JPH057578A - Image area determining method and image processing method and apparatus for radiation image - Google Patents

Abstract

PURPOSE:To achieve the optimum image processing by a method wherein image data of a radiation image is analyzed to determine a desired image area and an image area for determining image processing conditions is extracted optimally to ensure the optimum image processing data while achieving a handiness of the apparatus. CONSTITUTION:An X-ray irradiator 1 make X rays irradiate an object 2. On the other hand, an image reader 14 forms an latent image corresponding to irradiation X rays with an accelerated phosphofluorescence layer of a radiation image conversion panel 3. The radiation image panel 3 generates an accelerated fluorescence proportional to a light beam to be irradiated from an accelerated fluorescence excitation light source 4 while the filter 5 allows an accelerated phosphofluo-rescence alone to be impinged into a photoelectric converter 6. which 6 outputs a current signal proportional to the phos-phofluorescence to an A/D conversion section 7. Moreover, the A/D conversion section 7 transfers image data to an image processor 8 at a specified timing. The image processor 8 is provided with an image area determining means 1, a contrast processing/determining means 12, a contrast processing means and the like to execute an image processing.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for determining an image area used when image processing a radiation image captured by radiation such as X-rays, an image processing method using the determination method, and an image processing apparatus. Regarding More specifically, a method for determining an image area for determining an image processing condition from a radiation image, an image processing method for making the radiation image easier to see based on image data included in the determined image area, and an apparatus thereof. Regarding

[0002]

2. Description of the Related Art Radiation images such as X-ray images are often used for diagnosing diseases such as the human body. To obtain this X-ray image, X-rays transmitted through an object are phosphor layers (fluorescent screens).
The so-called radiography in which a silver salt film is developed by irradiating the same with visible light to generate visible light and irradiating the silver salt film with the visible light in the same manner as in ordinary photography, has heretofore been widely used.

However, in recent years, there has been proposed a method of directly extracting an irradiation line image from a phosphor layer without using the phosphor screen and the silver salt film. Specifically, for example, US Pat. No. 3,859,527 and JP-A-55-12.
Japanese Patent Publication No. 144, etc. discloses a radiation image conversion method using a stimulable phosphor and visible light or infrared rays as stimulable excitation light. This conversion method uses a radiation image conversion panel in which a stimulable phosphor layer is formed on a support, and this conversion panel is irradiated with the radiation that has passed through the subject and the radiation corresponding to the radiation transmittance of each part of the subject. Energy is stored to form a latent image. afterwards,
By scanning the conversion panel with stimulable excitation light, the radiation energy accumulated in the conversion panel is emitted as fluorescence, and the fluorescence is photoelectrically converted to obtain a radiation image signal.

The radiation image signal is processed as it is or after being subjected to image processing to obtain a silver salt film,
Although it is output to a monitor such as a CRT or the like for visualization, it is often digitized for image processing by a computer at this time. The image data of the digitized radiation image is stored in an image storage device such as a semiconductor storage device, a magnetic storage device, an optical disk storage device, or a magneto-optical storage device, and then taken out from the image storage device as needed. In some cases, it is output to a silver salt film, monitor, etc. and visualized.

There is also a method in which a silver salt film storing a radiation image is irradiated with light from a light source such as a laser or a fluorescent lamp, and the light transmitted through the silver salt film is photoelectrically converted to be digitized. The apparatus for obtaining digital image data from this silver salt film is configured such that a light beam is scanned over the silver salt film in one direction and the silver salt film is conveyed in a direction orthogonal to the scanning direction, and is provided on the opposite side of the light source. To detect transmitted light with a photodetector, or to attach a silver salt film to the outside of a transparent drum containing a light source and rotate the drum, and at the same time to provide an aperture for guiding the transmitted light to the photodetector of the drum. There is a device that moves it substantially parallel to the rotation axis.

Here, when reproducing the image data obtained by the above method, the density (luminance in the case of a monitor) of the region of interest (the region including the image portion necessary for diagnosis in medical use) in the reproduced image is determined. Depending on the purpose of making it uniform and the purpose of outputting the structure of the human body and the shadows of lesions in a way that is easier to see, the image data is subjected to image processing such as gradation processing and spatial frequency processing before being visualized and used for diagnosis. There is.

In the image processing, it is desired to output the image data of the region of interest under good conditions and reproduce the data of the region of interest in an easy-to-see manner. Therefore, the image data corresponding to the region of interest in the radiation image is specified. However, it is necessary to determine the image processing condition based on the specified result. For example, when radiography is performed on a subject such as the chest of a human body to diagnose the lung condition, it is necessary to perform image processing so that the lung field, which is the region of interest, is easy to see, and image data for the lung field is specified. As a method for doing so, a method for obtaining from a histogram of image data of a radiation image and a method for obtaining from statistical properties (maximum value, minimum value, average value, etc.) are known (Japanese Patent Publication No. 1-383818, Japanese Patent Laid-Open Publication No. Sho. 63-
No. 31641, etc.).

[0008]

However, since the histogram and the statistical properties are largely different for each image data depending on the body type of the patient, imaging conditions (for example, radiation dose, imaging voltage, etc.), imaging positioning, etc., Even if the image data corresponding to the region of interest is specified from the histogram, the image data cannot be specified accurately, so that there is a problem that the optimum image condition cannot always be set and the image processing cannot be performed.

The present invention has been made in view of such circumstances, and makes it possible to secure the optimum image processing data for each image without being affected by the body shape of the patient, imaging conditions, imaging positioning, and the like. The purpose is to

[0010]

Therefore, according to the first aspect of the present invention, the image data of the radiation image formed corresponding to the transmission amount of the radiation passing through each part of the subject is analyzed to obtain a desired value. An image area determination method for a radiographic image for determining an image area is provided. Further, in claim 2, after the desired image area in claim 1 is divided into a plurality of image areas by comparing the radiation image data with a threshold value and the radiation image is divided into a plurality of image areas, I decided to make a decision.

Further, according to a third aspect of the present invention, the image data of the radiation image formed corresponding to the transmission amount of the radiation passing through each part of the subject is analyzed to determine the desired image area, and the determined desired area is determined. Provided is an image processing method of a radiation image, which determines an image processing condition based on image data of an image region, and image-processes the radiation image based on the determined image processing condition.

Further, in the present invention, the image area determining means for analyzing the image data of the radiation image formed corresponding to the transmission amount of the radiation passing through each part of the subject and determining the desired image area is determined. A radiation image processing apparatus including: an image processing condition determining unit that determines an image processing condition based on image data of a desired image region; and an image processing unit that performs image processing on the radiation image based on the determined image processing condition. I will provide a.

According to a fifth aspect of the present invention, the image area determining means according to the fourth aspect compares the radiation image data with a threshold value to divide the radiation image into a plurality of image areas, and the plurality of divided images. It was configured to determine the desired image area from the area.

[0014]

In the first aspect, the radiation image is analyzed to determine the desired image area, and in the second aspect, the radiation image is divided into a plurality of image areas and then the desired image area is determined from those image areas. It was decided so that optimum image processing could be performed. Further, in claims 3 and 4, a desired image region is determined from the image data of the radiation image, and then image processing conditions are determined from the image data of the determined desired image region to optimally process the radiation image. I chose
In the fifth aspect, the desired image area is determined from the image areas after determining the plurality of image areas from the image data of the radiation image and the threshold value.

[0015]

Embodiments of the present invention will be described below with reference to the drawings. In the present embodiment, X-rays will be described as an example of radiation. FIG. 1 showing an embodiment of the present invention includes medical radiation including an image processing apparatus for a radiation image according to the present invention, which embodies an image area determining method for a radiation image and an image processing method for a radiation image according to the present invention. An image recording / reading device.

That is, the X-ray irradiator 1, which is a radiation source, is controlled by an X-ray controller (not shown) and irradiates the subject 2 with X-rays. The image reading device 14 includes a radiation image conversion panel 3 that faces the X-ray irradiation device 1 with the subject 2 interposed therebetween. This conversion panel 3 is provided with a stimulable phosphor layer, and the stimulable phosphor layer follows the transmittance distribution of X-rays transmitted through the subject 2 in accordance with the X-ray dose irradiated from the X-ray irradiation device 1. It stores energy and forms a latent image. The radiation image conversion panel 3 is provided with the stimulable phosphor layer on a support, and the stimulable phosphor layer is shielded or covered by a protective member to prevent adverse effects and damages from the environment. There is. Examples of the stimulable phosphor material include, for example, JP-A-61-272091 and JP-A-59.
No. 7500.

The stimulable excitation light source 4 generates a visible light ray or an infrared ray whose emission intensity is controlled, and irradiates the radiation image conversion panel 3 with the light ray by a scanning method via various optical systems. To do. By this irradiation, the radiation conversion panel 3
Emits stimulable fluorescence in proportion to the stored energy. The filter 5 separates the light beam from the stimulated excitation light source 4 and the stimulated fluorescent light and allows only the stimulated fluorescent light to enter the photoelectric converter 6,
The photoelectric converter 6 outputs a current signal proportional to the amount of stimulated fluorescence to the analog / digital (A / D) converter 7.

The A / D converter 7 converts the input current signal into digital image data and outputs it to the image processing device 8. As a specific hardware configuration of the A / D conversion unit 7, as shown in FIG. 2, the output voltage of the current / voltage converter 7A that converts the output current of the photoelectric converter 6 into a voltage signal is output via the amplifier 7B. It is input to the A / D converter 7C. Here, the amplifier 7B may be a logarithmic amplifier. The A / D converter 7C converts the analog signal into a digital signal (digital image data) and outputs it to the control circuit 7D. The control circuit 7D adjusts the gains of the current / voltage converter 7A and the amplifier 7B, adjusts the input dynamic range of the A / D converter,
At the same time, the read gain of the radiation image information is adjusted comprehensively, and the image data is transferred to the image processing device 8 at a predetermined timing.

In the image processing device 8, as shown in FIG.
Binarization / labeling means 9 and image area determination means 10 described later.
A signal value detecting means 11, a resolution processing determining means 12 and a gradation processing means 13 are provided. As a specific hardware configuration of the image processing device 8, as shown in FIG. 2, a central processing unit (hereinafter abbreviated as CPU) 21 is provided, and the CPU 21 has a confirmation monitor 22 and a display control unit 23 and an image. It is connected via the bus VB.

A frame memory 24 for storing image processing data and the like is connected to the CPU 21 via a frame memory controller 25 and an image bus VB. Further, a keyboard 26 for inputting identification information of a subject (name, sex, date of birth, etc.) and a display device 27 for displaying this input information are provided, and the keyboard 26 and the display device 27 are an interface.
It is connected to the CPU 21 via 28.

A timing control unit 29 for outputting a timing control signal is provided. The timing control unit 29 outputs the timing control signal to the X-ray control unit of the X-ray irradiation apparatus 1 via the adapter 30, and the control Output to the circuit 7D. A magnetic memory 31 for recording image data is provided,
The magnetic memory 31 stores image-processed image data and unprocessed image data according to a signal from the magnetic memory control unit 32. The image data may be recorded by an external optical disk device or a magnetic tape device as shown by the broken line in FIG.

A memory 34 stores a control program and the like. Reference numeral 33 is an I / O interface for an external device (for example, the image output device 15, a host computer, etc.), and the image processing device 8 can be connected to the external device. Here, the CPU 21 constitutes the image area determining means, the image processing condition determining means, and the image processing means in claim 4.

Further, the function of the CPU 21 according to the present invention will be described in detail with reference to FIG. That is, the original radiation image data read from the stimulable phosphor layer of the radiation image conversion panel 3 determines the image area for determining the gradation processing condition as the image processing condition before performing the gradation processing. In order to do so, the binarization unit 42 compares it with a predetermined threshold value, binarizes it, and assigns an identification code to each pixel. For example, the identification code 1 is given to the pixels equal to or more than the threshold value, and the identification code 0 is given to the pixels less than the threshold value. The threshold is set by the threshold setting unit 42, and the initial value of the threshold is set according to the radiation image data. The binarization process is not performed on the entire image when there is an unnecessary portion in the radiation image, the X-ray irradiation unit is regulated, or special information about the subject is not used as an effective image. Although it may be performed only for the region, it is preferable to perform it for the entire image unless such a situation occurs. The labeling unit 43 performs a labeling process on the identification coded image data by determining the continuity of the identification code. Specifically, for example, when a plurality of pixels having the identification code 1 are adjacent and continuous, the same label A, for example, is given to all the pixels of the continuous pixel group. If a plurality of pixels having the identification code 1 are adjacent to each other in other portions and are consecutive, a different label B is given to all the pixels in the consecutive pixel group. This operation is performed for all consecutive pixel groups.

Here, "a plurality of pixels are adjacent to each other in succession" means that the pixels adjacent to a predetermined pixel in the vertical and horizontal directions have the same identification code, and the pixels in the diagonal vertical direction in addition to the above direction have the same identification code. , Say. Further, when performing the labeling process, in addition to the continuity of the identification code, the number of consecutive pixels is added to the criterion, and for example, n or more (n is an integer of 2 or more) consecutive pixels You may label only a group. Further, labeling may be performed for both identification codes, but it is preferable to use only one of them depending on the purpose. Specifically, when it is desired to obtain the maximum value from the image data, it is better to label only the image of the identification code 1, and when it is desired to obtain the minimum value, it is better to label only the pixel of the identification code 0. .

The image area determination unit 44 determines a desired image area from the labeled image areas. Specifically, a target pixel group (desired image area) is selected from a plurality of labeled pixel groups, for example, based on information on their positions in the radiation image and signal magnitudes. The selected pixel group may be one or plural. At this time, when the desired image area is not obtained, the threshold value setting unit 41 changes the threshold value (for example, decreases it) and repeats the binarization process again.

Therefore, the threshold value setting section 41, the binarization section 42, the labeling section 43 and the image area determination section 44 constitute the image area determination means in claim 4. When the desired image area is determined, the image data extraction unit 45 statistically processes the image data in the image area within the image area, and the information required by the next gradation processing condition determination unit 46. And outputs the extracted information to the gradation processing condition determining unit 46. The gradation processing condition determining unit 46 as the image processing condition determining means determines the gradation processing condition based on the information from the image data extracting unit 45. Specifically, when a radiation image is reproduced on the image output device 15, the gradation processing table is corrected based on the information so that the image has density and gradation characteristics that are diagnostically easy to read. Has become. When the gradation processing condition is determined in this way, finally, the gradation processing unit 47 as an image processing means gradation-processes the original radiation image data and outputs the image data to the image output device 15.

The image data used in each unit other than the gradation processing unit 47 does not have to be used for all the data (all pixels) read by the image reading device 14, and the image data thinned out from the original image data, Alternatively, image data obtained by averaging original image data may be used to simplify the process for determining gradation processing conditions. In addition, "look-ahead" as disclosed in JP-A-58-67240
The gradation processing condition may be determined based on the image data obtained by, and the image data obtained by the “main reading” may be processed according to the determined processing condition.

Further, although gradation processing is taken up as image processing in the above embodiment, it goes without saying that it can be applied to image processing such as frequency processing. Next, a human chest side image is taken as an example of the subject, and description will be given based on FIGS. 4 to 12. Figure 4 is a schematic representation of a side view of the human chest.
8 is composed of 2464 pixels in the vertical direction. The signal from each pixel is A / D converted after passing through a logarithmic converter, and the number of gradations is 1024.

First, the photographing of a subject will be described. When a photographing button (not shown) is pressed, the timing control unit 29 causes the X
The X-ray irradiator 1 is driven via the line adapter 30 to irradiate the subject 2 with a predetermined amount of X-rays. The X-rays are transmitted through the subject 2 and input to the radiation image conversion panel 3, and a latent image is formed according to the X-ray dose transmitted through the conversion panel 3. Then, when the photographing is completed and the stimulable excitation light from the stimulable excitation light source 4 is applied to the conversion panel 3, the conversion panel 3 emits light in proportion to the latent image energy, and the amount of this light is converted into the photoelectric converter 6. Is photoelectrically converted in and input to the A / D converter 7 as a current signal. This output current is converted into digital image data via the current / voltage converter 7A, the amplifier 7B and the A / D converter 7C of the A / D converter 7, and is input to the control circuit 7D. Then, it becomes digital image data and is input to the control circuit 7D. Then, the digital image data is transferred to the control circuit 7D.
Is transferred to and stored in the frame memory 24 via the timing control unit 29, the image bus VB, and the frame memory control unit 25. Then, the CPU 21 reads the control program from the memory 34 via the system bus SB, reads the radiation image data from the frame memory 24 via the image bus VB and the frame memory control unit 25, and according to the control program, the present invention is applied. Determination of a desired image area and image processing are executed as follows.

First, for the purpose of reducing the number of pixels and improving the processing speed, the image data of the original human chest side image is thinned to 1/16 in both the vertical and horizontal directions before the binarization processing, and the horizontal direction is reduced. Set the number of pixels to 128 and the number of vertical pixels to 154. Further, for the same purpose as the thinning processing, from the thinned whole image (rectangular area surrounded by broken line in FIG. 4) to the effective image area (rectangular area surrounded by solid line in FIG. 4) Make a decision based on the characteristics.

That is, in the case of the side of the chest of the human body, the first
In the right and left of the image, there is no subject (blank part), secondly there is an unnecessary upper part of the abdomen in the lower part of the image, and thirdly there is an unnecessary neck image in the upper part of the image. . Therefore, the effective image area consisting of 88 × 118 pixels is determined from the entire image except for 10 lines at the upper end, 26 lines at the lower end, and 20 lines at the left and right ends.

Here, in addition to the method of determining the effective image area by utilizing the features common to the images as described above,
It may be determined individually for each image based on profile information of the entire image. In particular, it is desirable to determine the effective image area so as to eliminate the blank portion as much as possible because the blank portion is binarized. Also, X
Image of a subject with a narrowed line field (see Fig. 6,
In the case where the hatched area is the irradiation field aperture area), the effective image area may be determined by analyzing the image data or detecting the irradiation field based on the signal from the X-ray irradiation device.

Then, from the image data of the effective image area, 2
Set the initial value of the threshold for the binarization process. here,
In the present embodiment, the first object is to detect the maximum image data (maximum signal value) of the lung field area as the area of interest, and in order to achieve this object, the threshold for binarization is set from the larger one. It is necessary to gradually decrease toward the optimum threshold value. Therefore, the initial value is determined by adding 100 to the maximum signal value of 8 × 8 pixels in the substantially central portion (hatched area in FIG. 4) of the effective image area. Here, when gradually decreasing the threshold value toward the optimum threshold value, the initial value needs to be larger than the optimum threshold value, so the maximum value of the signal values of the entire image that can be treated as data may be set as the initial value. However, at this time, the difference between the initial value and the optimum threshold is large and the processing time becomes long, which is not preferable.

As another example of the method of setting the initial value,
As shown in FIG. 7, the maximum value (see FIGS. 8 and 9) of the image data is obtained by removing the blank portions of several lines (for example, 40, 45 ... 65 lines) in the approximate center of the effective image area. To set the initial value, to obtain the histogram of the signal value in the effective image area as shown in Fig. 10, and to set the maximum value excluding the blank part of this histogram to the initial value, There is a method of setting a signal value of a predetermined percentage to an initial value.

On the other hand, when gradually increasing the threshold value from the smaller value toward the optimum threshold value, it is necessary to set the value smaller than the optimum threshold value to the initial value. Therefore, the minimum value of the predetermined range of the image data is set to the initial value. However, for the same reason as above, it is desirable to set the initial value according to the characteristics of each image. Then, the set threshold value and the signal value of each pixel are compared for each pixel, and pixels above the threshold value and pixels below the threshold value are distinguished. For example, an identification code of 1 is attached to pixels that are equal to or more than a threshold value, and an identification code of 0 is attached to pixels that are less than the threshold value. And

Then, the identification code of each pixel in the binarized effective image area is determined and labeled to divide into a plurality of areas which are a set of the same identification code. Specifically, in the present embodiment in which the threshold value is gradually decreased and set, since the area consisting of the pixel group having the pixel data equal to or larger than the threshold value is the area of interest, only the pixels having the identification code of 1 are labeled. However, for example, when three or more adjacent identification codes are 1 in succession, all the pixels are labeled with the same label A. This determination of continuity is made not only when the identification codes in the vertical and horizontal directions are the same, but also when the identification codes in the diagonal vertical direction are the same. Similarly, a different label is attached to each pixel whose identification code is 1. In this way, the region where the signal value is equal to or greater than the threshold value is labeled for each pixel group, and
For example, five areas A to E are determined as shown in FIG.

Here, when the threshold value is set to be gradually increased, the area consisting of the pixel group having the image data less than the threshold value is the area to be noticed, so that only the pixels having the identification code 0 are labeled. Then, for example, when two or more adjacent identification codes are 0 consecutively, all the pixels are labeled with the same label. Next, a desired image area is determined based on the labeled image areas A to E and the relative positional relationship of the image areas in the effective image area.

For example, as shown in FIG. 5, when the five areas A to E are labeled, the
A region (A, D, E in this example) that touches any one of the lower, left, and right ends is rejected, and the other regions B and C are determined as desired image regions. This is because the lung field region as the region of interest exists in the center of the image, so it should not be in contact with the edge of the image, and conversely, the void portion is always in contact with the edge of the image. .

When the desired image area does not exist, the threshold value is decreased and the binarization process is repeated. The reduction width of this threshold value is preferably about 1 to 50 steps. Further, as a method of determining the desired image area, there is a method of utilizing the statistical property of the data in the labeled image area in addition to the method based on the relative positional relationship. For example, it is a method of comparing the maximum value of the signal value in the labeled image area with the maximum value of the signal value in the effective image area and determining the desired image area depending on whether or not the difference is within a predetermined range. Of course, a plurality of determination methods may be used together to increase the accuracy, and this method is desirable.

Next, the image data in the determined B and C areas is statistically processed to extract the average value of the image data in the desired image area. This average value represents the substantially maximum image data (maximum signal value) S1 of the lung field region as the region of interest.
In addition, as statistical processing, in addition to the average value, the maximum value, the minimum value,
Intermediate values, histogram processing, etc. can be used. Further, although only the data in the desired image area is used in the present embodiment, an area wider than the desired image area is determined with the desired image area as the center or the end point, and the lung field area is approximated based on the data in this area. The maximum image data may be obtained. For example, as shown in FIG. 11, a rectangular area centered on the desired image area F is set. This method is effective when the desired image area is relatively small, and is particularly effective when the image data contains peculiar data such as noise.

Next, the gradation processing condition is determined based on the substantially maximum image data S1 of the extracted lung field region. Specifically, for example, as shown in FIG. 12, the gradation conversion table is determined so that the pixel corresponding to the maximum image data S1 in the lung field region has the density D1 (for example, the transmission density of 2.0) that is easy to read for diagnosis. . Then, the entire original radiation image data is subjected to gradation processing based on the determined gradation conversion table.

As described above, the image data of the radiation image is divided into a plurality of image areas A to E based on the threshold value, and the lung field area is determined from these image areas and their position information. Since the radiation image is gradation-processed based on this gradation processing condition, the optimum image processing data of the lung field can be secured, the region of interest can be easily seen and the stable finish can be achieved, and the diagnostic performance can be improved. In particular, since the gradation processing condition is determined based on the maximum signal value of the lung field, the gradation processing condition can be determined by the maximum signal value of each radiographic image. Compared to the case of determining the processing conditions, even if the data is different for each radiographic image due to the different imaging conditions, etc., optimum image processing data can be secured for each radiographic image,
This can also improve the diagnostic performance.

Further, since the calculation for determining the desired image area is simple, the processing speed can be increased and the image processing can be simplified. Next, a human chest front image is taken as an example of the subject, and a detailed description will be given based on FIGS. 13 to 16. FIG. 13 schematically shows a front image of the human chest, and the number of pixels and gradations of the original front image of the human body are the same as those of the side image of the human chest.

First, before performing the binarization process, a filtering process is performed by a filter having a size of 60 × 60 pixels to reduce the high frequency component for the purpose of reducing the binarization process abnormality due to noise. In order to improve the speed, the number of pixels is set to 128 in the horizontal direction and 154 in the vertical direction by performing 1/16 thinning processing in both the vertical and horizontal directions. The filtering process may be performed after the thinning process, and at this time, the filtering process speed can be improved. In this example, the entire radiation image is set as the effective image area, and the number of pixels in the effective image area is 128 in the horizontal direction and 1 in the vertical direction.
Set to 54.

Next, the initial value of the threshold value for the binarization process is set. In the present embodiment, the first purpose is to obtain the maximum image data (maximum signal value) of the lung field region as the region of interest, and the second is to obtain the minimum image data (minimum signal value) in the vertical / horizontal / abdominal region. To aim. In order to achieve these objects, it is necessary to gradually change the threshold for binarization from the larger one to the optimum threshold and at the same time to gradually change another threshold from the smaller one to the optimum threshold.

First, as the initial value P1 for obtaining the maximum image data of the lung field region, for example, the image data excluding the blank portions in the lines 50, 60, 70 and 80 at the approximate center of the effective image region is used. The maximum value is obtained and 100 is added to this maximum value to set it as the initial value P1. Then, after comparing the set threshold value with the signal value of each pixel and binarizing it in the same manner as the example of the chest side image, labeling processing is performed based on the identification code and, for example, as shown in FIG. One image area G to L is determined. However, in this example, since the image noise is processed in advance by the filtering process, labeling is performed when two or more identification codes of adjacent images are continuous.

Next, among the plurality of labeled image areas, the area (G to J in this example) which is in contact with any one of the upper, lower, left and right end portions of the effective image area is a blank portion. And the other K and L areas are determined as the desired image area. At this time, when the desired image region does not exist, the threshold value is decreased and the binarization process is repeated. Next, the image data in the determined K and L image areas is statistically processed to extract the maximum value of the image data in the desired image area. This maximum value is in addition to the approximate maximum value of the lung field region as the region of interest,
Average values, minimum values, intermediate values, histogram processing, etc. may be used.

Next, an initial value P2 of the threshold value for obtaining the minimum image data of the mediastinum / abdominal region is set. That is,
The minimum value of all the image data in the effective image area is calculated, 100 is subtracted from this minimum value, and the value is set to the initial value P2.
This is because in the front chest image, the image data of the entire mediastinum / thorax of the subject becomes the smallest. Then, after comparing the set threshold value with the signal value of each pixel and binarizing the same as in the example of the chest side image, labeling processing is performed based on the identification code. At this time, contrary to the example of the chest side image, pixels having a signal value equal to or less than the threshold value are given an identification code 1, and pixels having a signal value exceeding the threshold value are given an identification code 0. In addition, since the mediastinum / abdominal region has a small signal value and relatively large noise, when the identification codes of four or more pixels are consecutively the same, a label is attached to perform labeling processing. . In this way, for example, as shown in FIG. 15, three image areas M, N, O are determined.

Next, the desired image area is determined from the labeled image areas M, N, O based on the relative positional relationship in the effective image area. That is, of the labeled image areas, the area of M that touches the lower end of the effective image area is determined as the desired image area, and the areas of N and O that do not touch the lower end are rejected. This is because there is a region where the image data is the smallest near the lower end of the effective image region, and when there is a labeled region other than the lower end, the possibility of non-human body components such as a pacemaker is high. is there. At this time, when the desired image area does not exist, the threshold value is increased and the binarization processing is repeated.

Next, the image data in the determined desired image area M is subjected to cumulative histogram processing, and the signal value at the 20% position is extracted from the smaller one of the histogram. This signal value represents substantially the minimum image data (minimum signal value) S3 of the mediastinum / abdominal region as the region of interest. In addition to the histogram processing, maximum value, minimum value, intermediate value, average value processing, etc. of signal values can be used.

Next, the gradation processing condition is determined based on the substantially maximum signal value S2 of the lung field region and the substantially minimum signal value S3 of the mediastinum / abdominal region. That is, for example, as shown in FIG. 16, the pixels corresponding to the maximum signal value S2 of the lung field region and the pixel corresponding to the minimum signal value S3 of the mediastinum / abdominal region are respectively the densities D2 and D3 which are easy to interpret in diagnosis. (For example, the transmission density in the lung field
2.0, mediastinum / abdomen 0.3), set the gradation conversion table.

Then, the entire original radiation image data is subjected to gradation processing based on the gradation conversion table thus determined. Even in this case, the same effect as the image processing of the chest side image can be obtained. In the setting of the gradation conversion table, the basic gradation conversion table is used for the gradation processing by rotating or translating it on the coordinates as disclosed in JP-A-59-83149. You may do it.

In the above embodiment, gradation processing was described as an example of image processing. However, in addition to gradation processing, the conditions of spatial frequency processing as disclosed in Japanese Patent Publication No. 62-62376 are described. It may be set based on the image data extracted as described above. Further, the radiation image data subjected to the gradation processing according to the present invention may be visualized immediately by the image output device 15, but once stored in the image filing system,
Read out when necessary and output to film, or CRT
May be displayed on the screen.

When the radiation image is stored in the filing system, the processed radiation image data that has been subjected to the gradation processing according to the present invention may be stored. The gradation processing condition (gradation conversion table) determined based on the invention may be stored as a pair and the gradation processing may be performed at the time of reading.

In this embodiment, the stimulable phosphor is X.
Although the line detector is used and the image data obtained by photoelectrically reading the line detector is subjected to gradation processing, the system is not limited to a system using a stimulable phosphor, and a configuration using other radiation detectors or Alternatively, the radiation image recorded on the silver salt film may be photoelectrically read and gradation processing may be performed on the radiation image.

Further, in the above embodiment, the human chest is described as an example, but it may be an image mainly composed of bones such as legs or an image mainly composed of meat such as chest. ,
It does not limit the subject.

[0057]

As described above, according to the present invention, in claim 1, the image data of the radiation image is analyzed to determine the desired image area. Therefore, the image for determining the image processing condition is obtained. Since the region can be extracted optimally, the optimum image processing data can be secured, the processing speed and the apparatus can be simplified, and the optimum image processing can be performed without being affected by the photographing conditions and the like. Further, in claim 2, since the image data of the radiation image and the threshold value are compared to determine a plurality of image regions and then a desired image region is determined from those image regions, the image processing condition is determined. Therefore, the image area for can be extracted more optimally, and optimal image processing can be performed.

Further, in claims 3 and 4, after the image processing condition is determined from the image data of the desired image region determined by analyzing the radiation image data, the radiation image is imaged based on the image processing condition. Since the processing is performed, the optimum image processing can be performed and the diagnostic performance can be secured without being affected by the shooting conditions and the like. Further, in claim 5, the image data of the radiation image is compared with a threshold value to determine a plurality of image regions, and then a desired region is determined, and the image processing condition is determined based on the image data of the desired image region of the garden. Therefore, the optimum image processing condition can be determined and the optimum image processing can be performed.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing an embodiment of the present invention.

FIG. 2 is a hardware configuration diagram of main parts of the above.

FIG. 3 is a functional block diagram of the above.

FIG. 4 is a diagram showing an example of the same effect.

5 is an enlarged view of a main part of FIG.

FIG. 6 is a diagram when the irradiation field in the above is narrowed down.

FIG. 7 is a diagram showing another example of initial value setting described above.

FIG. 8 is a signal characteristic diagram of a predetermined line in FIG.

FIG. 9 is a signal characteristic diagram of another predetermined line of FIG.

[Fig. 10] Histogram of data in the effective image area

FIG. 11 is a diagram showing another example of obtaining the maximum image data of the above.

[FIG. 12] Same as the above gradation conversion characteristic diagram

FIG. 13 is a diagram showing another example of the same effect.

FIG. 14 is a diagram for obtaining the lung field region of the same as above.

FIG. 15 is a diagram for obtaining the mediastinum / abdominal region in the same as above.

FIG. 16 is a gradation conversion characteristic diagram of the same as above.

[Explanation of symbols]

1 X-ray irradiation device 2 subject 3 Radiation image conversion panel 8 Image processing device

Claims (5)

[Claims] 1. A desired image area is determined by analyzing image data of the radiation image formed from image data of the radiation image formed corresponding to the amount of radiation transmitted through each part of a subject. A method for determining an image area of a radiation image, which is characterized in that: 2. A desired image area is divided into a plurality of image areas by comparing the image data of the radiation image with a threshold value, and then determined from the plurality of divided image areas. Image area determination method for radiographic images of humans. 3. When performing image processing on a radiation image formed corresponding to the amount of radiation transmitted through each part of a subject,
The image data of the radiation image is analyzed to determine a desired image area, the image processing condition is determined based on the image data of the determined desired image area, and the radiation image is determined based on the determined image processing condition. An image processing method for a radiation image, wherein the image processing is performed. 4. A radiographic image processing apparatus adapted to image-process a radiographic image formed corresponding to the amount of transmission of radiation passing through each part of a subject, wherein image data of the radiographic image is analyzed and a desired image area is defined. Image region determining means for determining, image processing condition determining means for determining image processing condition based on image data of the determined desired image region, and image for image processing the radiation image based on the determined image processing condition A radiation image processing apparatus comprising: a processing unit. 5. The image area determining means compares the image data of the radiation image with a threshold value to divide the radiation image into a plurality of image areas, and determines a desired image area from the plurality of divided image areas. The radiographic image processing device according to claim 4.