X-ray irradiation, detection delay artifacts similarly occur in the gain image. Further, when the detection delay artifact of the gain image is corrected by some image processing, the pixel in which the detection delay artifact has occurred cannot correctly represent the original gain, and thus the gain correction may not be performed correctly. However, if a gain image is acquired by connecting a conventional X-ray imaging apparatus and an X-ray generation apparatus and performing imaging by performing synchronous communication, the detection delay artifact described above does not occur, but it is contrary to the advantage that connection is unnecessary. Resulting in. In addition, there is a situation where it is desired to use in combination with an X-ray generator without such a connection part in the first place. In such a case, an imaging method for detecting the start of X-ray irradiation, and a gain image using the imaging method are used. It is difficult to create both.

  The present invention has been made in view of the above-described problems, and has an object to obtain a gain image with high accuracy using an image captured by automatically detecting the start of radiation irradiation. To do.

In order to achieve the above object, a radiation imaging apparatus according to a position aspect of the present invention comprises the following arrangement. That is,
An imaging means for detecting the start of radiation irradiation based on the radiation dose detected by the radiation detector and capturing a radiation image using the radiation detector;
An acquisition means for acquiring a plurality of base images by irradiating with radiation in a state in which no subject exists and performing imaging by the imaging means;
A discriminating means for discriminating an artifact region caused by a time difference from the actual irradiation start to the detection of the irradiation start for the plurality of base images
Creation means for creating a gain image for gain correction by synthesizing the plurality of base images without using image information of the artifact region determined by the determination means.

  According to the present invention, it is possible to obtain a gain image with high accuracy using an image captured by automatically detecting the start of radiation irradiation.

The figure which shows the apparatus structural example of the X-ray imaging system by embodiment. 6 is a flowchart illustrating X-ray image capturing processing according to the embodiment. 6 is a flowchart illustrating a gain image creation process according to the embodiment. The flowchart which shows the process which discriminate | determines a detection delay artifact line. The flowchart which shows a normalization process. The figure explaining a detection delay artifact line. The figure explaining the interpolation process of the profile in a detection delay artifact line. The figure explaining creation of the gain image by the synthesis | combination of a normalization image.

  Hereinafter, an example of a preferred embodiment of the present invention will be described with reference to the accompanying drawings.

  The radiation imaging system of the present embodiment corrects an artifact region caused by radiation detection delay and gain correction for a radiation image obtained by imaging using a radiation detector composed of a plurality of pixels. The gain image used for the gain correction is created so as to reduce the influence of the artifact region caused by the radiation detection delay, thereby realizing highly accurate gain correction. In the following, an embodiment in which X-rays are used as radiation will be described, but the present invention can also be applied to radiation imaging apparatuses other than X-rays.

  FIG. 1 is a diagram illustrating a configuration example of the X-ray imaging system according to the present embodiment. An X-ray generation apparatus 101 as a radiation generation apparatus irradiates X-rays toward an FPD 102 as a radiation detector. The FPD 102 and the image processing apparatus 103 constitute an X-ray imaging apparatus, and the FPD 102 transmits an image signal corresponding to the detected X-ray dose to the connected image processing apparatus 103.

  The image processing apparatus 103 stores an I / O unit 106 that transmits and receives data, a CPU 107 (central processing unit) that controls various operations of the image processing apparatus 103, a program executed by the CPU 107, and reads and writes various data and the like. The memory 108 is provided. The image processing apparatus 103 includes a storage medium 109 for recording / saving image data and the like. Further, the image processing apparatus 103 is connected to a display apparatus 104 that displays processing results, images, and the like, and an operation apparatus 105 that receives user operations.

  An X-ray imaging process in the X-ray imaging system having the above configuration will be described. FIG. 2A is a flowchart for explaining the X-ray imaging process according to the present embodiment. In step S <b> 201, the CPU 107 detects the start of X-ray irradiation based on the radiation dose (X-ray dose) detected by the FPD 102. In the present embodiment, the CPU 107 detects the start of radiation irradiation based on the X-ray dose sequentially read from the FPD 102 in units of rows. When the start of irradiation is detected, the CPU 107 starts capturing an X-ray image using the FPD 102. In imaging, the FPD 102 sends X-ray image data (hereinafter referred to as an original image) obtained by converting the X-ray dose received by each internal X detection element into a digital signal to the I / O unit 106 of the image processing apparatus 103. In step S203, the CPU 107 uses, for example, a method described in Patent Documents 1 and 2 to indicate an artifact region (detection delay artifact) caused by a time difference from the actual start of irradiation to the detection of the start of irradiation in the original image. (Region) image information is corrected. Thereafter, the CPU 107 performs offset correction for removing the dark current component from the original image after the artifact correction in S203 (S204), and performs gain correction using the gain image for the original image after the offset correction (S205).

  Next, a process for creating a gain image used in S205 will be described. The outline of the operation in the X-ray imaging apparatus of this embodiment when creating a gain image is as follows.

  The operator uses the operation device 105 to instruct the creation of a gain image and inputs an operation for starting shooting. This operation refers to an operation for exchanging information for preparing for shooting between the FPD 102 and the image processing apparatus 103. Here, the information for preparation for photographing indicates that the image processing apparatus is informed that the FPD 102 is ready for photographing. The state in which the FPD 102 can be photographed means that the initialization process of the embedded software has been completed, and that the standby until the hardware becomes stable from the unstable state when the power is turned on has been completed. Etc.

  The X-ray generation apparatus 101 irradiates the FPD 102 with radiation without a shielding object (subject). Upon receiving the X-ray, the FPD 102 detects the start of X-ray irradiation and images image data (original image) obtained by converting the X-ray dose received by each internal X detection element into a digital signal, as in S201 and S202. The data is sent to the I / O unit 106 of the processing device 103. The image processing apparatus 103 stores the received original image in the memory 108 or the storage medium 109. The internal processing of the image processing apparatus 103 is realized by the CPU 107. The image processing apparatus 103 creates a gain image from the received original image (details of the process for creating the gain image will be described later with reference to the flowchart in FIG. 2B), and sends the processing result to the display apparatus 104. Here, examples of display of the processing result include displaying a message indicating that the gain image has been successfully created, displaying the created gain image, and the like.

  Next, the flow of processing for creating a gain image will be described in detail with reference to the flowchart of FIG. 2B. The process of creating a gain image includes a base image acquisition process (S210) for creating a plurality of captured images (hereinafter referred to as a base image) as a basis of the gain image, and a gain image acquisition process for creating a gain image from the plurality of base images. (S220).

  In the base image acquisition process (S210), the base image is repeatedly generated until the necessary number of base images for gain image generation can be acquired. The plurality of base images necessary for creating the gain image must include two base images in which the detection delay artifact regions do not overlap. When it is desired to reduce noise components such as quantum noise and noise on an electric circuit, it is preferable to create three or more base images and use them for gain image creation. The presence or absence of overlapping detection delay artifact regions can be determined based on, for example, the position of a row in the image when the FPD 102 detects the start of X-ray irradiation.

  In this case, for example, a row in which the start of irradiation is detected and a predetermined number of rows in which detection scanning immediately before the detection start may be determined as detection delay artifact regions. In this case, the predetermined number can be obtained from, for example, [time per line that requires scanning for detection] × [time difference from the start of actual irradiation until irradiation is detected]. Here, the actual irradiation start time may be acquired by the image processing apparatus 103 from the X-ray generation apparatus 101. For example, the clocks of the X-ray generation apparatus 101 and the image processing apparatus 103 are synchronized with each other, and information on the time when irradiation is started is acquired from the X-ray generation apparatus. In this case, communication between the X-ray generation apparatus 101 and the image processing apparatus 103 is performed. Even in this case, communication between the X-ray generator 101 and the FPD 102 is unnecessary. Further, the image processing apparatus 103 receives the time when the irradiation is detected from the FPD 102. By calculating the difference between these two times, the image processing apparatus 103 calculates the above-described time difference. Alternatively, the base image captured on the display device 104 may be displayed on the display device 104 so that the user can determine whether the artifact regions overlap.

  The contents of the basic image acquisition process (S210) will be described. In step S <b> 211, the image processing apparatus 103 performs X-ray irradiation without a shielding object, and acquires a plurality of original images in which detection delay artifact regions do not overlap. Processing for automatically detecting the start of X-ray irradiation and performing imaging is as described above in S201 and S202. Next, in S212, the CPU 107 performs an offset correction process in order to remove the dark current component from the original image. For example, after obtaining each original image, an offset correction process is performed by reading a dark current image from the FPD 102 in a state where X-rays are not irradiated and subtracting the dark current image from the original image. The offset correction process is not limited to this. For example, the offset correction process may be performed by reading a dark current image before photographing and subtracting it from the original image. In this way, a plurality of base images taken without a subject are acquired.

  In the gain image acquisition process (S220), a gain image is created using two or more base images in which detection delay artifacts created by the process described above do not overlap. Hereinafter, the gain image creation processing S220 will be described.

  First, in S221, the CPU 107 performs a discrimination process for discriminating detection delay artifact rows that constitute a detection delay artifact region for each base image acquired in the base image acquisition process (S210). In the detection delay artifact row determination process, it is determined whether or not the detection delay artifact is generated for each row of each base image. Details of the discrimination processing will be described later. Next, in S222, the CPU 107 performs a normalization process using each base image and the detection delay artifact row determination result to create a normalized image. This normalization process is performed to normalize the X-ray dose when each base image is taken. In step S223, the CPU 107 creates a gain image using the result of the detection delay artifact row discrimination process (S221) and the normalized image (S212).

  Next, the detection delay artifact line discrimination process (S221) will be described in detail. FIG. 3 is a flowchart showing the detection delay artifact line discrimination process (S221). As shown in FIG. 3, the determination process includes a profile creation process (S301), an approximate expression creation process (S302), and a flag creation process (S303) for each base image. Note that the result of the detection delay artifact line determination process is used in a normalization process (S222) and a gain image generation process (S223), which are subsequent processes.

  First, in the profile creation process (S301), the CPU 107 creates a profile of the feature amount calculated for each row in the base image. In this embodiment, a profile of an average value of pixel values (luminance values) of pixels arranged in the row direction is created. Note that the feature amount is not limited to the average value of the pixel values in the row, and may be another feature amount. For example, there are feature values obtained from a median or a frequency graph (histogram). As an example of the feature amount obtained from the histogram, there is an average value or a median value after excluding 1% of the frequency having the larger value and the smaller value. Next, in the approximate expression creation process (S302), the CPU 107 creates an approximate expression representing the distribution of the X-ray dose input to the X-ray detector from the profile created in S301. Specifically, the CPU 107 obtains an approximate expression by fitting a model such as a quadratic expression or a linear expression to the profile created in S301 and using a least square method or the like. At this time, since the detection delay artifact row has a value that does not follow the X-ray dose distribution, it is preferable to use a method that can weight the error, such as a robust estimation method.

  Next, in the flag creation process (S303), the CPU 107 determines whether each row of the base image is a detection delay artifact row, and records the determination result as a flag. In this embodiment, the CPU 107 compares the approximate expression derived by the approximate expression creation process (S302) with the row average value profile created by the profile creation process (S301) for each line. Then, the CPU 107 determines whether or not it is a detection delay artifact based on the comparison result. For example, when the row average value is equal to or less than a predetermined ratio with respect to the value represented by the approximate expression, the row is determined as a detection delay artifact row. The CPU 107 stores the flag = 1 in association with each row of the base image, for example, if the artifact row, and flag = 0 otherwise.

  In the above-described determination process, an example using an approximate expression of a profile has been described, but the present invention is not limited to this. For example, detection of detection delay artifact rows as described in the base image acquisition process (S210) may be used. That is, a predetermined number of lines immediately before the line in which the irradiation start is detected and the reading order in the detection scanning may be determined as detection delay artifact lines.

  Next, the normalization process (S222) of FIG. 2 will be specifically described. FIG. 4 is a flowchart showing the flow of normalization processing (S222). In this normalization process, a profile creation process (S401), an interpolation process (S402), an average value calculation process S223, and a normalized image creation process (S404) are performed on each base image. In the normalization process (S222), normalization is performed using a value that eliminates the influence of detection delay artifacts, thereby eliminating the influence of variations in the X-ray dose irradiated from the X-ray generation apparatus 101 in a plurality of base images. To be done.

  First, in the profile creation process (S401), the CPU 107 creates a profile of feature quantities in units of lines. In the present embodiment, the CPU 107 creates a profile of the average value in the row direction of the base image. Since this process is the same as the profile creation process (S301) in the detection delay artifact line discrimination process, the process result of S301 may be used.

  In the interpolation process (S402), the feature amount of the detection delay artifact region is determined by interpolation based on the feature amount of the region other than the detection delay artifact region. That is, the CPU 107 uses the row average value other than the row determined to be the detection delay artifact row in the row average value profile obtained in the profile creation process (S401), to determine the row determined to be the detection delay artifact row. The row average value is determined by interpolation. In this way, an interpolation average value profile based on the average value of the pixel values of each row other than the detection delay artifact region and the interpolation value of the row of the detection delay artifact region is created. For example, as shown in FIG. 6, an approximate expression is created from the profile of line average values excluding detection delay artifact lines, and the artifact lines are interpolated using the values of the approximate expressions. The approximate expression may be, for example, a linear expression or a secondary expression.

  Next, in the average value calculation process (S403), the CPU 107 calculates the average value of the entire profile from the interpolation average value profile obtained in the interpolation process (S402). The average value of the profile obtained here can be handled as an average value that is not affected by the detection delay artifact row of the base image.

  Next, in the normalized image creation process (S404), the CPU 107 normalizes the base image using the average value calculated in the average value calculation process (S403) to create a normalized image. For example, the CPU 107 obtains a normalized image by dividing the pixel value of each pixel of the base image by the average value calculated in the average value calculation process (S403).

  In the gain image creation process (S223) of FIG. 2B, a gain image is created using a plurality of normalized images created as described above. In the present embodiment, the CPU 107 performs gain correction by combining a plurality of normalized images without using the image information of the detection delay artifact row determined in the detection delay artifact row determination processing (S221). Create a gain image. For example, the CPU 107 creates a gain image by calculating an average value of pixel values of the same coordinates in each base image (normalized image) created in the normalization process (S222) as a pixel value of the gain image. At this time, the CPU 107 does not include the pixel value of the row determined to be the detection delay artifact row by the determination processing (S221) in the calculation of the average value.

  FIG. 7 is a diagram for explaining a gain image creation process. Assume that a first normalized image 701 and a second normalized image 721 are obtained as normalized images that are combined to create a gain image. A detection delay artifact region 702 is detected in the first normalized image 701, and a detection delay artifact region 722 is detected in the second normalized image 721. When the normalized image is combined to create the gain image 741, the average value of pixels that do not belong to the detection delay artifact region is calculated and used as the pixel value of the gain image. For example, the pixel 703 of the first normalized image 701 and the pixel 723 of the second normalized image 721 are both pixels having coordinates (x1, y1), and the average value of these pixel values is the gain image 741. This is the pixel value of the pixel 743 at coordinates (x1, y1). On the other hand, in the case of the pixel 704 and the pixel 724 at the coordinates (x2, y2), the pixel 704 belongs to the detection delay artifact region and is not used for calculating the average value. Therefore, in this example, the pixel value of the pixel 724 becomes the pixel value of the pixel 744 in the gain image 741 as it is. In the case of the pixel 705 and the pixel 725 at the coordinates (x3, y3), the pixel 725 belongs to the detection delay artifact region and thus is not used for calculating the average value. The pixel value of the pixel 705 is directly used as the pixel 745 in the gain image 741. Of the pixel value.

  As described above, according to the present embodiment, the created gain image can accurately represent the original gain, and a gain image that can appropriately perform gain correction even when detection delay artifact occurs is obtained. be able to.

  The present invention can also be realized by executing the following processing. That is, software (program) that realizes the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media, and a computer (or CPU, MPU, or the like) of the system or apparatus reads the program. It is a process to be executed.

101 X-ray generation device, 102 FPD, 103 image processing device, 104 display device, 105 operation device, 106 I / O unit, 107 CPU, 108 memory, 109 storage medium

Claims (10)

A radiography apparatus,
An imaging means for detecting the start of radiation irradiation based on the radiation dose detected by the radiation detector and capturing a radiation image using the radiation detector;
An acquisition means for acquiring a plurality of base images by irradiating with radiation in a state in which no subject exists and performing imaging by the imaging means;
A discriminating means for discriminating an artifact region caused by a time difference from the actual irradiation start to the detection of the irradiation start for the plurality of base images,
A radiation imaging apparatus comprising: a creation unit that creates a gain image for gain correction by combining the plurality of base images without using image information of the artifact region determined by the determination unit .   The radiographic apparatus according to claim 1, wherein the acquisition unit acquires the plurality of base images so as to include at least two base images that do not overlap an artifact region caused by the time difference.   The radiographic apparatus according to claim 2, wherein the acquisition unit determines whether or not an artifact region overlaps based on a position of a row where the irradiation start is detected in each of the plurality of base images.   The determination unit obtains an approximate expression that represents a profile of a feature amount calculated in units of rows for each of the plurality of base images, and the feature amount of a row with respect to the feature amount indicated by the approximate expression is equal to or less than a predetermined ratio. The radiation imaging apparatus according to claim 1, wherein the line is determined as a detection delay artifact line. The imaging means detects the start of radiation irradiation based on the radiation dose sequentially read from the radiation detector in line units,
4. The method according to claim 1, wherein the discriminating unit sets, as the artifact region, a region including a row in which the irradiation start is detected and a predetermined number of rows immediately before the row in the reading order. The radiographic apparatus according to the item. Interpolating means for interpolating the feature amount of the artifact region based on the feature amount of the region other than the artifact region;
Normalization means for normalizing image information of the base image based on the feature amount of the artifact region obtained by the interpolation means and the feature amount of the region other than the artifact region in each of the plurality of base images; Further comprising
The radiographic apparatus according to claim 1, wherein the creating unit creates the gain image using the normalized plurality of base images.   The interpolating unit obtains an approximate expression of a feature amount profile calculated in units of rows for rows excluding the row belonging to the artifact region, and interpolates the feature amount of the row belonging to the artifact region using the approximate equation. The radiation imaging apparatus according to claim 6.   The said creation means uses the average value of the image information obtained from areas other than the artifact areas of each of the plurality of base images as the image information of the gain image. The radiation imaging apparatus described in 1. A gain image creation method by a radiation imaging apparatus having an imaging means for detecting radiation start based on a radiation dose detected by a radiation detector and imaging a radiation image using the radiation detector,
An acquisition step of acquiring a plurality of base images by irradiating with radiation in a state where no subject exists, and performing imaging by the imaging unit;
For the plurality of base images, a determination step of determining an artifact region caused due to a time difference from the actual irradiation start to the detection of the irradiation start,
Creating a gain image, comprising: creating a gain image for gain correction by combining the plurality of base images without using the image information of the artifact region discriminated in the discrimination step Method.
  The program for making a computer perform each process of the gain image production method described in Claim 9.