JP5179946B2 - Radiation image processing method and apparatus - Google Patents

Description

  The present invention relates to a radiographic image processing method and apparatus for correcting defective pixels in an image photographed by a radiation detector.

  2. Description of the Related Art Conventionally, in the medical field and the like, various types of radiation detectors that record a radiation image related to a subject by irradiation with radiation that has passed through the subject have been proposed and put into practical use.

  As the radiation detector, for example, as described in Patent Document 1, there is a radiation detector using a semiconductor such as amorphous selenium that generates a charge by irradiation of radiation, and as such a radiation detector, a so-called radiation detector is used. An optical reading type and a TFT reading type have been proposed.

  If such a radiation detector is used, image information obtained by imaging can be acquired as digital data, so that affinity with a diagnosis support apparatus using a computer can be increased.

Since the radiation detector as described above may include a defective pixel, for example, as described in Patent Document 2, in order to prevent deterioration of image quality due to the defective pixel, the position of the defective pixel is previously determined. Is obtained by measurement, and the pixel value detected by the defective pixel is corrected. At this time, defective pixel information is obtained when the radiation detector is irradiated with uniform radiation, and defective pixel positions are acquired and correction conditions are calculated using the defective pixel information.
JP 2000-105297 A JP 2007-129339 A

  By the way, a defective pixel generated in an image captured by the radiation detector as described above has a longer accumulation time in the radiation detector or a larger amount of radiation irradiation to the radiation detector, thereby causing a defective pixel as a defective pixel. There are deformed defective pixels in which the recognized area expands to the periphery, and non-deformed defective pixels in which the area recognized as a defective pixel is constant regardless of the accumulation time in the radiation detector and the radiation dose to the radiation detector.

  Deformed defective pixels often occur particularly in radiation detectors having a photoconductive layer mainly composed of amorphous selenium. This is because some of the amorphous selenium in the photoconductive layer is crystallized, and if the accumulation time in the radiation detector becomes long or the radiation dose to the radiation detector increases, it will be recognized as a defective pixel starting from the crystallization part. This is because the area of the defective pixel group to be expanded to the periphery.

  For non-deformed defective pixels, for example, in a TFT reading type radiation detector, if a part of TFT constituting each pixel becomes nonfunctional due to disconnection or the like, the part becomes a dead pixel and can detect radiation at all. It becomes impossible. This dead pixel area is always constant regardless of the accumulation time in the radiation detector and the amount of radiation applied to the radiation detector. The region recognized as a defective pixel does not expand to the periphery due to an increase in the amount of radiation applied to the vessel.

  As described above, there are two types of defective pixels in an image photographed by the radiation detector, that is, a deformed defective pixel and a non-deformed defective pixel. For these deformed defective pixels, defect pixel information is acquired in advance. Even if correction is performed for a defective pixel in an image captured by the radiation detector, the accumulation time in the radiation detector and / or the amount of radiation applied to the radiation detector at the time of actual imaging is determined as defective pixel information. If the accumulation time in the radiation detector and / or the amount of radiation applied to the radiation detector is different from that obtained when the image data is acquired, there is a problem that the defective pixel cannot be corrected correctly.

  Therefore, a plurality of defective pixel information is acquired and stored in the database for each accumulation time in the radiation detector and / or for each radiation dose to the radiation detector, and the database is searched for each defective pixel in the normal image. Then, it is conceivable to perform correction based on the accumulation time in the radiation detector and / or the radiation dose to the radiation detector when the image is actually taken. In such an embodiment, a database is provided for each defective pixel. Therefore, there is a problem that it takes a lot of time for image correction.

  The present invention has been made in view of the above circumstances, and in a radiation image processing method and apparatus for correcting a defective pixel in an image photographed by a radiation detector, a deformed defective pixel and an undeformed defect in the image An object of the present invention is to provide a radiation image processing method and apparatus capable of shortening the correction processing time when there are two types of pixels.

  According to the first radiation image processing method of the present invention, radiation detection is performed on defective pixel information based on a uniform irradiation image obtained by uniformly irradiating radiation to a radiation detector or an unexposed image. Stored in the database for each accumulation time in the detector and / or for each radiation dose to the radiation detector, and in the defective pixel information in the normal image obtained by detecting the radiation transmitted through the subject using the radiation detector The radiation dose at the position of the radiation detector corresponding to the position of the deformed defective pixel whose size changes according to the accumulation time in the radiation detector and / or the radiation dose to the radiation detector is determined in the normal image For each deformed defective pixel, the default is based on the accumulation time in the radiation detector and / or the radiation dose to the radiation detector. The corresponding defective pixel information is extracted from the database, the image data corresponding to each deformed defective pixel in the normal image is corrected based on the extracted defective pixel information, and the accumulation time in the radiation detector in the normal image and / or The image data corresponding to each non-deformed defective pixel in the normal image is corrected by predetermined image processing for each non-deformed defective pixel whose size does not change according to the radiation dose to the radiation detector. Is the method.

  In addition, the second radiation image processing method of the present invention is configured to obtain defective pixel information based on a uniform irradiation image obtained by uniformly irradiating radiation to a radiation detector or an unexposed image. The size depends on the accumulation time in the radiation detector and / or the radiation dose to the radiation detector in the normal image that is stored in the database and detected by detecting the radiation that has passed through the subject using the radiation detector. For the deformed defective pixel that changes, the image data corresponding to each deformed defective pixel in the normal image is based on the revised defective pixel information obtained by uniformly expanding the area of each deformed defective pixel in the defective pixel information stored in the database. Correction is performed and the size does not change according to the accumulation time and / or radiation dose to the radiation detector in the normal image For each form defective pixel is a method characterized by correcting the image data corresponding to each non-deformed defective pixels during normal image by predetermined image processing.

  Note that uniformly enlarging means a process of enlarging one or a plurality of pixels around a defective pixel serving as a reference.

  As described above, deformation defect pixels often occur particularly in a radiation detector having a photoconductive layer mainly composed of amorphous selenium. This is because some of the amorphous selenium in the photoconductive layer is crystallized, and if the accumulation time in the radiation detector becomes long or the radiation dose to the radiation detector increases, it will be recognized as a defective pixel starting from the crystallization part. This is due to the reason that the area to be expanded around. In this case, when the radiation detector is irradiated with uniform radiation, many signals are output for the deformed defective pixel portion as compared with other normal pixels.

  On the other hand, many non-deformed defective pixels have become insensitive pixels and cannot detect radiation at all. In this case, no signal is output even if radiation is irradiated to the radiation detector. There is no longer to do.

  Therefore, in the radiographic image processing method of the present invention, a portion with a large pixel value is identified as a deformed defective pixel and a portion with a small pixel value is compared with general pixel data in defective pixel information. It is preferable to identify as

  When imaging the signal detected by the radiation detector, the higher the detection signal value, the lower the luminance, the lower the detection signal value, the higher the luminance, and the higher the detection signal value, the higher the luminance. There are two types of cases where the lower the detection signal value is, the lower the brightness is displayed. The “pixel value high part” here corresponds to the part of the defective pixel information where the detection signal value at the radiation detector is high. The “part with a small pixel value” means a part corresponding to a part with a low detection signal value in the radiation detector in the defective pixel information.

  The first radiation image processing apparatus of the present invention is based on a radiation detector and a uniform irradiation image or an unirradiated image acquired by uniformly irradiating the radiation detector with radiation. In a database storing defect pixel information for each accumulation time in the radiation detector and / or for each radiation dose to the radiation detector, and in an image acquired by the radiation detector, the accumulation time and / or radiation in the radiation detector. The position of the deformed defect pixel whose size changes according to the radiation dose to the detector and the position of the non-deformed defect pixel whose size does not change according to the accumulation time in the radiation detector and / or the radiation dose to the radiation detector. It is acquired by detecting the radiation that has passed through the subject using the defective pixel identification means to identify and the radiation detector. Radiation dose specifying means for specifying the radiation dose at the position of the radiation detector corresponding to the position of the deformed defect pixel in the normal image, and for each deformed defect pixel in the normal image, the accumulation time in the radiation detector and / or Based on the radiation dose to the radiation detector, the corresponding defective pixel information is extracted from the database, the image data corresponding to the deformed defective pixel is corrected based on the extracted defective pixel information, and each non-deformed in the normal image The defective pixel is characterized by comprising correction means for correcting image data corresponding to the non-deformed defective pixel by predetermined image processing.

  The second radiation image processing apparatus of the present invention is based on a radiation detector and a uniform irradiation image or an unexposed image acquired by uniformly irradiating the radiation detector with radiation. The position of the deformed defective pixel whose size changes according to the accumulation time in the radiation detector and / or the radiation dose to the radiation detector in the database storing the defective pixel information and the image acquired by the radiation detector; and A defect pixel specifying means for specifying a position of an undeformed defective pixel whose size does not change according to an accumulation time in the radiation detector and / or a radiation dose to the radiation detector, and a database for each deformed defective pixel in a normal image. Deformation defect image based on revised defect pixel information that uniformly enlarges the area of each deformation defect pixel in the stored defect pixel information Correction means for correcting the image data corresponding to the non-deformed defective pixel by a predetermined image processing for each non-deformed defective pixel in the normal image. To do.

  In the radiological image processing apparatus of the present invention, the defective pixel specifying means identifies a portion having a large pixel value as a deformed defective pixel and compares a portion having a small pixel value as compared with general pixel data in defective pixel information. Preferably, the pixel is identified as a deformed defective pixel.

  The database in the first radiographic image processing apparatus may store only information on representative deformed defective pixels for each accumulation time in the radiation detector and / or for each radiation dose to the radiation detector. preferable.

  The radiation detector of the radiation image processing method and apparatus described above is an image signal that represents a radiation image related to a subject by converting incident radiation directly or once into light and then converting it into light and outputting the charge to the outside. A solid state detector capable of obtaining the above can be used.

  There are various types of solid-state detectors. For example, from the aspect of the charge generation process that converts radiation into electric charge, the photoconductive layer detects fluorescence emitted from the phosphor when irradiated with radiation. The signal charge obtained in this way is temporarily stored in the power storage unit, the stored charge is converted into an image signal (electrical signal) and output, or the photoconductive layer is irradiated with radiation. The signal charge generated in step 1 is collected by the charge collection electrode, temporarily stored in the power storage unit, and the stored charge is converted into an electrical signal and output, or the direct conversion type solid state detector that reads the stored charge to the outside From the aspect of the reading process, a TFT reading method that scans and reads a TFT (thin film transistor) connected to the power storage unit, or an optical reading method that reads the reading light (electromagnetic wave for reading) by irradiating the detector with the reading light. Such as those of news, there is such an improved direct conversion type which is proposed in the Patent Document 1 filed by the present applicant that combines the direct conversion type and the optical readout type.

  In the first radiographic image processing method and apparatus, the defective pixel information stored in the database is actually imaged for each accumulation time in the radiation detector and / or for each radiation dose to the radiation detector. In addition to the mode acquired by the above, the process of enlarging the periphery of the deformed defective pixel serving as a reference with reference to certain defective pixel information as a reference in consideration of the accumulation time and the dose of radiation, and the deformed defective pixel as described above A plurality of pieces of defective pixel information may be generated by performing a process of uniformly expanding the deformed defective pixel region for every change in the accumulation time or the radiation dose assuming that the region is enlarged.

In particular, when acquiring defective pixel information based on an unexposed image, paying attention to the accumulation time, for example, the circumference of the reference deformed defective pixel is expanded by one round until the accumulation time is 1 second, and thereafter every second. A plurality of pieces of defective pixel information may be generated by a process of enlarging one round at a time. (One-round enlargement means a process of enlarging one pixel around the deformation defective pixel group, such as one pixel enlargement in the upper, lower, left, and right directions of the reference deformation defective pixel, or one pixel expansion in eight directions.)
In the first radiographic image processing method and apparatus, the radiation dose at the position corresponding to the deformed defective pixel position in the defective pixel information is specified in the normal image, and for each deformed defective pixel in the normal image, Corresponding defective pixel information is extracted from the database based on the accumulation time in the radiation detector and / or the radiation dose to the radiation detector, and corresponding to each deformed defective pixel in the normal image based on the extracted defective pixel information. Although the image data is corrected, the present invention is not limited to a mode in which defective pixel information that completely matches the accumulation time and the specified radiation dose is extracted from the database, but some defective pixel information (for example, the accumulation time and the specified radiation dose are close to each other) When defective data is stored as described above based on the defective pixel information extracted at the stage of image data correction. The process of enlarging the periphery of the reference deformation defect pixel as a defective pixel in consideration of the radiation dose and the expansion of the deformation defect pixel area as described above is uniform for every change in accumulation time and radiation dose. The correction area may be adjusted by performing a process for enlarging the deformation defective pixel area.

  The radiation dose at the position corresponding to the deformed defective pixel position in the defective pixel information may be identified by estimation based on, for example, the radiation detection amount of pixels around the deformed defective pixel in the normal image. Alternatively, it may be specified uniformly from imaging conditions such as radiation intensity and accumulation time.

  According to the first radiographic image processing method and apparatus of the present invention, in the radiographic image processing method and apparatus for correcting defective pixels in an image photographed by a radiation detector, the radiation detector is uniformly irradiated. The defective pixel information acquired by detecting the emitted radiation is stored in a database for each accumulation time in the radiation detector and / or for each radiation dose to the radiation detector, and transmitted through the subject using the radiation detector. Corresponds to the position of deformed defective pixels whose size changes according to the accumulation time in the radiation detector in the defective pixel information and / or the radiation dose to the radiation detector in the normal image acquired by detecting radiation The radiation dose at the position of the radiation detector to be identified is specified, and the radiation for each deformation defect pixel in the normal image is determined. Based on the accumulation time in the output device and / or the radiation dose to the radiation detector, the corresponding defective pixel information is extracted from the database, and the image corresponding to each deformed defective pixel in the normal image based on the extracted defective pixel information For each non-deformed defective pixel whose size does not change according to the accumulation time in the radiation detector in the normal image and / or the radiation dose to the radiation detector in the normal image, the data is corrected by predetermined image processing. Image data corresponding to each non-deformed defective pixel is corrected, and when non-deformed defective pixels in a normal image are corrected, correction is performed by predetermined image processing without referring to a database. Therefore, when there are two types of defective pixels and non-deformed defective pixels in the image, the correction processing time can be shortened. .

  According to the second radiographic image processing method and apparatus of the present invention, in the radiographic image processing method and apparatus for correcting defective pixels in an image captured by the radiation detector, radiation is applied to the radiation detector. Defective pixel information based on uniform irradiation images acquired by uniform irradiation or non-irradiation images is stored in a database and acquired by detecting radiation that has passed through the subject using a radiation detector For each of the deformed defective pixels whose size changes in accordance with the accumulation time in the radiation detector and / or the radiation dose to the radiation detector in the normal image thus obtained, each deformed defective pixel in the defective pixel information stored in the database is displayed. The image data corresponding to each deformed defective pixel in the normal image is corrected based on the revised defective pixel information obtained by uniformly expanding the area, and Corresponding to each non-deformed defective pixel in the normal image by predetermined image processing for each non-deformed defective pixel whose size does not change according to the accumulation time in the radiation detector in the image and / or the radiation dose to the radiation detector When correcting non-deformed defective pixels in a normal image by correcting the image data to be processed, predetermined image processing is performed without performing processing for uniformly expanding the area of each defective pixel in the defective pixel information. Therefore, if there are two types of deformation defective pixels and non-deformation defective pixels in the image, the correction processing time can be shortened.

  In the radiological image processing method and apparatus, a portion having a large pixel value is identified as a deformed defective pixel and a portion having a small pixel value is identified as a non-deformed defective pixel as compared with general pixel data in defective pixel information. As a result, it is possible to appropriately distinguish the deformation defective pixel from the non-deformation defective pixel.

  Further, in the first radiographic image processing apparatus, the database stores only information on representative deformation defective pixels for each accumulation time in the radiation detector and / or for each radiation dose to the radiation detector. As a result, the amount of data stored in the database can be reduced.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a schematic side view showing a radiographic image capturing system according to a first embodiment to which the radiographic image processing method of the present invention is applied, and FIG. 2 is a schematic diagram of a solid state detector used in the radiographic apparatus of the system. is there.

  The radiation imaging system 1 includes a radiation irradiation device 2 including a radiation source, a radiation imaging device 3 including a solid detector 20 (radiation detector) for detecting radiation, the radiation irradiation device 2, and the like. The computer 4 is connected to the radiation imaging apparatus 3 and the monitor 5 is connected to the computer 4.

  The radiation irradiation apparatus 2 includes a base 10, a support column 11 fixed to the base 10, and a radiation irradiation unit 12 containing a radiation source therein. It is attached to be movable in the direction (vertical direction in FIG. 1).

  In the radiation irradiating apparatus 2, the operation of the radiation source and the movement operation of the radiation irradiating unit 12 are all integrated and controlled by a control means (not shown).

  The radiation imaging apparatus 3 includes a base 13, a column 14 fixed to the base 13, a radiographic imaging unit 15 that houses a solid detector 20 therein, and the radiographic imaging unit 15 and a subject. 1 and a handrail 17 fixed to the base 13, and the radiographic image capturing unit 15 is attached to be movable in the longitudinal direction (vertical direction in FIG. 1) of the column 14.

  The system of the present embodiment can perform two types of shooting: normal shooting (only one shooting) and long shooting (image combining after shooting multiple times). It is configured so as to be detachable from the base 13, and when taking a long picture, a screen 16 is attached for photographing, and when not taking a long picture, the screen 16 is removed for photographing. In addition, a sensor for detecting the attachment / detachment state of the partition 16 is attached to the base 13.

  The solid state detector 20 is arranged inside the radiographic image capturing unit 15 so that its radiation detection surface is parallel to the radiation incident surface of the radiographic image capturing unit 15.

  As shown in FIG. 2, the solid state detector 20 includes a first conductive layer 24 made of a-Si TFT on a glass substrate 25, a photoconductivity that generates electric charge by receiving radiation and exhibits conductivity. The layer 23, the second conductive layer 22, and the insulating layer 21 are laminated in this order.

  The first conductive layer 24 is formed with a TFT corresponding to each pixel, the output of each TFT is connected to an IC chip 26, and the IC chip 26 is an image signal processing unit (not shown) on the printed circuit board 27. It is connected to the.

  When the solid state detector 20 irradiates the photoconductive layer 23 with radiation when an electric field is formed between the first conductive layer 24 and the second conductive layer 22, the solid state detector 20 enters the photoconductive layer 23. Charge pairs are generated, and latent image charges corresponding to the amount of the charge pairs are accumulated in the first conductive layer 24. When reading the accumulated latent image charge, the TFTs of the first conductive layer 24 are sequentially driven to output an analog signal corresponding to the latent image charge corresponding to each pixel, and this analog signal is processed by image signal processing. The detection is performed for each pixel in the unit, and the analog signals detected for each pixel are combined in the pixel arrangement order. The combined analog signal is AD converted by an AD converter (not shown) to generate a digital image signal. The generated digital image signal is transmitted from the image signal processing unit to the computer 4 via the memory.

  In the radiation imaging apparatus 3, the operation of the solid state detector 20 and the movement operation of the radiation image capturing unit 15 are all integrated and controlled by a control unit (not shown). The control means also has a function of notifying the computer 4 of the attachment / detachment state of the partition 16 on the base 13.

  The computer 4 detects defective pixel information based on the uniform irradiation image acquired by uniformly irradiating the solid detector 20 with radiation at every accumulation time in the solid detector 20 and the radiation with respect to the solid detector 20. A database DB for storing each dose is provided, and in the image acquired by the solid detector 20, the size changes according to the accumulation time in the solid detector 20 and / or the radiation dose to the solid detector 20. A function as a defective pixel specifying means for specifying the position of the deformed defective pixel and the position of the non-deformed defective pixel whose size does not change according to the accumulation time in the solid detector 20 and / or the radiation dose to the solid detector 20, The radiation dose at the position of the solid state detector 20 corresponding to the position of the deformation defect pixel in the normal image is specified. For each deformed defective pixel in the normal image and the function as the radiation dose specifying means, the corresponding defective pixel information is extracted from the database DB based on the accumulation time in the solid detector 20 and the radiation dose to the solid detector 20. The image data corresponding to the deformed defective pixel is corrected based on the extracted defective pixel information, and the image data corresponding to the non-deformed defective pixel is corrected by predetermined image processing for each non-deformed defective pixel in the normal image. It has a function as a correction means to perform.

  Here, the database DB will be described in detail with reference to the drawings. FIG. 3 is a diagram showing a state in which the size of a deformation defect pixel group in an image photographed by a solid-state detector changes depending on an accumulation time in the solid-state detector and a radiation dose to the solid-state detector, and FIG. 4 is a management structure of the database. FIG. 5 is a diagram showing an example of defective pixel information stored in the database.

  The defective pixel information necessary for correcting the defective pixel in the image captured by the solid state detector 20 is acquired by irradiating the solid state detector 20 with uniform radiation and performing the image capturing. Compare each pixel value in the defective pixel information obtained in this way with the average pixel value in the image and the intermediate value / average value of the pixels in the surrounding image area. A portion having a small pixel value (a portion displayed in white in the image) is identified as a non-deformed defective pixel.

  The defective pixel information stored in the database is not limited to that based on the uniform irradiation image acquired by uniformly irradiating radiation, but may be based on the non-irradiation image.

  As shown in FIG. 3, with respect to the deformation defect pixels generated in the solid state detector 20 used in the present embodiment, the accumulation time in the solid state detector 20 becomes long, or the radiation irradiation amount to the solid state detector 20 As a result of the increase, the area recognized as a defective pixel expands to the periphery. In FIG. 3, the change in the accumulation time in the radiation detector is shown in the horizontal direction, and the change in the radiation dose to the solid detector 20 is shown in the vertical direction.

  Therefore, even if defective pixel information is acquired in advance and correction is performed for the deformed defective pixel in the image captured by the solid state detector 20, the accumulation time and / or the solid state detector 20 when the actual image is captured are described. Alternatively, when the radiation dose to the solid detector 20 is different from the accumulation time in the solid detector 20 and / or the radiation dose to the solid detector 20 when the defective pixel information is acquired, the deformation defective pixel is corrected correctly. There is a problem that can not be done.

  In order to solve such a problem, in the present embodiment, as shown in FIG. 4, a plurality of defects photographed by changing the accumulation time in the solid-state detector 20 and the radiation dose to the solid-state detector 20 respectively. Pixel information RD (m, n) is acquired and stored in the database DB in the computer 4.

  An example of the defective pixel information RD (m, n) is shown as defective pixel information RD (1,1) taken with an accumulation time in the solid state detector 20 of 100 ms and a radiation dose to the solid state detector 20 of 0.1 mR. Is as shown in FIG. The defective pixel information RD (1,1) includes three deformed defective pixel groups BDPa (1,1), BDPb (1,1), BDDP (1,1), and two non-deformed defective pixel groups WDPa (1). , 1) and WDPb (1, 1).

  Defective pixel information RD (1,2) obtained by taking the accumulation time in the solid state detector 20 as 100 ms and the radiation dose to the solid state detector 20 as 0.5 mR is as shown in FIG. In the defective pixel information RD (1,2), three modified defective pixel groups BDPa (1,2), BDPb (1,2), BDPc (1,2) are located at the same position as the defective pixel information RD (1,1). However, in these defective pixel groups, an area recognized as a defective pixel is expanded to the periphery as compared with the deformed defective pixel group in the defective pixel information RD (1, 1). Similarly, in the defective pixel information RD (1,2), there are two non-deformed defective pixel groups WDPa (1,2) and WPb (1,2) at the same position as the defective pixel information RD (1,1). However, these non-deformed defective pixel groups do not change the areas recognized as defective pixels as compared to the non-deformed defective pixel groups in the defective pixel information RD (1, 1).

  Defective pixel information RD (1, 3) obtained by taking the accumulation time in the solid state detector 20 as 100 ms and the radiation dose to the solid state detector 20 as 1.0 mR is as shown in FIG. In the defective pixel information RD (1,3), three modified defective pixel groups BDPa (1,3), BDPb (1,3), BDPc (1,3) are located at the same position as the defective pixel information RD (1,1). 3), however, these defective pixel groups have a region that is recognized as a defective pixel further expanded to the periphery as compared with the deformed defective pixel group in the defective pixel information RD (1, 2). Similarly, in the defective pixel information RD (1, 3), there are two non-deformed defective pixel groups WDPa (1, 3) and WDDP (1, 3) at the same position as the defective pixel information RD (1, 1). However, these non-deformed defective pixel groups do not change the areas recognized as defective pixels as compared to the non-deformed defective pixel groups in the defective pixel information RD (1, 2).

  In the database DB of the present embodiment, the accumulation time in the solid state detector 20 is changed to 100 ms, 500 ms, 1000 ms, and 1500 ms, and the radiation dose to the solid state detector 20 is further set to 0.1 mR and 0.5 mR for each accumulation time. , 1.0 mR, 5.0 mR, and 10 mR, a total of 20 pieces of defective pixel information RD (m, n) are acquired.

  Next, the operation of the radiographic imaging system 1 configured as described above will be described. FIG. 6 is a diagram illustrating an example of an image photographed by the present system, and FIG. 7 is a diagram illustrating an example after correction of an image photographed by the present system.

  First, when acquiring defective pixel information, the photographer designates a predetermined irradiation radiation intensity, time, etc., and presses a radiation exposure switch (not shown) to start photographing.

  When imaging is started, imaging is performed by irradiating radiation from the radiation irradiation unit 12 toward the radiation image capturing unit 15.

  When radiation is irradiated, latent image charges carrying radiation image information are accumulated in the solid state detector 20. Since the amount of accumulated latent image charge is substantially proportional to the amount of radiation transmitted through the subject, this latent image charge carries an electrostatic latent image.

  After a predetermined time elapses, recording on the solid state detector 20 is stopped and photographing is ended, and then an analog signal corresponding to the latent image charge is output from the solid state detector 20 and an image obtained by AD conversion in the image signal processing unit. The defective pixel information signal RD (m, n) detected based on the above is generated. The generated defective pixel information signal RD (m, n) is transmitted from the image signal processing unit to the computer 4 via the memory.

  When the computer 4 receives the digital defective pixel information signal RD (m, n) from the radiation imaging apparatus 3, the computer 4 stores the digital defective pixel information signal RD (m, n) in the database DB for each accumulation time in the solid state detector 20 and for each radiation dose to the solid state detector 20. End shooting.

  In this embodiment, it is necessary to acquire 20 pieces of defective pixel information RD (m, n), but this operation is preferably performed automatically.

  With the defective pixel information RD (m, n) necessary for the database DB stored, the photographer sets the irradiation radiation intensity, time, etc. after placing the subject at a predetermined photographing position, and radiation (not shown) When the exposure switch is pressed, shooting is started.

  Imaging is performed in the same manner as described above. After a predetermined time elapses, recording on the solid state detector 20 is stopped and the imaging is terminated. Then, an analog signal corresponding to the latent image charge is output from the solid state detector 20 and an image signal processing unit. A normal image signal ND is generated by A / D conversion. The generated normal image signal ND is transmitted from the image signal processing unit to the computer 4 via the memory.

  When the computer 4 receives the normal image signal ND from the radiation imaging apparatus 3, the computer 4 corrects the defective pixel in the normal image signal ND.

  Here, this correction processing will be described in detail.

  In the database DB, the defective pixel information RD (1,1) with the smallest accumulation time in the solid state detector 20 and the radiation dose to the solid state detector 20 appears the smallest in the defective pixel group, and the deformed defective pixel group and the non-deformed defect group. Since it is easy to specify the center position of both pixel groups, first, the coordinate positions of the centers of the deformed defective pixel group and the non-deformed defective pixel group in the image are specified based on the defective pixel information RD (1, 1).

  As shown in FIG. 6, since the coordinates specified above are defective pixels in the normal image ND, the defective pixel portion is corrected.

  In the normal image ND, there are three deformation defect pixel groups BDPaN, BDPbN, BDPcN and two non-deformation defect pixel groups WDPaN, WPDPbN as in the defective pixel information RD (1, 1). Each of the corrections is performed individually.

  First, correction of the deformation defect pixel group will be described.

  As described above, for the deformed defective pixel group, if the accumulation time in the solid state detector 20 and / or the radiation dose to the solid state detector 20 is different, the region of the defective pixel also changes. Therefore, defective pixel information RD ( m, n) need to be corrected. Among them, the accumulation time in the solid state detector 20 is recognized by the computer 4, and all the defective pixels in the same screen are naturally photographed with the same accumulation time. Will be.

  However, in the solid-state detector 20, since the radiation dose to the portions corresponding to the defective pixel groups BDPaN, BDPbN, and BDPcN in the normal image ND varies depending on the state of the subject, each defective pixel group BDPaN, BDPbN, BDPcN portion. It is necessary to specify the amount of radiation irradiation at the time of imaging for the part corresponding to.

  For specifying the radiation dose, the central portion of the deformed defect pixel group is based on the pixel values in the vicinity of the center of each deformed defect pixel group in the normal image ND, excluding the assumed defective pixel area of the maximum size. It is sufficient to estimate how much radiation dose is. Here, as for the defective pixel region of the maximum size assumed, it is only necessary to refer to a portion having the largest radiation dose for each accumulation time in the database DB. For example, when the accumulation time in the solid state detector 20 is set to 100 ms, the determination may be made based on the defective pixel information RD (1, 5) having the largest radiation dose to the solid state detector 20 among them.

  Note that the method of specifying the radiation dose is not limited to the above method. For example, the pixel value of each pixel is determined from the center to the outside of each deformation defect pixel group, and is closest to the center. In addition to estimating radiation dose based on pixels that are not saturated (pixels that are not the maximum value), estimating based on intermediate values or average values of surrounding pixel areas, irradiation radiation intensity, accumulation time, etc. Any method may be used such as uniformly specifying from the shooting conditions.

  In this way, the radiation dose is specified for each of the three deformation defect pixel groups BDPaN, BDPbN, and BDPcN in the normal image ND, and each deformation defect pixel group in the corresponding defect pixel information RD (m, n) is referred to. To do. And after specifying the area | region to correct | amend about each deformation | transformation defect pixel group, image correction is performed with respect to a defective pixel part with a predetermined | prescribed image processing method. The image processing method for correcting the deformed defective pixel portion is not particularly limited, and any method may be used, for example, a complementary process is performed based on surrounding normal pixels.

  Next, correction of an undeformed defective pixel group will be described.

  The two non-deformed defective pixel groups WDPaN and WDDPbN in the normal image ND are defective because the region that becomes the defective pixel does not change even when the accumulation time in the solid state detector 20 and / or the radiation dose to the solid state detector 20 changes. If the coordinate positions of the non-deformed defective pixel groups WDPaN and WDPabN in the normal image ND are specified with reference to the pixel information RD (1, 1), the image correction is performed on the corresponding position as it is by a predetermined image processing method. . The image processing method for correcting the defective pixel portion is not particularly limited, and any method may be used, for example, a complementary process is performed based on surrounding normal pixels.

  Thus, when correcting the non-deformed defective pixel group, it is not necessary to refer to the database DB, and the correction process can be performed at a higher speed than the correction of the deformed defective pixel group.

  By adopting such an aspect, when there are two types of deformation defect pixel group and non-deformation defect pixel group in the image, it is possible to shorten the correction processing time.

  After the above process is completed, the computer 4 displays the normal image ND ′ corrected for the defective pixel portion as shown in FIG. 7 on the monitor 5 and ends the series of processes.

  Next, a radiographic imaging system according to a second embodiment of the present invention will be described. In the present embodiment, the image stored in the database is changed as compared with the first embodiment. FIG. 8 is a diagram showing a database management structure of the present embodiment, and FIG. 9 is a diagram showing an example of representative defective pixel information stored in the database.

  The change in the defective pixel area of the deformed defective pixel group shows a similar change tendency for each accumulation time in the solid state detector 20 and for each radiation dose to the solid state detector 20 even in different deformed defective pixel groups.

  Therefore, as shown in FIG. 8, in the database DB ′ of the present embodiment, the accumulation time in the solid detector 20 and the radiation dose to the solid detector 20 are the smallest (the accumulation time in the solid detector 20 is 100 ms, the solid The radiation dose to the detector 20 is 0.1 mR) Only defective pixel information RD (1,1) is acquired in order to specify the center positions of the deformed defective pixel group and the non-deformed defective pixel group, and the remaining remaining information For the nineteen portions, the accumulation time in the solid state detector 20 is changed to 100 ms, 500 ms, 1000 ms, and 1500 ms, and the radiation dose to the solid state detector 20 is further set to 0.1 mR, 0.5 mR, 1 for each accumulation time. Acquire representative representative defective pixel information PD (m, n) of the deformed defective pixel group when changed to 0.0 mR, 5.0 mR, and 10 mR. To have.

  An example of the representative defective pixel information PD (m, n) is as follows. The representative defective pixel information PD (1,1) is taken with the accumulation time in the solid state detector 20 being 100 ms and the radiation dose to the solid state detector 20 being 0.5 mR. 2) is as shown in FIG. In this representative defective pixel information PD (1,2), only one representative deformed defective pixel group BDP (1,2) is shown, and the size of the entire image of the representative defective pixel information PD (1,2). Is significantly smaller than the defective pixel information RD (m, n) in which the signal of the entire region of the solid state detector 20 as shown in FIG. 5 is copied, and thereby the size of the image data is also reduced. Can do.

  The representative defective pixel information PD (1, 3) obtained by taking the accumulation time in the solid detector 20 as 100 ms and the radiation dose to the solid detector 20 as 1.0 mR is as shown in FIG. 9B. The image size of the representative defective pixel information PD (1, 3) is the same as the image size of the representative defective pixel information PD (1, 3), and the representative defective pixel information PD (1, 3) includes the representative defective pixel information PD. Although there is a deformed defective pixel group BDP (1,3) at the same position as (1,2), this defective pixel group is defective in comparison with the deformed defective pixel group in the representative defective pixel information PD (1,2). The area recognized as is expanding around.

  The representative defective pixel information PD (1, 4) obtained by taking the accumulation time in the solid state detector 20 as 100 ms and the radiation dose to the solid state detector 20 as 5.0 mR is as shown in FIG. 9C. The image size of the representative defective pixel information PD (1, 4) is the same as the image size of the representative defective pixel information PD (1, 2), and the representative defective pixel information PD (1, 4) includes the representative defective pixel information PD. Although there is a deformed defective pixel group BDP (1, 4) at the same position as (1, 2), this defective pixel group is defective in comparison with the deformed defective pixel group in the representative defective pixel information PD (1, 3). The area recognized as is further expanded to the surroundings.

  In the present embodiment, the deformation defect pixel group in the normal image is corrected by specifying the corresponding radiation defect amount for each deformation defect pixel group in the normal image and corresponding representative defect pixel information PD (m, n). As referred to, there is no representative defect pixel information PD (m, n) unique to each deformation defect pixel group, and the deformation defect pixels having the same accumulation time in the solid state detector 20 and the radiation dose to the solid state detector 20 are the same. If it is a group, the defective pixel range is specified using the same representative defective pixel information PD (m, n). Subsequent processing is the same as in the first embodiment.

  The correction of the non-deformed defective pixel group in the normal image in the present embodiment is the same as that in the first embodiment.

  By setting it as such an aspect, the data amount memorize | stored in database DB 'can be decreased.

  The preferred embodiment of the present invention has been described above. However, the correction of the deformation defect pixel in the normal image according to the present invention is the defect appropriately extracted from the database for each deformation defect pixel group in the normal image as described above. It is not limited to a mode in which correction is performed based on pixel information, but based on revised defective pixel information in which the area of each deformed defective pixel in the defective pixel information is uniformly expanded at the time of correction using single defective pixel information. It is good also as an aspect which correct | amends.

  Further, the present invention is not limited to the radiographic imaging system described above, and may be applied to any imaging system such as a breast imaging system.

  In addition, although a TFT readout type is used as the solid state detector, the same effect can be obtained even with other types of solid state detectors such as an optical readout type.

Schematic which shows the radiographic imaging system of the 1st Embodiment of this invention Schematic diagram of a solid state detector used in the radiographic apparatus of the above system The figure which shows a mode that the size of the defective pixel group in the image image | photographed with the solid state detector of the said system changes with the accumulation time in a solid state detector, and the radiation dose with respect to a solid state detector. Diagram showing the database management structure of the above system The figure which shows an example of the defective pixel information memorize | stored in the database of the said system The figure which shows an example of the image image | photographed by the said system The figure which shows an example after the correction | amendment of the image image | photographed by the said system The figure which shows the management structure of the database of the radiographic imaging system of the 2nd Embodiment of this invention The figure which shows an example of the representative defect pixel information memorize | stored in the database of the said system

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Radiographic imaging system 2 Radiation irradiation apparatus 3 Radiography apparatus 4 Computer 5 Monitor 10 Base 11 Support column 12 Radiation irradiation part 13 Base 14 Support column 15 Radiation image capturing part 16 Screen 17 Handrail 20 Solid state detector

Claims (5)

Defect pixel information based on a uniform irradiation image or non-irradiation image acquired by uniformly irradiating the radiation detector with radiation is stored at the radiation detector and / or the radiation detection. Stored in the database for each radiation dose to the device,
In the normal image acquired by detecting the radiation transmitted through the subject using the radiation detector, the accumulation time in the radiation detector and / or the radiation dose to the radiation detector in the defective pixel information. Identify the radiation dose at the position of the radiation detector corresponding to the position of the deformed defective pixel whose size changes accordingly,
For each deformed defective pixel in the normal image, based on the accumulation time in the radiation detector and / or the radiation dose to the radiation detector, extract the relevant defective pixel information from the database,
Based on the extracted defective pixel information, the image data corresponding to each deformation defective pixel in the normal image is corrected,
For each non-deformed defective pixel whose size does not change in accordance with the accumulation time in the radiation detector in the normal image and / or the radiation dose to the radiation detector, each non-deformed defective pixel in the normal image is subjected to predetermined image processing. A radiation image processing method comprising correcting image data corresponding to a deformed defective pixel. Claims wherein compared to common pixel data in the defective pixel information, the part is large pixel values identified as the deformed defective pixels, and wherein the identifying and the non-deformed defective pixel portions pixel value is small Item 2. The radiographic image processing method according to Item 1 . A radiation detector;
Defect pixel information based on a uniform irradiation image or non-irradiation image acquired by uniformly irradiating the radiation detector with radiation is stored for each accumulation time in the radiation detector and / or the radiation. A database that stores each radiation dose to the detector;
In the image acquired by the radiation detector, the position of the deformed defect pixel whose size changes according to the accumulation time in the radiation detector and / or the radiation dose to the radiation detector, and the accumulation in the radiation detector. A defective pixel specifying means for specifying a position of an undeformed defective pixel whose size does not change in accordance with time and / or a radiation dose to the radiation detector;
Radiation dose specifying means for specifying the radiation dose at the position of the radiation detector corresponding to the position of the deformation defect pixel in the normal image acquired by detecting the radiation transmitted through the subject using the radiation detector When,
For each deformed defective pixel in the normal image, the defective pixel information corresponding to the extracted defective pixel is extracted from the database based on the accumulation time in the radiation detector and / or the radiation dose to the radiation detector. The image data corresponding to the deformation defect pixel is corrected based on the pixel information, and the image data corresponding to the non-deformation defect pixel is corrected by predetermined image processing for each non-deformation defect pixel in the normal image. A radiation image processing apparatus comprising: a correction unit. The defective pixel specifying unit identifies a portion having a large pixel value as the deformed defective pixel and a portion having a small pixel value as the non-deformed defective pixel as compared with general pixel data in the defective pixel information. The radiographic image processing apparatus according to claim 3, wherein Wherein the database, according to claim 3, characterized in that is configured to store only the representative information of the deformation defective pixels for each radiation dose on the accumulation time and / or each said radiation detector in the radiation detector or 4. The radiographic image processing apparatus according to 4 .