WO2020159201A1 - Radiographic image processing method and radiography apparatus - Google Patents

Abstract

The present invention relates to a radiographic image processing method and a radiography apparatus, whereby abnormal pixel values are detected from a plurality of radiographic images acquired by time. At least two radiographic images, among the plurality of radiographic images acquired by time, are selected, and by considering pixel value variations between the selected at least two radiographic images, a determination may be made, by reflecting variations in an input signal by time, on whether the respective pixel values of target pixels are abnormal pixel values including random noise.

Description

Radiation image processing method and radiography apparatus

The present invention relates to a radiographic image processing method and a radiographic imaging apparatus, and more particularly, to a radiographic image processing method and a radiographic imaging apparatus for detecting abnormal pixel values from a plurality of radiographic images obtained over time.

In general, a radiographic imaging apparatus includes a radiation generator that irradiates radiation to a subject and a radiation detector that detects radiation that has transmitted through the subject with a radiation detection panel.

The radiation detection panel may be made of a material such as a complementary metal-oxide semiconductor (CMOS) or a crystal such as polycrystalline silicon (Poly-Si), and noise may occur according to the characteristics of the material.

Since the noise has a relatively high value, when a radiographic image is generated using a pixel value including such noise, the radiographic image may be distorted, thereby degrading the quality of the radiographic image.

In order to solve this, a method for removing noise from a radiographic image has been attempted, but conventional methods detect noise pixels by determining only a tendency for each region in a single radiographic image, and an input signal over time. In the case of random noise generated randomly with respect to time in a plurality of radiographic images acquired over time because the change of the time is not taken into consideration, there is a limit in detecting a noise pixel.

(Patent Document 1) Korean Registered Patent Publication No. 10-0617517

The present invention provides a radiographic image processing method and a radiographic imaging apparatus for detecting abnormal pixel values from a plurality of radiographic images acquired over time.

A radiographic image processing method according to an embodiment of the present invention includes: (a) obtaining a plurality of radiographic images over time using a radiation detection panel having a plurality of pixels; (b) selecting a target pixel among the plurality of pixels from two or more selected radiation images among the plurality of radiation images; (c) calculating a pixel value change amount over time of the selected target pixel; (d) comparing the calculated amount of change of the pixel value with time of the target pixel with a preset reference value; And (e) determining a pixel value of the target pixel as an abnormal value in at least one of the selected two or more radiographic images when the amount of change in the pixel value over time of the target pixel is equal to or greater than the preset reference value. It can contain.

Further comprising the step of verifying the pixel value of the target pixel determined to be an abnormal value, and further comprising the step of verifying the pixel value of the target pixel, the neighboring pixels adjacent to the target pixel calculated from the selected two or more radiographic images Comparing the pixel value change with time and the preset reference value; And checking the pixel value of the target pixel determined as the abnormal value as an abnormal value when the number of surrounding pixels having a change in the pixel value over time is equal to or less than the preset determination number. .

Further comprising the step of verifying the pixel value of the target pixel determined to be an abnormal value, and further comprising the step of verifying the pixel value of the target pixel, the neighboring pixels adjacent to the target pixel calculated from the selected two or more radiographic images Comparing a pixel value change amount with time and a pixel value change amount with time of the target pixel; And the pixel value of the target pixel determined as the abnormal value is abnormal when the number of pixel values variation over time and the number of surrounding pixels having a pixel value variation within a preset allowable deviation range is equal to or less than a preset determination number. It may include the process of confirming by value.

The process of (b) to (e) is repeated for the whole of the plurality of radiographic images, and a process of storing the position of the target pixel in which the pixel value determined as the abnormal value occurs is further included. In the step (b) after the process of storing the position, the pixel storing the position may be excluded from the selection of the target pixel, and the pixel value of the pixel storing the position may be determined as an abnormal value.

The preset reference value may be generated even when radiation is not irradiated from each pixel as a whole to the plurality of pixels, or may be greater than the inherent noise value generated by the radiation source.

The (b) process may be performed by selecting two or more radiographic images that are temporally continuous among the plurality of radiographic images.

In the step (b), two radiographic images that are temporally continuous among the plurality of radiographic images are selected, and in the step (e), two radiographic images in which the amount of change in a pixel value over time of the target pixel is greater than or equal to the preset reference value Among the pixels, the pixel value of the target pixel may be determined as an abnormal value in a radiographic image having a high pixel value of the target pixel.

And detecting a pixel region to which radiation is incident at an intensity within a predetermined range among the plurality of pixels. In the process (b), the target pixel may be selected from the pixel regions.

The process may further include correcting a pixel value of a target pixel determined as an abnormal value by interpolating a pixel value of adjacent neighboring pixels on the same radiation image.

A radiographic imaging apparatus according to another embodiment of the present invention includes a plurality of pixels, and a radiation detection panel that generates electrical signals for acquiring a plurality of radiographic images over time by irradiated radiation; A change amount calculating unit for calculating a pixel value change amount over time of a target pixel selected from the plurality of pixels from two or more radiation images selected from the plurality of radiation images and a pixel value change amount over time of neighboring pixels adjacent to the target pixel; A change amount comparison unit comparing the calculated change amount of the pixel value with time of the target pixel with a preset reference value; And a pixel value abnormality determination unit determining a pixel value of the target pixel as an abnormal value in at least one radiation image among the selected two or more radiographic images when the amount of change of the pixel value over time of the target pixel is equal to or greater than the preset reference value. It may include.

A first pixel value verification unit configured to determine a pixel value of the target pixel determined as the abnormal value as an abnormal value when the number of surrounding pixels having a change amount of the pixel value over time is equal to or less than a preset determination number; It may further include.

The pixel value of the target pixel determined as the abnormal value is an abnormal value when the number of pixel values variation over time and the number of surrounding pixels having a pixel value variation within a preset allowable deviation range is equal to or less than a preset determination number. It may further include; a second pixel value verification unit determined by.

It may further include a location storage unit for storing the location of the target pixel where the pixel value determined as the abnormal value occurs.

It may further include a reference value setting unit for setting the preset reference value that is greater than the intrinsic noise value generated even when radiation is irradiated from each pixel as a whole, or is generated by the radiation source.

In the radiographic image processing method according to an embodiment of the present invention, by selecting two or more radiographic images among a plurality of radiographic images acquired over time, an input signal over time ( Input signal) may be reflected to determine whether a pixel value of each target pixel is an abnormal pixel value including random noise. Accordingly, by correcting by removing abnormal pixel values including random noise from a plurality of radiographic images, distortion of the radiographic image can be prevented, and quality of the radiographic image can be improved.

In addition, by comparing the amount of change of the pixel value with time of the target pixel and a preset reference value, and determining the abnormal pixel value again, verifying that the abnormal pixel value is correct through the surrounding pixels, effectively reflecting the change in the input signal due to the movement of the subject. Thus, it can be more accurately determined whether the pixel value of each target pixel is an abnormal pixel value including random noise.

Also, because of the nature of random noise, the probability of random noise is high in a pixel where random noise has occurred once, so the location of the target pixel where the abnormal pixel value is generated is stored and the pixel value of the pixel is regarded as an abnormal pixel value in the subsequent radiation image. By doing so, the number of pixels that need to be determined to be abnormal pixel values can be reduced, and the time for image processing to determine and correct whether there are abnormal pixel values for the entire plurality of radiographic images can be shortened.

1 is a flowchart illustrating a radiographic image processing method according to an embodiment of the present invention.

2 is a conceptual diagram illustrating a target pixel and surrounding pixels according to an embodiment of the present invention.

3 is a conceptual diagram illustrating verification of a pixel value of a target pixel through a preset reference value according to an embodiment of the present invention.

4 is a conceptual diagram illustrating verification of a pixel value of a target pixel through comparison of a pixel value change amount with time of a target pixel and a change amount of pixel value with time of a surrounding pixel according to an embodiment of the present invention.

5 is a block diagram showing the configuration of a radiographic apparatus according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in more detail with reference to the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various different forms, and only the present embodiments allow the disclosure of the present invention to be complete, and the scope of the invention to those skilled in the art is completely It is provided to inform you. In the description, the same reference numerals are assigned to the same components, and the drawings may be exaggerated in size in order to accurately describe the embodiments of the present invention, and the same reference numerals in the drawings refer to the same elements.

1 is a flowchart illustrating a radiographic image processing method according to an embodiment of the present invention.

Referring to FIG. 1, a radiographic image processing method according to an embodiment of the present invention includes (a) obtaining a plurality of radiographic images over time using a radiation detection panel 110 having a plurality of pixels 111. Process (S100); (b) selecting a target pixel 111a among the plurality of pixels 111 from two or more selected radiation images 10 and 20 among the plurality of radiation images (S200); (c) calculating the amount of change in the pixel value over time of the selected target pixel 111a (S300); (d) comparing the calculated amount of change of the pixel value over time of the target pixel 111a with a preset reference value R (S400); And (e) when the amount of change in pixel value over time of the target pixel 111a is greater than or equal to the preset reference value R, the target pixel in at least one of the selected two or more radiographic images 10 and 20 ( And determining the pixel value of 111a) as an abnormal value (S500).

First, (a) a plurality of radiation images are acquired over time using the radiation detection panel 110 including a plurality of pixels 111 (S100). By irradiating radiation to the radiation detection panel 110 using a radiation generator, a radiation image can be generated, and charges generated by irradiation of radiation are stored in a plurality of pixels 111 of the radiation detection panel 110. The radiation image may be generated according to charges (amounts) stored in the plurality of pixels 111. Here, the generated radiation image may be directly obtained, or the generated and stored radiation image may be obtained. Such a plurality of X-ray images may be used in computed tomography (CT) such as fluorescence (Fluoroscopy), computed tomography (Cone Beam Computed Tomography; CBCT).

Next, (b) a target pixel 111a is selected from the plurality of pixels 111 from two or more selected radiation images 10 and 20 among the plurality of radiation images (S200). Among the plurality of radiographic images, two or more radiographic images 10 and 20 to check whether there are abnormal pixel values may be selected, and an abnormality for a pixel value of each pixel 111 in the selected two or more radiographic images 10 and 20 You can judge cognition. Here, the target pixel 111a may be one or a plurality of pixels.

Next, (c) a pixel value change amount according to the time of the selected target pixel 111a is calculated (S300). The amount of change in pixel values over time of each target pixel 111a may be calculated, or the amount of change in pixel values over time of all target pixels 111a may be calculated at once. Through this, the amount of change in the pixel value over time of the entire target pixel 111a may be calculated.

Then, (d) the calculated change in pixel value over time of the target pixel 111a is compared with a preset reference value R (S400). The calculated amount of change in pixel value over time of the target pixel 111a can be compared with a preset reference value R, and the amount of change in pixel value over time of each target pixel 111a can be compared with a preset reference value R, , It can be compared with a preset reference value R for all target pixels 111a. Here, basically, by removing the value of the element (or noise) that may be interposed in all the pixels 111 basically with time, the difference from the pixel value change amount over time of substantially other pixels (eg, surrounding pixels) The predetermined reference value R may be used to identify an element (for example, random noise) generating.

Then (e) the pixel value of the target pixel 111a in the selected two or more radiographic images 10 and 20 when the amount of change in the pixel value over time of the target pixel 111a is greater than or equal to the preset reference value R It is determined as an abnormal value (S500). Elements commonly included in all the pixels 111 may be excluded by the preset reference value R, and when the amount of change in the pixel value over time of the target pixel 111a is greater than or equal to the preset reference value R, the target pixel ( Since a different element (or noise) is interposed more than other pixels in 111a), a pixel value of the target pixel 111a may be determined as an abnormal value in the selected two or more radiographic images 10 and 20. That is, the portion in which the pixel value of the target pixel 111a changes in common with the pixel value of the peripheral pixel 111b is removed by the preset reference value R, so that the portion where only the pixel value of the target pixel 111a is changed can be seen. In addition, the pixel value of the target pixel 111a may be determined as an abnormal value in the selected two or more radiographic images 10 and 20 through a portion in which only the pixel value of the target pixel 111a is independently changed.

In the present invention, by taking into account the amount of change in the pixel value over time of the target pixel 111a between the selected two or more radiographic images 10 and 20, the amount of change in pixel value over time of the target pixel 111a and a preset reference value (R) Through comparison, it may be determined whether the pixel value of the target pixel 111a is abnormal.

The step (b) (S200) may be performed by selecting two or more consecutive radiographic images 10 and 20 among the multiple radiographic images. Among the plurality of radiographic images, two or more temporally continuous radiographic images 10 and 20 may be selected, and a temporal tendency may be maintained (or continued) between two or more temporally continuous radiographic images 10 and 20. Therefore, the amount of change in pixel value over time can be more effective in detecting abnormal pixel values.

For example, between temporally discontinuous radiation images, changes in the intensity of irradiation of radiation due to device manipulation and/or errors, and intentional and/or accidental changes in the position of a subject (eg, movement or movement) ), the probability of occurrence of a change in the input signal for each pixel 111 is high, and when a change in radiation intensity and/or a position of the subject gradually changes with time The amount of change in pixel value according to is inevitably increased. In this case, the amount of change in the pixel value over time mostly exceeds the preset reference value R, so that only the comparison of the amount of change in the pixel value over time with the preset reference value R of the target pixel 111a results in the pixel value of the target pixel 111a. It is difficult to determine whether this is an abnormal value. In addition, the overall pixel value variation is large, so the ratio of random noise to the pixel value variation may be small, which may make it difficult to verify the pixel value of the target pixel 111a through the peripheral pixel 111b. .

Here, the random noise refers to noise generated randomly with respect to time (or along) for each pixel 111 independently, for example, a complementary metal-oxide semiconductor; CMOS) or polycrystalline silicon (Poly-Si), such as crystals, may be generated depending on the material characteristics of the radiation detection panel 110, and may also be referred to as random telegraph signal (RTS) noise (hereinafter RTS noise). have. The random noise has a relatively high value. Therefore, when a radiographic image is generated using a pixel value including random noise, the radiographic image may be distorted and the quality of the radiographic image may be deteriorated. Here, RTS noise is a noise generated when one or several electrons or holes in a channel of a switching device are trapped or de-traded in an oxide, and discontinuously. It shows characteristics that repeatedly display multiple signal levels at very low frequencies. As a result, in the radiographic image, the signal of the pixel 111 leads to a defect that appears randomly at several levels. In addition, RTS noise is a phenomenon that occurs in almost all semiconductors, and when utilized in a sensor, RTS noise is generally removed by increasing the number of sampling and taking a median value. Although a method is used, it is impossible to perform sampling several times in an image sensor of a passive pixel sensor (PPS) type.

In the step (b), in step S200, two consecutive radiographic images 10 and 20 among the plurality of radiographic images may be selected, and in step (e), in step S500, at the time of the target pixel 111a. The pixel value of the target pixel 111a in the radiographic image 10 or 20 where the pixel value of the target pixel 111a is relatively high among the two radiographic images 10 and 20 in which the amount of change in the pixel value is greater than or equal to the preset reference value R Can be determined as an abnormal value. In general, the pixel values of the target pixel 111a may be determined as both abnormal values in the two radiographic images 10 and 20, but any one of the two radiographic images 10 and 20 may be a radiographic image 10 or 20. Since the pixel value of the target pixel 111a may be a normal value, the pixel value of the target pixel 111a, which is the normal value among the two radiographic images 10 and 20, may be filtered to be continuously used in the radiographic image. Through this, the radiographic image can be imaged similarly to the optimal radiographic image generated only with charge generated by radiation without noise and/or correction. Since the random noise increases the pixel value, the pixel value including the random noise is forced to be higher, and the target pixel 111a in the radiographic image 10 or 20 where the pixel value of the target pixel 111a is relatively high. The pixel value of can be determined as an abnormal value.

The preset reference value R may be generated even when radiation is not irradiated from each pixel 111 as a whole of the plurality of pixels 111 or may be greater (or more) than the inherent noise value generated by radiation. The preset reference value R may be set larger than the intrinsic noise value common to all the plurality of pixels 111. The intrinsic noise may have a different value, but refers to noise common to all of the plurality of pixels 111, and may be noise generated by a device offset. Since it occurs in common with the pixels 111 of, it is not a big problem if all the device offset values are removed from the pixel values of the plurality of pixels 111 constituting the radiation image. For example, the intrinsic noise is generated without causing radiation due to a cause such as a thermal cause, poor insulation, etc., and causes dark current to accumulate charge in a plurality of pixels 111 even when radiation is not irradiated. ) And photons, and may include quantum noise generated by a radiation source such as a radiation generator.

Here, since the intrinsic noise values may be different for each pixel 111, the preset reference value R is an average value of the intrinsic noise values of the plurality of pixels 111 or an intermediate value among the intrinsic noise values of the plurality of pixels 111. Alternatively, it can be set larger (or higher) than the maximum value. On the other hand, since the random noise value is much different from the intrinsic noise value, the preset reference value R may be set to about twice the average value of the intrinsic noise values of the plurality of pixels 111, or the plurality of pixels 111 It can be set to about twice the minimum, median or maximum value of the intrinsic noise value of

In the present invention, the preset reference value R is set to be larger than the intrinsic noise value generated in each pixel 111 as a whole of the plurality of pixels 111, thereby thereby forming a plurality of pixels between the selected two or more radiographic images 10 and 20. The natural variation due to the intrinsic noise can be excluded from the variation in pixel values over time in (111), and the random noise can be detected from the pixel values of the plurality of pixels 111 constituting the radiation image more effectively. have.

Meanwhile, the preset reference value R may be set by using a difference value between the maximum and minimum values of the intrinsic noise that may occur in each pixel 111 and is greater than the difference between the maximum and minimum values of the intrinsic noise ( Above). At this time, for each pixel 111, the difference value between the maximum value and the minimum value of the intrinsic noise may be calculated to set a preset reference value R above a minimum value, a median value, a maximum value, or an average value of the calculated difference values. The preset reference value R may be set to be equal to or greater than a difference value between the maximum value of the intrinsic noise and the minimum value of the intrinsic noise in the entire plurality of pixels 111. In addition, it may be based on 1/2 of the difference between the maximum and minimum values of the inherent noise.

2 is a conceptual diagram for explaining a target pixel and a surrounding pixel according to an embodiment of the present invention, and FIG. 3 is a conceptual diagram for explaining a pixel value verification of a target pixel through a preset reference value according to an embodiment of the present invention 3(a) shows a case determined by an abnormal pixel value, and FIG. 3(b) shows a case determined by a normal pixel value.

2 and 3, the radiographic image processing method according to the present invention may further include a process of verifying a pixel value of the target pixel 111a determined as an abnormal value (S600 ).

In addition, the pixel value of the target pixel 111a determined to be an abnormal value may be verified (S600 ). By verifying the pixel value of the target pixel 111a determined as the abnormal value, whether the amount of change in the pixel value over time of the target pixel 111a is greater than or equal to a preset reference value R is due to a change in the input signal, or the random It is possible to check whether it is caused by noise or both, and the pixel value of the target pixel 111a determined as an abnormal value due to only a change in the input signal is determined as a normal value, thereby forming a radiation image. Can be used as is. Through this, only the abnormal pixel value including (or attributable to) the random noise may be corrected to prevent distortion of the radiographic image, and quality of the radiographic image may be improved.

That is, after comparing the amount of change of the pixel value over time of the target pixel 111a with a preset reference value R to determine the abnormal pixel value, and verifying that the abnormal pixel value is correct through the surrounding pixels 111b, the subject moves. It is possible to more accurately determine whether the pixel value of each target pixel 111a is an abnormal pixel value including the random noise by effectively reflecting the change of the input signal due to the like.

In the process of verifying the pixel value of the target pixel 111a (S600), the amount of change in the pixel value over time of the surrounding pixel 111b adjacent to the target pixel 111a calculated from the selected two or more radiographic images 10 and 20 Comparing the preset reference value R with a step S610; And the pixel value of the target pixel 111a determined as the abnormal value is abnormal when the number of surrounding pixels 111b that is greater than or equal to the preset reference value R is less than or equal to the preset determination number n1. It may include a process of checking by value (S620).

The amount of change in pixel value over time of the neighboring pixel 111b adjacent to the target pixel 111a calculated from the selected two or more radiographic images 10 and 20 and a preset reference value R may be compared (S610 ). Similar to the target pixel 111a, the amount of change in the pixel value over time of the adjacent pixel 111b adjacent to the target pixel 111a calculated from the selected two or more radiographic images 10, 20 is also compared with a preset reference value R Can. Through this, it is possible to grasp the tendency of the amount of change in the pixel value over time in the peripheral area of the target pixel 111a to be similar to the amount of change in the pixel value over time of the target pixel 111a.

Here, the target pixel 111a may be represented by a in the drawing, and a plurality of peripheral pixels 111b may be represented by b1 to b8 as shown in FIGS. 2 and 3. Further, the selected two or more radiographic images 10 and 20 may include a first radiographic image 10 and a second radiographic image 20, and the first radiographic image 10 is generated at a first time t1. It may be a radiographic image, and the second radiographic image 20 may be a radiographic image generated at a second time t2. In this case, in the first radiographic image 10, the target pixel 111a may be represented as a1 as shown in FIG. 2B by reflecting the first time t1, and the peripheral pixels 111b may be represented as b11 to b81. In the second radiographic image 20, the target pixel 111a may be represented by a2 as shown in FIG. 2(c) by reflecting the second time t2, and the peripheral pixels 111b may be expressed by b12 to b82. Can be represented.

The pixel value of the target pixel 111a determined as the abnormal value is abnormal when the number of surrounding pixels 111b that is greater than or equal to the preset reference value R is less than or equal to the preset determination number n1. It can be confirmed (or confirmed) by the value (S620). As shown in (a) of FIG. 3, when the number of surrounding pixels 111b that is greater than or equal to the preset reference value R is less than or equal to the preset determination number n1, the target pixel 111a and the target pixel The reason why the amount of change in the pixel value over time in the peripheral region of (111a) cannot be confirmed is similar, and the reason that the amount of change in the pixel value over time of the target pixel 111a is greater than or equal to the preset reference value R can be considered to be attributed to the random noise. In addition, the pixel value of the target pixel 111a determined as the abnormal value may be determined (or determined) as the abnormal value. And, as shown in (b) of FIG. 3, when the number of surrounding pixels 111b having a change amount of the pixel value over the preset reference value R is greater than a preset determination number n1, the target pixel ( It can be considered that the amount of change in the pixel value over time in the peripheral region of 111a) and the target pixel 111a has a similar tendency, and it can be considered that there is a large change in the input signal rather than the random noise. Accordingly, the pixel value of the target pixel 111a, which is determined to be an abnormal value due to a change in the input signal, can be determined as a normal value through the overturning of the determination, thereby allowing the normal pixel value due to a change in the input signal to be an abnormal value. It can be excluded from judgment.

Here, the number of determinations (n1) is appropriately set to a number that can determine the tendency whether the amount of change in the pixel value over time of the target pixel 111a is similar to the amount of change in the pixel value over time in the peripheral area of the target pixel 111a. It can be set to three or more to reliably grasp similar trends.

4 is a conceptual diagram illustrating verification of a pixel value of a target pixel through comparison of a pixel value change amount with time of a target pixel and a change amount of pixel value with time of a surrounding pixel according to an embodiment of the present invention.

Referring to FIG. 4, a process (S600) of verifying a pixel value of the target pixel 111a is performed on a neighboring pixel 111b adjacent to the target pixel 111a calculated from the selected two or more radiographic images 10 and 20. Comparing the amount of change in the pixel value over time and the amount of change in the pixel value over time of the target pixel (111a) (S630); And the number of surrounding pixels 111b having the amount of change in the pixel value over time of the target pixel 111a and the amount of change in the pixel within the preset tolerance (d) range is less than or equal to the preset number of determinations (n2). And checking the pixel value of the target pixel 111a determined as an abnormal value (S640 ).

A pixel value change amount over time of a neighboring pixel 111b adjacent to the target pixel 111a calculated from the selected two or more radiographic images 10 and 20 and a pixel value change amount over time of the target pixel 111a may be compared. (S630). Through this, it is possible to grasp the tendency of the amount of change in the pixel value over time of the target pixel 111a and the amount of change in the pixel value over time in the peripheral area of the target pixel 111a.

Also, when the number of surrounding pixels 111b having the amount of change in the pixel value and the amount of change in the pixel within the preset tolerance (d) range (± d) according to the time of the target pixel 111a is less than or equal to the preset determination number (n2) The pixel value of the target pixel 111a determined as the abnormal value may be checked (or confirmed) as the abnormal value (S640). The target pixel 111a when the number of pixel values variation over time of the target pixel 111a and the number of surrounding pixels 111b having a pixel value variation within a preset allowable deviation (d) range is less than or equal to a preset determination number (n2) ) And the amount of change in the pixel value over time in the peripheral area of the target pixel 111a cannot be confirmed, and the reason for the amount of change in the pixel value over time of the target pixel 111a is greater than or equal to the preset reference value R is caused by the random noise. It can be considered to be due to the abnormality, and the pixel value of the target pixel 111a determined as the abnormal value may be determined as the abnormal value. Also, when the number of surrounding pixels 111b having the amount of change in the pixel value over time of the target pixel 111a and the amount of change in the pixel within the preset tolerance (d) range is larger than the number of predetermined determinations (n2). It can be considered that the amount of change in the pixel value over time in the peripheral region of the target pixel 111a and the target pixel 111a has a similar tendency, and it can be considered that there is a large change in the input signal rather than the random noise. Accordingly, the pixel value of the target pixel 111a, which is determined to be an abnormal value due to a change in the input signal, can be determined as a normal value through the overturning of the determination, thereby allowing the normal pixel value due to a change in the input signal to be an abnormal value. It can be excluded from judgment. At this time, the allowable deviation (d) may be appropriately set at a preset reference value (R), and may be appropriately determined within about 10% of a change in the pixel value over time of the target pixel (111a), which is not very large.

Here, the number of determinations n2 is suitably set to a number capable of determining the tendency of the amount of change in the pixel value over time of the target pixel 111a to be similar to the amount of change in the pixel value over time in the peripheral area of the target pixel 111a. It can be set to three or more to reliably grasp similar trends.

In addition, in the case of FIG. 4, when the number of determinations (n1, n2) is 3, at the time of the surrounding pixels 111b adjacent to the target pixel 111a calculated as the selected two or more radiographic images 10 and 20, When the amount of change of the pixel value according to the preset reference value R is compared, the number of surrounding pixels 111b having the amount of change of the pixel value over the preset reference value R is only two, so that the target pixel 111a The pixel value is determined as an abnormal value. However, the amount of change in pixel value over time of the adjacent pixel 111b adjacent to the target pixel 111a calculated from the selected two or more radiographic images 10 and 20 is compared with the amount of change in pixel value over time of the target pixel 111a. When doing so, the number of surrounding pixels 111b having the amount of change in pixel value over time of the target pixel 111a and the amount of change in pixel value within a preset tolerance (d) range is 8, and the pixel value of the target pixel 111a is set. It is decided by the normal value. Referring to FIG. 4, since a and b1 to b8 appear to have similar tendencies without large deviation, the time of the peripheral pixel 111b adjacent to the target pixel 111a calculated from the selected two or more radiographic images 10 and 20 When comparing the amount of change of the pixel value with time and the amount of change of the pixel value with time of the target pixel 111a, the amount of change with the pixel value with time of the target pixel 111a and the value of the pixel with time of the peripheral area of the target pixel 111a It is possible to grasp (or confirm) the tendency of the change amount more accurately.

As described above, in the present invention, by selecting two or more radiographic images 10 and 20 among a plurality of radiographic images acquired over time, by considering the amount of change in pixel values between the selected two or more radiographic images 10 and 20, over time, In response to a change in the input signal, it may be determined whether the pixel value of each target pixel 111a is an abnormal pixel value including the random noise. Accordingly, by correcting by removing the abnormal pixel value including the random noise from a plurality of radiographic images, distortion of the radiographic image can be prevented, and the quality of the radiographic image can be improved.

The (b) process (S200) to the (e) process (S500) may be repeated for the whole of the plurality of radiographic images, and the radiographic image processing method according to the present invention may generate a pixel value determined as an abnormal value. The process of storing the position of the pixel 111a (S700); may further include, and the pixel in which the position is stored in the (B) process (S200) after the process of storing the position of the target pixel 111a (S700) The pixel value of the pixel 111 in which the location is stored may be determined as an abnormal value by excluding (111) from the selection of the target pixel 111a.

The (b) process (S200) to the (e) process (S500) may be repeated, and the (b) process (S200) to (e) process (S500) may be performed on the whole of the plurality of radiographic images. can do. Further, it may be determined whether the pixel value is an abnormal value for all the plurality of pixels 111 or all target pixels 111a for which inspection (or determination) is desired.

In addition, the position of the target pixel 111a in which the pixel value determined (or determined) as the abnormal value has occurred may be stored (S700 ). Due to the nature of the random noise, since the probability of the random noise occurring again is high in the pixel 111 where the random noise has once occurred, the position of the target pixel 111a in which an abnormal pixel value has occurred is stored, and the position in the subsequent radiation image It may be regarded as an abnormal pixel value without performing the steps (b) (S200) to (e) (S500) of determining whether the stored pixel 111 is an abnormal pixel value.

In the step (B) after the process of storing the position of the target pixel 111a (S700) (S200), the pixel storing the position is excluded from the selection of the target pixel 111a, and the pixel storing the position ( The pixel value of 111) may be determined (or considered) as an abnormal value. That is, because of the nature of the random noise, since the probability of generating the random noise is high in the pixel 111 where the random noise has once occurred, the location of the target pixel 111a where the abnormal pixel value has occurred is stored in the subsequent radiation image. By considering the pixel value of the corresponding pixel 111 as an abnormal pixel value, the number of pixels 111 to determine whether the pixel value is abnormal can be reduced. In addition, the time of image processing for determining and correcting whether there are abnormal pixel values for all of the plurality of radiographic images may be shortened.

The radiographic image processing method according to the present invention may further include a step (S150) of detecting a pixel region to which radiation is incident at an intensity within a predetermined range among a plurality of pixels 111 (S200). ), a target pixel 111a may be selected from the pixel areas.

In addition, a pixel region in which radiation is incident at an intensity within a predetermined range among the plurality of pixels 111 may be detected (S150 ). In the radiographic image, a complex region in which the pixel value difference between adjacent pixels 111 is severe (or very large), such as an edge region of the subject and a boundary region between parts of the subject, and a central region of each part of the subject, such as A simple region in which the difference in pixel values between the pixels 111 is not large (or insignificant) may exist. Since the input signal to each pixel 111 is almost entirely different in the complex area, the tendency of the pixel value change over time of the target pixel 111a and the pixel value change over time of the peripheral area of the target pixel 111a can be grasped. It is difficult to detect the abnormal pixel value including the random noise, and even if the random noise is included, it is not a big problem in the part where medical judgment is made from the radiographic image because it is unrecognizable to the eye in the radiographic image. However, since the simple region can grasp the tendency of the amount of change in the pixel value over time of the target pixel 111a and the amount of change in the pixel value over time in the peripheral area of the target pixel 111a, the abnormal pixel value including the random noise When it is possible to detect and the random noise is included, it may be a problem when the medical image is identified as the radiographic image containing the random noise, which is identified as an abnormal part in the radiographic image. Accordingly, a pixel region (ie, the simple region) in which radiation is incident at an intensity within a predetermined range among the plurality of pixels 111 is detected, and only the pixel(s) of the pixel region in which radiation is incident at an intensity within the predetermined range is pixels. It can be determined whether the value is an abnormal value.

In the step (b) (S200), a target pixel 111a may be selected from the pixel areas. Target pixel 111a among the pixel areas to determine whether only the pixel(s) of the pixel area where radiation is incident at an intensity within the preset range capable of detecting the abnormal pixel value including the random noise is an abnormal value You can choose In addition, when the target pixel 111a is selected from the pixel areas, all of the plurality of pixels 111 do not determine whether the pixel values are abnormal values, but only the pixel(s) corresponding to some of the pixel areas have abnormal pixel values. By determining whether it is a value, the time for detecting the abnormal pixel value including the random noise in the selected radiation image can be shortened.

The method of processing a radiographic image according to the present invention further includes a process of correcting a pixel value of a target pixel 111a determined as an abnormal value by interpolating a pixel value of adjacent peripheral pixels 111b on the same radiographic image (S800). Can.

Further, the pixel value of the target pixel 111a determined (or determined) as an abnormal value may be corrected by interpolating the pixel value of the adjacent pixels 111b on the adjacent (one) radiation image (S800). For example, the pixel value of the target pixel 111a determined (or determined) as an abnormal value may be replaced with an average value or a median value of the pixel values of the adjacent peripheral pixels 111b. At this time, when the adjacent peripheral pixel 111b includes a pixel 111 having a pixel value determined (or determined) as an abnormal value, a peripheral pixel having a pixel value determined as an abnormal value in calculation of an average value or a median value (111b) can be excluded. The method of correcting the pixel value of the adjacent pixel 111b with the pixel value of the target pixel 111a determined to be an abnormal value may be performed in various ways in addition to this method.

5 is a block diagram showing the configuration of a radiographic imaging apparatus according to an embodiment of the present invention.

The radiographic imaging apparatus according to another embodiment of the present invention will be described in more detail with reference to FIG. 5, and the overlapping with the above-described part in relation to the radiographic image processing method according to an embodiment of the present invention will be omitted.

The radiographic imaging apparatus 100 according to another embodiment of the present invention includes a plurality of pixels 111, and a radiation detection panel for generating electrical signals for acquiring a plurality of radiographic images over time by irradiated radiation ( 110); The amount of change in pixel value over time of the selected target pixel 111a among the plurality of pixels 111 from two or more selected radiographic images 10 and 20 among the plurality of radiographic images and surrounding pixels adjacent to the target pixel 111a ( A change amount calculating unit 121 for calculating a pixel value change amount over time in 111b); A change amount comparison unit 122 comparing the calculated change amount of the pixel value with time of the target pixel 111a with a preset reference value R; And when the amount of change in the pixel value over time of the target pixel 111a is equal to or greater than the preset reference value R, the target pixel 111a is selected from at least one of the selected two or more radiographic images 10 and 20. It may include; a pixel value abnormality determination unit 123 for determining the pixel value as an abnormal value.

The radiation detection panel 110 may include a plurality of pixels 111 and may generate electrical signals for acquiring a plurality of radiation images over time by irradiated radiation. A radiographic image may be generated and obtained by the electric signal thus generated.

The change amount calculating unit 121 may change the pixel value and the target pixel over time of the selected target pixel 111a among the plurality of pixels 111 from two or more selected radiation images 10 and 20 among the plurality of radiation images. The amount of change in the pixel value over time of the peripheral pixel 111b adjacent to 111a) may be calculated. Meanwhile, the radiographic imaging apparatus 100 of the present invention is a pixel setting unit (not shown) that selects a target pixel 111a among a plurality of pixels 111 from two or more radiographic images 10 and 20 selected from the plurality of radiographic images. ); may be further included.

The change amount comparison unit 122 may compare the calculated change amount of the pixel value according to the time of the target pixel 111a with a preset reference value R, and the pixel according to the time of the peripheral pixel 111b adjacent to the target pixel 111a The value change amount may be compared with a pixel value change amount and/or a preset reference value R over time of the target pixel 111a.

The pixel value anomaly determining unit 123 may select the target pixel 111a from the selected two or more radiographic images 10 and 20 when the amount of change in the pixel value over time of the target pixel 111a is greater than or equal to a preset reference value R. The pixel value of can be determined as an abnormal value.

The radiographic imaging apparatus 100 of the present invention may further include an image processing unit 120 that performs image processing of the generated or obtained radiographic image, and the image processing unit 120 compares the change amount calculation unit 121 and the change amount The unit 122 and the pixel value abnormality determination unit 123 may be included.

The radiographic imaging apparatus 100 according to the present invention is determined to be the abnormal value when the number of surrounding pixels 111b equal to or greater than the preset reference value R is less than or equal to the preset determination number n1. It may further include a first pixel value verification unit (not shown) that determines the pixel value of the target pixel 111a as an abnormal value.

The first pixel value verification unit (not shown) determines the abnormal value when the number of surrounding pixels 111b equal to or greater than the preset reference value R is less than or equal to the preset determination number n1. The pixel value of the target pixel 111a may be determined as an abnormal value to accurately verify whether the pixel value of the target pixel 111a is an abnormal value.

The radiographic imaging apparatus 100 according to the present invention is the number of surrounding pixels 111b having a pixel value change amount over time and a predetermined value (d) within a range (± d) of the target pixel 111a. A second pixel value verification unit (not shown) that determines the pixel value of the target pixel 111a determined as the abnormal value as an abnormal value when is less than or equal to the preset number of determinations n2 may be further included.

The second pixel value verifying unit (not shown) determines that the number of surrounding pixels 111b having the amount of change in pixel values over time of the target pixel 111a and the amount of change in pixel values within a preset tolerance (d) range is preset. When the number n2 is less than or equal to the pixel value of the target pixel 111a determined as the abnormal value, the pixel value of the target pixel 111a can be accurately verified.

The radiographic imaging apparatus 100 according to the present invention may further include a location storage unit (not shown) that stores the location of the target pixel 111a where the pixel value determined as the abnormal value has occurred.

The location storage unit (not shown) may store the location of the target pixel 111a where the pixel value determined as the abnormal value occurs. Due to the nature of random noise, since the probability of the random noise occurring again is high in the pixel 111 where the random noise has once occurred, the position of the target pixel 111a where the abnormal pixel value has occurred is stored, and the position in the subsequent radiation image is The stored pixel 111 may be regarded as an abnormal pixel value without a separate process of determining whether it is an abnormal pixel value.

The radiographic imaging apparatus 100 according to the present invention is a reference value setting unit for setting the preset reference value greater than the intrinsic noise value generated even when radiation is irradiated from each pixel as a whole to the plurality of pixels or generated by a radiation source (Not shown); may further include.

The reference value setting unit (not shown) is a preset reference value (or greater than or equal to) that is generated even when radiation is not irradiated from each pixel 111 as a whole, or greater than or equal to the intrinsic noise value generated by the radiation source. R) can be set. The preset reference value R may be set to be larger (or higher) than the intrinsic noise value common to all the plurality of pixels 111. The intrinsic noise may have a different value, but refers to noise common to all the plurality of pixels 111, and may be noise generated by a device offset, and the noise may be any of the plurality of pixels 111 ), it is not a big problem if all of the device offset values are removed from the pixel values of the plurality of pixels 111 constituting the radiation image. For example, the intrinsic noise is generated due to a cause of thermal (thermal), poor insulation, and the like due to radiation, and dark current and quantum of photons accumulating charge in the plurality of pixels 111 It is caused by nature, and may include quantum noise generated by a radiation generator.

As described above, in the present invention, by selecting two or more radiographic images among a plurality of radiographic images acquired over time, and taking into account the amount of change in pixel values between the selected two or more radiographic images, the change in the input signal over time is reflected. It may be determined whether the pixel value of each target pixel is an abnormal pixel value including random noise. Accordingly, by correcting by removing abnormal pixel values including random noise from a plurality of radiographic images, distortion of the radiographic image can be prevented, and quality of the radiographic image can be improved. In addition, by comparing the amount of change of the pixel value with time of the target pixel and a preset reference value, and determining the abnormal pixel value again, verifying that the abnormal pixel value is correct through the surrounding pixels, effectively reflecting the change in the input signal due to the movement of the subject. Thus, it can be more accurately determined whether the pixel value of each target pixel is an abnormal pixel value including random noise. Also, because of the nature of random noise, the probability of random noise is high in a pixel where random noise has occurred once, so the location of the target pixel where the abnormal pixel value is generated is stored and the pixel value of the pixel is regarded as an abnormal pixel value in the subsequent radiation image. By doing so, the number of pixels that need to be determined to be abnormal pixel values can be reduced, and the time for image processing to determine and correct whether there are abnormal pixel values for the entire plurality of radiographic images can be shortened.

Although the preferred embodiments of the present invention have been shown and described above, the present invention is not limited to the above-described embodiments, and general knowledge in the field to which the present invention pertains without departing from the gist of the present invention claimed in the claims Anyone who has a will understand that various modifications and other equivalent embodiments are possible. Therefore, the technical protection scope of the present invention should be defined by the following claims.

Claims (14)

1. (a) obtaining a plurality of radiation images over time using a radiation detection panel having a plurality of pixels;

   (b) selecting a target pixel among the plurality of pixels from two or more selected radiation images among the plurality of radiation images;

   (c) calculating a pixel value change amount over time of the selected target pixel;

   (d) comparing the calculated amount of change of the pixel value with time of the target pixel with a preset reference value; And

   (e) determining a pixel value of the target pixel as an abnormal value in at least one of the selected two or more radiographic images when the amount of change in the pixel value over time of the target pixel is greater than or equal to the preset reference value; Radiographic image processing method.
2. The method according to claim 1,

   Further comprising the step of verifying the pixel value of the target pixel determined to be an abnormal value;

   The process of verifying the pixel value of the target pixel,

   Comparing the pixel value change amount with time of a neighboring pixel adjacent to the target pixel calculated from the selected two or more radiographic images and the preset reference value; And

   And checking the pixel value of the target pixel determined as the abnormal value as an abnormal value when the number of surrounding pixels over the preset reference value is less than or equal to the preset determination number. Way.
3. The method according to claim 1,

   Further comprising the step of verifying the pixel value of the target pixel determined to be an abnormal value;

   The process of verifying the pixel value of the target pixel,

   Comparing a pixel value change amount over time of a neighboring pixel adjacent to the target pixel calculated from the selected two or more radiographic images and a pixel value change amount over time of the target pixels; And

   The pixel value of the target pixel determined as the abnormal value is an abnormal value when the number of pixel values variation over time and the number of surrounding pixels having a pixel value variation within a preset allowable deviation range is equal to or less than a preset determination number. A radiographic image processing method comprising the step of confirming.
4. The method according to claim 1,

   The above steps (b) to (e) are repeated for the whole of the plurality of radiographic images,

   Further comprising; storing the position of the target pixel where the pixel value determined as the abnormal value occurs;

   A radiographic image processing method of determining a pixel value of a pixel in which the location is stored as an abnormal value by excluding a pixel in which the location is stored in the selection of the target pixel in the step (b) after the process of storing the location of the target pixel.
5. The method according to claim 1,

   The preset reference value is generated even when radiation is not irradiated from each pixel as a whole to the plurality of pixels, or a method of processing a radiographic image larger than an inherent noise value generated by a radiation source.
6. The method according to claim 1,

   The step (b) is a radiographic image processing method performed by selecting two or more temporally continuous radiographic images among the plurality of radiographic images.
7. The method according to claim 6,

   In the step (b), two consecutive radiographic images are selected from the plurality of radiographic images,

   In the step (e), a pixel value of the target pixel is determined as an abnormal value in a radiographic image having a high pixel value of the target pixel among two radiographic images in which the amount of change in the pixel value over time of the target pixel is greater than or equal to the preset reference value. Radiographic image processing method.
8. The method according to claim 1,

   Further comprising the step of detecting a pixel region to which the radiation is incident at an intensity within a predetermined range among the plurality of pixels;

   In the step (b), a radiographic image processing method of selecting the target pixel among the pixel areas.
9. The method according to claim 1,

   And correcting the pixel value of the target pixel determined as the abnormal value by interpolating the pixel value of adjacent surrounding pixels on the same radiation image.
10. A radiation detection panel having a plurality of pixels and generating electrical signals for acquiring a plurality of radiation images over time by the irradiated radiation;

    A change amount calculating unit for calculating a pixel value change amount over time of a target pixel selected from the plurality of pixels from two or more radiation images selected from the plurality of radiation images and a pixel value change amount over time of neighboring pixels adjacent to the target pixel;

    A change amount comparison unit comparing the calculated change amount of the pixel value with time of the target pixel with a preset reference value; And

    A pixel value abnormality determination unit determining a pixel value of the target pixel as an abnormal value in at least one radiation image among the selected two or more radiographic images when the amount of change of the pixel value over time of the target pixel is equal to or greater than the preset reference value; Radiography apparatus comprising.
11. The method according to claim 10,

    A first pixel value verification unit configured to determine a pixel value of the target pixel determined as the abnormal value as an abnormal value when the number of surrounding pixels having a change amount of the pixel value with respect to the time is equal to or less than a preset determination number; Radiographic apparatus further comprising.
12. The method according to claim 10,

    The pixel value of the target pixel determined as the abnormal value is an abnormal value when the number of pixel values change over time and the number of surrounding pixels having a pixel value variation within a preset allowable deviation range is equal to or less than a preset determination number. And a second pixel value verifying unit determined as.
13. The method according to claim 10,

    And a location storage unit which stores a location of a target pixel in which a pixel value determined as an abnormal value occurs.
14. The method according to claim 10,

    And a reference value setting unit configured to set the preset reference value greater than the intrinsic noise value generated even when radiation is irradiated from each pixel as a whole, or greater than the inherent noise value generated by the radiation source.

Images