JP2001238868A - Method of image processing and its apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable each domain to output an image with an appropriate density by analyzing the domains of an object and by subdividing each domain more minutely. SOLUTION: After obtaining a radiated image created based on radioactive rays which have passed through a breast, a recognition of the domains of the breast is made. By analyzing the inside domains of the breast which have already been recognized, the classification of the inside domains by characteristic features will be made. Based on the result of this classification, a standard signal value will be decided on which an image processing will be done.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an image processing method and an image processing apparatus for obtaining a radiation image from radiation that has passed through a subject.

[0002]

2. Description of the Related Art In recent years, devices capable of directly taking a radiation image as a digital image have been developed. For example, as a device that detects the amount of radiation applied to a subject and obtains a radiation image formed in accordance with the detected amount as an electric signal, a method using a detector using a stimulable phosphor is disclosed in Japanese Patent Application Laid-Open No. H10-163,873. No. 55-12429, JP-A-63-18985
Many publications such as Japanese Patent Publication No. 3 are disclosed. In such an apparatus, a stimulable phosphor is applied to a sheet-like substrate, or a detector fixed by vapor deposition or the like is once irradiated with radiation that has passed through a subject, so that the stimulable phosphor absorbs the radiation. Then, by exciting the stimulable phosphor with light or heat energy, the radiation energy accumulated by absorption by the stimulable phosphor is emitted as fluorescence, and the fluorescence is photoelectrically converted to form an image signal. I'm trying to get.

On the other hand, an electric charge corresponding to the intensity of the irradiated radiation is generated in a photoconductive layer, the generated electric charge is stored in a plurality of two-dimensionally arranged capacitors, and the stored electric charge is taken out. Has been proposed. Such a radiation image detecting apparatus uses what is called a flat panel detector (FPD). This type of FPD is disclosed in JP-A-9-90.
As described in Japanese Patent Application Publication No. 048, fluorescence can be detected by a photodiode or a CCD or C-MOS sensor. A similar FPD is described in Japanese Patent Application Laid-Open No. 6-342098.

[0004] In these apparatuses, in a process of obtaining a radiation image generated based on radiation passing through a subject, for example, as described in JP-A-63-31640, a cumulative histogram is created and a desired histogram is created. An image signal range is obtained, the image signal range is corrected according to the characteristic value of the cumulative histogram, and a gradation processing condition is determined so that the corrected image corresponds to a predetermined electrical image signal range. 0
As described in Japanese Patent Application Laid-Open No. 6-130518, an image processing condition is determined from a cumulative histogram of an area at a certain distance from an area excluding a direct radiation part by a cumulative histogram.

[0005]

However, when a cumulative histogram is obtained and conditions are determined based on the cumulative histogram or the like, it greatly depends on the characteristics of the image. It is difficult to output the cumulative histogram at a stable density because the cumulative histogram is an imaged region having a very large individual difference such as the amount of the mammary gland present in the breast region.

[0006] Further, since the mammogram has a wide interpretation point, it is necessary to analyze the subject area, divide the subject area into smaller areas, and output each area accurately.

In particular, when a very large aneurysm shadow exists in the breast region, the density in the breast region is very low (whitish). The finished image has a higher density (blackish) than when there is no large seed nodule shadow.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and it is an object of the present invention to analyze a subject area and further divide the subject area into smaller areas so that each area can be output with an appropriate density. It is an object.

[0009]

Means for Solving the Problems In order to solve the above problems and achieve the object, the present invention has the following constitution.

[0010] The first aspect of the present invention is an image processing method for processing a radiation image generated based on radiation that has passed through a breast. The image processing method includes the steps of: Image processing characterized by analyzing the breast region into characteristic regions, determining a reference signal value based on the classification result, and performing image processing based on the reference signal value. Method. ].

According to the first aspect of the present invention, it is possible to perform stable image processing by analyzing the inside of a breast region as a subject and determining a reference signal value in a region necessary for diagnosis. Become. Further, the method of analyzing the subject and determining the reference signal value based on the analysis result can set the reference signal value to a point where the reference signal value is desired to be determined more accurately than the case where the reference signal value is determined based on a cumulative histogram or the like of the entire subject region.

According to a second aspect of the present invention, there is provided the image processing method according to the first aspect, wherein the analysis in the breast region determines a threshold value in a local region. ].

According to the second aspect of the present invention, the threshold value in the local region is determined and binarized, so that the detection accuracy of the edge portion where the density greatly changes is improved, and the edge of the mammary gland region is particularly improved. It is possible to accurately extract a part, a pectoral muscle region, and the like.

According to a third aspect of the invention, there is provided the image processing method according to the first aspect, wherein the reference signal value is determined based on pixels around a mammary gland region. ].

According to the third aspect of the present invention, since the detection accuracy of the edge portion where the density greatly changes is high, the peripheral portion of the mammary gland region can be accurately extracted. Therefore, by setting the reference signal value at the periphery of the mammary gland region, it is possible to output a stable density.

According to a fourth aspect of the present invention, there is provided the image processing method according to the first aspect, wherein the reference signal value is determined based on a pectoral muscle region. ].

According to the fourth aspect of the present invention, since the detection accuracy of the edge portion where the density greatly changes is high, the peripheral portion of the pectoral muscle region can be accurately extracted. Therefore, by setting the reference signal value in the pectoral muscle region, it is possible to output with a stable density.

According to a fifth aspect of the present invention, there is provided the image processing method according to the first aspect, wherein the reference signal value is determined based on a fat area. ].

According to the fifth aspect of the present invention, since the detection accuracy of the edge portion where the density largely changes is high, the fat region can be accurately extracted. For this reason, by setting the reference signal value in the fat area, it is possible to output a stable density.

According to a sixth aspect of the present invention, there is provided the image processing method according to the first aspect, wherein the reference signal value is determined based on the vicinity of the periphery of the breast. ].

According to the sixth aspect of the present invention, since the detection accuracy of the edge portion where the density largely changes is high, the vicinity of the periphery of the breast can be accurately extracted. This other
By setting the reference signal value at the periphery of the mammary gland region, it is possible to output with a stable density.

According to a seventh aspect of the present invention, there is provided a method according to the first aspect, wherein, among the reference signal values, an area in which one or more reference signal values are determined in accordance with a feature of an image is automatically determined. The image processing method according to any one of claims 1 to 6. ].

According to the seventh aspect of the present invention, among the reference signal values, one or more regions for determining the reference signal values can be automatically determined in accordance with the characteristics of the image, so that the stable density and the stable Can be output in the key.

[0024] The invention according to claim 8 is characterized in that "one or more reference signal value determination areas among the reference signal values can be manually determined." The image processing method according to 1. ].

According to the eighth aspect of the present invention, one or more reference signal value determination areas among the reference signal values can be manually determined, and the point at which the interpreting doctor freely determines the reference signal value is determined. Since the image can be changed, a doctor's favorite image to be interpreted can be output.

According to a ninth aspect of the present invention, there is provided the method according to the third aspect, wherein the area for determining the reference signal value is determined in advance, and the reference signal value is determined from the determined area for determining the reference signal value. Item 7. The image processing method according to any one of Items 6. ].

According to the ninth aspect of the present invention, since an area for determining a reference signal value can be stored in advance, image processing can be performed according to an area to be diagnosed which is known in advance. Image processing can be performed.

According to a tenth aspect of the present invention, there is provided an image processing apparatus for processing a radiation image generated based on radiation passing through a breast, wherein the breast area recognizing means for recognizing a breast area, and the breast area recognizing means. Means for classifying the inside of the breast region into characteristic regions by analyzing the inside of the breast region recognized by the method, and a reference for determining a reference signal value based on the result of the classification by the inside breast region classification means An image processing apparatus, comprising: signal value determination means, and performing image processing based on the reference signal value. ].

According to the tenth aspect of the present invention, it is possible to perform stable image processing by analyzing the inside of a breast region as a subject and determining a reference signal value in a region necessary for diagnosis. Become. Further, in the method of analyzing a subject and determining the reference signal value based on the analysis result, the reference signal value can be set to a point where the reference signal value is desired to be determined more accurately than when the reference signal value is determined based on a histogram of the entire subject area.

According to an eleventh aspect of the present invention, there is provided an image processing apparatus according to the tenth aspect, wherein the intra-breast region classification unit includes a local region threshold value determination unit that determines a threshold value in a local region.
An image processing apparatus according to claim 1. ].

According to the eleventh aspect of the present invention, the threshold value in the local area is determined and binarized,
In particular, the detection accuracy of the edge portion where the density greatly changes is improved, and the region and the like can be accurately extracted.

According to a twelfth aspect of the present invention, in the image processing apparatus according to the tenth aspect, the reference signal value is determined based on pixels around a mammary gland region. ].

According to the twelfth aspect of the present invention, since the detection accuracy of the edge portion where the density greatly changes is high, the peripheral portion of the pectoral muscle region can be accurately extracted. Therefore, by setting the reference signal value in the pectoral muscle region, it is possible to output with a stable density.

According to a thirteenth aspect of the present invention, the image processing apparatus according to the tenth aspect, wherein the reference signal value is determined based on a pectoral muscle region. ].

According to the thirteenth aspect of the present invention, since the detection accuracy of the edge portion where the density greatly changes is high, the peripheral portion of the pectoral muscle region can be accurately extracted. Therefore, by setting the reference signal value in the pectoral muscle region, it is possible to output with a stable density.

According to a fourteenth aspect of the present invention, the image processing apparatus according to the tenth aspect, wherein the reference signal value is determined based on a fat area. ].

According to the fourteenth aspect of the present invention, since the detection accuracy of the edge portion where the density largely changes is high, the fat region can be accurately extracted. For this reason, by setting the reference signal value in the fat area, it is possible to output a stable density.

According to a fifteenth aspect of the present invention, the image processing apparatus according to the tenth aspect, wherein the reference signal value is determined based on the vicinity of the periphery of the breast. ].

According to the fifteenth aspect, since the detection accuracy of the edge portion where the density changes greatly is high, it is possible to accurately extract the vicinity of the periphery of the breast. However, by setting the reference signal value at the periphery of the mammary gland region, it is possible to output with a stable density.

According to a sixteenth aspect of the present invention, there is provided a method for automatically selecting a reference signal value which can automatically determine at least one region of the reference signal values in accordance with the characteristics of an image. The image processing apparatus according to claim 10, wherein the image processing apparatus includes: ].

According to the sixteenth aspect of the present invention, among the reference signal values, one or more regions for determining the reference signal values can be automatically determined in accordance with the characteristics of the image, so that the density and the stabilization are stable. Can be output in the key.

According to a seventeenth aspect of the present invention, there is provided an image forming apparatus comprising: a manual reference signal value determining area selecting means for manually determining one or more reference signal value determining areas among the reference signal values. The image processing apparatus according to claim 12. ].

According to the seventeenth aspect of the present invention, one or more reference signal value determination areas among the reference signal values can be manually determined, and the point at which the interpreting doctor freely determines the reference signal value is determined. Since the image can be changed, a doctor's favorite image to be interpreted can be output.

The invention according to claim 18 is characterized in that "the reference signal value determining area storage means for storing in advance the area for determining the reference signal value, and the reference signal value is determined by the reference signal value determining area storing means. The image processing apparatus according to any one of claims 12 to 15, wherein the processing is performed. ].

According to the eighteenth aspect of the present invention, since an area for determining a reference signal value can be stored in advance, image processing can be performed according to an area to be diagnosed which is known in advance. Image processing can be performed.

[0046]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, embodiments of the present invention will be described in detail with reference to the drawings, but the present invention is not limited to the description of the embodiments and the drawings.

FIG. 1 is a diagram showing the configuration of the image processing apparatus. The radiation generator 30 is controlled by the controller 10, and the radiation emitted from the radiation generator 30 is applied to the imaging panel mounted on the front of the radiation image reader 40 through the subject 5.

FIG. 2 shows the structure of the image pickup panel. The imaging panel 41 has a substrate having a thickness enough to obtain a predetermined rigidity, and a detection element 412 that outputs an electric signal in accordance with the dose of the irradiated radiation is provided on the substrate.
_{-(1,1) to} 412- _{(m, n)} are two-dimensionally arranged. Also,
Scan lines 415 _{-1 to} 415 _-m and signal lines 416 _{-1 to} 416 _-n
Are arranged, for example, orthogonally.

Scanning line 415 of imaging panel 41_-1~ 415
_-mAre connected to the scanning drive unit 44. Scan driver 4
4 to scanning line 415_-1~ 415_-mOne of the scan lines 4
Fifteen _-p(P is any value from 1 to m)
When S is supplied, this scanning line 415_-pThe test connected to
Electrical signal S corresponding to the dose of radiation emitted from the output element
V_-1~ SV_-nIs detected, and the signal line 416 is detected._-1~ 416_-nTo
The image data is supplied to the image data generation unit 46 via the image data generation unit 46.

The detection element 412 may be any element that outputs an electric signal corresponding to the dose of the irradiated radiation. For example, when a detection element is formed using a photoconductive layer in which an electron-hole pair is generated upon irradiation with radiation and the resistance value changes, an amount corresponding to the amount of radiation generated in the photoconductive layer is used. Is stored in the charge storage capacitor, and the charge data stored in the charge storage capacitor is supplied to the image data generator 46 as an electric signal. The photoconductive layer desirably has a high dark resistance value, and is preferably amorphous selenium, lead oxide, cadmium sulfide, mercuric iodide,
Or, a photoconductive organic material (including a photoconductive polymer to which an X-ray absorbing compound is added) is used,
Particularly, amorphous selenium is desirable.

When the detection element 412 is formed using, for example, a scintillator or the like that generates fluorescence when irradiated with radiation, a photodiode generates an electric signal based on the intensity of the fluorescence generated by the scintillator. The information may be supplied to the image data generator 46. The image data generation unit 46 sequentially selects the supplied electric signals SV based on the output control signal SC from the reading control unit 48 and converts them into digital image data DT. The image data DT is supplied to the reading control unit 48.

The reading control unit 48 is connected to the controller 10.
In the reading control unit 48, the controller 10
Scanning control signal R based on the control signal CTD supplied from
C and an output control signal SC are generated. This scanning control signal C
The scan control signal is supplied to the scan drive unit 44 based on the TD to control the scan.
Scan line 415 based on signal RC_-1~ 415_-mAgainst
The supply of the read signal RS is performed. Further, the output control signal S
C is supplied to the image data generator 46. This reading control
The scanning control signal RC and the output control signal SC from the section 48
Therefore, for example, the imaging panel 41 is (m × n) as described above.
When the detection element 412 is used, the detection element
Child 412_-(1,1)~ 412_{-(m, n)}Based on the electric signal SV from
Data DP_(1,1)~ DP_{8m, n)}Then
Data DP _(1,1), DP_(1,2), ... DP_{(1, n)}, D
P_(2,1), ..., DP_{(m, n)}Image data as order
DT is generated, the image data is generated, and
The image data DT is transferred from the image data generation unit 46 to the reading control unit.
48. Also, the reading control unit 48
Processing for transmitting image data DT to the controller 10 is also performed.
U.

The image data DT obtained by the radiation image reader 40 is supplied to the controller 10 via the reading controller 48. When the image data obtained by the radiation image reader 40 is supplied to the controller 10 and the log-converted image data is supplied, the controller 10
Can simplify the processing of the image data.

The radiation image reader 40 is not limited to the one using the FPD, but may use a stimulable phosphor.

FIG. 3 shows a configuration in which a radiation image reader 60 using a stimulable phosphor is used. In a conversion panel 61 irradiated with radiation, a stimulable phosphor layer is provided on a support. Is provided by a vapor phase deposition of a stimulable phosphor or a stimulable phosphor paint. This stimulable phosphor layer is shielded or covered by a protective member in order to block adverse effects and damage due to the environment.

The light beam generator (gas laser, solid-state laser, semiconductor laser, etc.) 62 generates a light beam whose emission intensity is controlled. This light beam reaches the scanning unit 63 via various optical systems, is deflected by the scanning unit 63, is further deflected by the reflecting mirror 64, and is guided to the conversion panel 61 as stimulating excitation scanning light. .

At the light-collecting end of the light-collecting body 65 formed of an optical fiber or a sheet-like light guide member, stimulating light excitation light is guided as scanning light. The light-collecting end of the light-collecting body 65 formed of an optical fiber or a sheet-shaped light guide member is disposed close to the conversion panel 61 on which the stimulating light is scanned, and scans the light beam from the light beam generator 62. As a result, stimulated emission having an emission intensity proportional to the latent image energy generated in the conversion panel 61 is received.

The filter 66 passes only the light in the stimulating emission wavelength region from the light introduced from the light collector 65, and the light passing through the filter 66 enters the photomultiplier 67. The photomultiplier 67 generates a current signal corresponding to incident light by photoelectric conversion. This current signal is supplied to the current / voltage converter 70 and converted into a voltage signal. Further, after the voltage signal is amplified by the amplifier 71, the voltage signal is converted into digital image data DT by the A / D converter 72. Here, a logarithmic conversion amplifier (log amplifier) is used as the amplifier 71. The image data DT is sequentially image-processed in the image processing device 80, and the image data DTC after the image processing is transmitted to the printer 83 via the interface 82.

CPU (Central Process)
ing Unit) 81 is for controlling image processing in the image processing device 80,
In the case of 0, various image processings (for example, spatial frequency processing, dynamic range compression, gradation processing, enlargement / reduction, movement, rotation, statistical processing, etc.) are performed on the image data DT, and an image in a form suitable for diagnosis is performed. Generate data DTC. This image data DTC is supplied to the printer 83 and
From 3, a hard copy of the radiation image of each part of the human body can be obtained. Note that a monitor such as a CRT may be connected to the interface 82, and a storage device (filing system) capable of storing image data of a plurality of radiation images (filing system) may be connected.

The reading controller 75 adjusts the light beam intensity of the light beam generator 62, adjusts the gain of the photomultiplier 67 by adjusting the power supply voltage of the high voltage power supply 76 for the photomultiplier, and amplifies the current / voltage converter 70. Part 71
Is adjusted, and the input dynamic range of the A / D converter 72 is adjusted, so that the reading gain is comprehensively adjusted. Image data DT obtained from the A / D converter 72
Is supplied to the controller 10 and controls the operation of the reading control unit 75 according to the control signal CTD from the controller 10.

The radiation image reader irradiates a silver halide film on which a radiation image is recorded with light from a light source such as a laser or a fluorescent lamp, and photoelectrically converts the transmitted light of the silver halide film to generate image data. May be. Further, a configuration may be employed in which radiation energy is directly converted into an electric signal using a radiation quantum counting type detector to generate image data.

Next, the configuration of the controller 10 is shown in FIG. CPU for controlling operation of controller 10
(Central Processing Unit)
11 is connected to a system bus 12 and an image bus 13, and an input interface 17. CPU 11 for controlling the operation of this controller 10
The operation of is controlled based on a control program stored in the memory 14. System pulse 12 and image bus 13
Are connected to a display control unit 15, a frame memory control unit 16, an output interface 18, a shooting control unit 19, a disk control unit 20, and the like. The operation of each unit is controlled by the CPU 11 using the system bus 12. The transfer of image data between the units via the image bus 13 is performed.

A frame memory 21 is connected to the frame memory control unit 16, and image data obtained by the radiation image reader 40 is stored via the imaging control unit 19 and the frame memory control unit 16. The image data stored in the frame memory 21 is read and supplied to the display control unit 15 and the disk control unit 20. The frame memory 21 may store the image data supplied from the radiation image reader 40 after the CPU 11 processes the image data.

An image display device 22 is connected to the display control unit 15, and a radiographic image based on the image data supplied to the display control unit 15 is displayed on the screen of the image display device 22. Here, when the number of display pixels of the image display device 22 is smaller than the number of pixels of the radiation image reader 40, the entire captured image can be displayed on the screen by thinning out and reading out the image data. In addition, if the image data of the area corresponding to the number of display pixels of the image display device 22 is read out, the captured image at the desired position can be displayed in detail.

From the frame memory 21 to the disk controller 2
When the image data is supplied to the disk controller 20, for example, the image data is read continuously and
The data is written to the O memory, and then sequentially recorded on the disk device 23. Further, image data read from the frame memory 21 and image data read from the disk device 23 can be supplied to the external device 90 via the output interface 18.

In the image processing section 26, the radiation image reader 4
0 to the image data D supplied via the photographing control unit 19
T irradiation field recognition processing, region of interest setting, normalization processing, gradation processing, and the like are performed. Further, frequency emphasis processing, dynamic range compression processing, and the like may be performed. In addition, as a configuration in which the CPU 11 also serves as the image processing unit 26,
Image processing and the like can also be performed.

The input interface 17 is connected to an input device 27 such as a keyboard. By operating the input device 27, management information such as information for identifying image data obtained by shooting and information on shooting is input. As the external device 90 connected to the output interface 18, a scanning laser exposure device also called a laser imager is used. In this scanning laser exposure apparatus, the intensity of a laser beam is modulated by image data, exposed to a conventional silver halide photographic material or a thermal phenomenon silver halide photographic material, and then subjected to an appropriate development process to thereby produce a radiation image. A hard copy is obtained.

Although the frame memory 21 stores the image data supplied from the radiation image reader 40, the image data supplied may be processed by the CPU 11 and then stored. Further, the disk device 23 stores image data stored in the frame memory 21, that is, image data supplied from the radiation image reader 40 and image data obtained by processing the image data by the CPU 11 together with management information and the like. be able to.

Next, the operation will be described. When obtaining a radiation image of the subject 5, it is assumed that the subject 5 is located between the radiation generator 30 and the imaging panel 41 of the radiation image reader 40, and the radiation emitted from the radiation generator 30 is And is incident on the radiation image capturing panel 41 transmitted through the subject 5. In addition, since the radiation image reader 40 is used instead of the radiation image reader 40, the description of the case where the radiation image reader 60 is used is omitted.

The controller 10 is input with management information indicating the identification of the subject 5 to be photographed and information relating to the photographing using the input device 27. The input of the management information using the input device 27 is performed by operating a keyboard, using a magnetic card, a bar code, an HIS (In-Hospital Information System: information management by network), and the like.

The management information includes, for example, an ID number, a name,
Date of birth, gender, date and time of imaging, location and position of imaging (for example, which part of the human body was irradiated with radiation from which direction), imaging method (simple imaging, contrast imaging, tomography, magnified imaging, etc.), imaging It consists of information such as conditions (tube voltage, tube current, irradiation time, use of scattered radiation removal grid, etc.).

The date and time of photographing can be obtained automatically from the CPU 11 by using a clock function built in the CPU 11. The management information to be input may be only information relating to the subject to be photographed at that time. A series of management information may be input in advance, and the subject may be photographed in the input order, or the management information may be input as needed. The information may be read and used.

When the power switch of the radiation image reader 40 is turned on, the control signal C
Based on the TD, the reading control unit 48 of the radiation image reader 40
The scanning panel 44 initializes the imaging panel 41. This initialization is for obtaining a correct electric signal corresponding to the radiation dose emitted from the imaging panel 41.

When the initialization of the imaging panel 41 is completed in the radiation image reader 40, the radiation from the radiation generator 30 can be irradiated. Here, when a switch for irradiating radiation is provided in the radiation generator 30, when the switch is scanned, the radiation
Radiation is irradiated for a predetermined time, and a signal DFS indicating the start of irradiation and a DFE indicating the end of irradiation
Is supplied to the controller 10.

At this time, the radiation dose of the radiation applied to the imaging panel 41 of the radiation image reader 40 is modulated by the subject 5 because the degree of absorption of the radiation by the subject 5 differs. Detection element 412 of imaging panel 41- _(1,1)
In ４ 412- _{(m, n)} , an electric signal based on the radiation modulated by the subject 5 is generated.

Next, in the controller 10, after a predetermined time has elapsed since the signal DFS was supplied, for example, when the irradiation time of the radiation is about 0.1 second, a time longer than the irradiation time (for example, about 1 second) elapses. Or immediately after the signal DFE is supplied, the radiation image reader 40 outputs the image data D
The control signal CTD is supplied to the reading control unit 48 of the radiation image reader 40 to start generation of T.

On the other hand, when a switch for irradiating radiation is provided in the controller 10, when this switch is operated, an irradiation start signal CST for starting irradiation of radiation is transmitted via the imaging control unit 19. The radiation is supplied to the radiation generator 30, and the radiation is emitted from the radiation generator 30 toward the subject 5 for a predetermined time. The irradiation time is set based on, for example, management information.

Next, the controller 10 outputs a control signal CT for starting the generation of image data in the radiation image reader 40 a predetermined time after outputting the irradiation start signal CST.
D is supplied to the reading control unit 48 of the radiation image reader 40. Note that the controller 10 detects the end of radiation irradiation by the radiation generator 30 and then sets the radiation image reader 4
Control signal CTD for starting generation of image data at 0
May be supplied to the radiation image reader 40.
In this case, generation of image data during irradiation of radiation can be prevented.

The reading control section 48 of the radiation image reader 40 generates a scanning control signal RC and an output control signal SC based on a control signal CTD for starting generation of image data supplied from the controller 10. The scanning control signal RC is supplied to the scanning driving unit 44, and the output control signal SC is supplied to the reading control unit 48 with the image data DT. This image data DT is sent to the controller 10 by the reading control unit 48.

The image data DT supplied to the controller 10 is stored in the frame memory 21 via the photographing control unit 19, the frame memory control unit 16 and the like. Using the image data stored in the frame memory 21, a radiographic image can be displayed on the image display device 22. Further, by supplying the image data stored in the frame memory 21 to the display control unit 15, the brightness, contrast, sharpness, and the like are adjusted, and a radiation image suitable for diagnosis or the like can be displayed. In addition, by supplying the image data subjected to the image processing to the external device 90, a hard copy of a radiation image suitable for diagnosis or the like can be obtained.

The image processing unit 26 obtains a radiation image generated based on the radiation that has passed through the subject. For example, as described in JP-A-63-31640, a cumulative histogram is created. Determine the desired image signal range, correct the image signal range according to the characteristic value of the cumulative histogram, determine the gradation processing conditions so that the corrected image corresponds to a predetermined electrical image signal range, As described in Japanese Patent Application Laid-Open No. 06-130518, an image processing condition is determined from a cumulative histogram of an area at a certain distance from an area excluding a direct radiation part by a cumulative histogram.

When a cumulative histogram is obtained and conditions are determined based on the cumulative histogram or the like, the condition largely depends on the characteristics of the image. It is difficult to output the cumulative histogram at a stable density because the cumulative histogram is an imaged region having a very large individual difference such as the amount of the mammary gland present in the breast region.

Further, since the mammogram has a wide interpretation point, it is necessary to analyze the inside of the subject area, divide the subject area into smaller areas, and output each area accurately. In particular, when a very large shade is present in the breast area, the density in the breast area is very low (whitish). Compared to the case where no shadow exists, the finished image has a high density (blackish).

Therefore, in the image processing apparatus of the present invention, the image processing section 26 is configured as shown in FIG.

The image processing unit 26 of the embodiment shown in FIG.
Is a breast region recognizing unit 100 for recognizing a breast region, an in-breast region analyzing unit 101 for analyzing the inside of the breast region recognized by the breast region recognizing unit 100, and an analysis result obtained by the breast region recognizing unit 101. And a reference signal value determining unit 103 for determining a reference signal value based on information in the region of interest. The image processing unit 104 determines the reference signal value based on the reference signal value. Perform image processing. The intra-breast region analyzing means 101 has a local area threshold value determining means 101a for determining a threshold value in a local area.

The reference signal value is determined by any one of the periphery of the mammary gland region, the pectoral muscle region, the fat region, and the vicinity of the breast margin.

The reference signal value determining means 103 includes a reference signal value automatic selecting means 103a which can automatically determine one or more reference signal value determining areas in the reference signal value according to the characteristics of the image. By automatically determining an area in which one or more reference signal values are determined according to the characteristics of an image among the reference signal values, it is possible to output with stable density and gradation.

The reference signal value determining means 103 has manual reference signal value determining area selecting means 103b which can manually determine one or more reference signal value determining areas among the reference signal values. Of these, one or more reference signal value determination areas can be manually determined, and the doctor who interprets the image can freely change the point at which the reference signal value is determined, so that the doctor's favorite image can be output.

As described above, the breast region is recognized by the breast region recognizing unit 100, the inside of the recognized breast region is analyzed by the breast region analyzing unit 101, and based on the analysis result, the region of interest is determined by the region of interest determining unit 102. To determine.

A method for recognizing a breast region by the breast region recognizing means 100 will be described. First, a histogram of the entire image is obtained, and a threshold value is determined by a discriminant analysis method. Here, an area lower than the threshold is defined as a breast area. By determining the threshold in the local region, the inside of the breast region is analyzed and classified into several characteristic regions.

The method will be described below. As shown in FIG. 6, a local region W having a predetermined size is determined. Local region W
Is, for example, a square of 28 mm to 112 mm (56 mm
Square) is desirable, but not limited thereto. Threshold value Th based on information obtained from the pixel value of local region W
Is determined, and the area is binarized.

Similarly, a local area W is set at a position separated by a predetermined distance D.
Is set, a threshold is determined, and binarization is performed. The same processing is repeated for the entire image. As a result, a binarized pixel can be obtained. In this way, by determining and binarizing the threshold value in the local region, in particular, the detection accuracy of the edge portion where the pixel value greatly changes is increased, and the margin of the mammary gland region, the pectoral muscle region, and the like are accurately detected. Can be extracted.

The local region W is preferably between 50 mm and 150 mm, but is not limited to this range. As other methods for determining the size of the local region W, the size may be determined quantitatively or at a fixed rate.
It may be changed depending on the shooting method, the type of the image generation device, and the like. Further, the predetermined distance D is preferably changed according to the local area W. For example, it is preferable that the predetermined distance D moves so as to overlap with the next local area, such as moving by a half distance of the local area W.

Hereinafter, an example of a method for recognizing each region of the breast region will be described with reference to FIG.

When recognizing the mammary gland region D1 and the pectoral muscle region D2, a threshold value determined by subtracting a threshold value determined for each local region from the average value Ave of the pixel values of the local region W by a fixed amount may be used as the threshold value. The average value Ave of the pixel values of the local area W that is missing a certain percentage R may be used as the threshold. In this case, the ratio R is set to a value smaller than 1. Here, the low density area is an area having a small pixel value, and the high density area is an area having a large pixel value. At this time, the area smaller than the threshold becomes the mammary gland area D1 or the pectoral muscle area D2.

By this method, the mammary gland region D1 and
The pectoral region D2 can be recognized, but the papilla D
3. The breast region D1 and the pectoral muscle region D2 can be separated based on the position information of the recognized candidate using information such as the imaging direction.

Further, by performing a morphological contraction process on the detected mammary gland region D1, the periphery of the mammary gland region D1 can be extracted. The morphological contraction process is, for example, a process of setting a specified pixel from an edge of a binarized region outside the region.

Around the skin line L and a fat area D6
When recognizing is, a value obtained by adding a threshold determined for each local region to the average value Ave of the pixel values of the local region W by a fixed amount may be used as the threshold, or a fixed ratio The value multiplied by R may be used as the threshold. In this case, the ratio R is set to a value larger than 1. At this time, an area larger than the threshold is near the skin line L.

According to this method, the vicinity of the skin line L,
In addition, the fat area D6 can be recognized. Further, of the areas near the skin line L and the fat area D6, the area near the end of the already recognized breast area D4 is set as the vicinity of the skin line L, and the other areas are the fat area D6.
6 can be set.

Which of these areas is used for setting the reference signal value may be determined by extracting the features of the image.

FIG. 8A shows an image photographed in the MLO direction (a direction in which the breast is obliquely sandwiched), and is characterized in that the pectoral muscle region is large and appears in the image. Also,
FIG. 8B is a photograph taken in the CC direction (a direction in which the breast is vertically sandwiched) and is characterized in that it has a shape close to a semicircle.

When the recognized pectoral muscle area is large together with the left and right breast images, the reference signal value may be determined according to the pectoral muscle area. If the area of the mammary gland is very small,
It may be determined that the mammary gland is regressed, and the reference signal value may be determined in the fat area. If both the mammary gland region and the fat region have appropriate areas, both may be set as reference signal values.

At this time, it is preferable to have a storage device for storing an area in which the reference signal value is set, so that it is possible to know from which area the reference signal value is obtained when performing image processing. By doing so, it is possible to set and output a desired density for each characteristic region, so that more accurate image processing can be expected. In particular, in the mammogram, the interpretation point is wide, and the regression state of the mammary gland area, the presence or absence of a lesion shadow, the imaging part whose image properties vary greatly depending on the imaging direction, etc. It is important to be able to output the area with the correct density.

When it is desired to enhance the contrast of a mammary gland region or the contrast of a fat region, for example, when it is desired to particularly emphasize each of the divided specific regions, it is not necessary to perform the processing with one image, but to emphasize each of them. In the case where separate images are created and output to a monitor in the process of performing the processing, the screen may be divided and displayed, or the screen may be switched and displayed. When outputting to a film, each image may be output to one or a plurality of films.

Further, the merits of classifying and outputting the respective regions in the mammogram are shown below.

The method of determining the reference signal value using the histogram is susceptible to the influence of a large tumor or the regression state of the mammary gland region. However, by obtaining the reference signal value in the periphery of the mammary gland region, the influence thereof is relatively small. It is possible to output with less density and stable density.

[0107] The pectoral region is mainly photographed by imaging in the MLO direction in which the breast is obliquely sandwiched, and when the reference signal value is obtained by using the pectoral region, the pectoral region is similar to the mammary gland. (The mammary gland area varies from individual to individual, but regresses with age. The breast area in which the mammary gland has regressed has a large fat area with low radiation absorption.) it can.

The fat area is a high-density area (relatively dark on the image), and the regression of the mammary gland area has a large individual difference.
It may be difficult to determine the reference signal value depending on the personality.
For this reason, by determining the reference signal value in the fat area, it is possible to output a stable density.

In the mammogram, it is important that the skin line can be traced. By determining the reference signal value in the vicinity of the skin line, it is possible to output the density in the vicinity of the skin line at a desired level.

The reference signal value obtained here is used for image processing described later.

In the image processing, even if the level distribution of the image data output from the imaging panel 41 fluctuates due to the difference in the radiation dose, the image data DT is always processed so that a stable radiation image can be obtained. Normalization processing is performed. Also, even if the distribution of the image data level fluctuates,
To obtain a radiation image having a suitable density and contrast for diagnosis or the like, gradation processing is performed on normalized image data DTreg which is image data after normalization processing. Further, in the image processing, the frequency enhancement processing for controlling the sharpness of the normalized radiation image with respect to the normalized image data DTreg, and the entire radiation image having a wide dynamic range can be performed without lowering the contrast of the fine structure portion of the subject. A dynamic range compression process for keeping the density within the easy-to-see density range may be performed.

In the normalization process of the image data DT, when the reference signal values D1 and D2 are set, the reference signal values D1 and D2 are referred to as shown in FIG. Normalization processing for level-converting D2 to desired reference values S1 and S2 is performed. Here, the characteristic curve C
C indicates the level of a signal output according to the radiation dose of the radiation applied to the imaging panel 41. Further, the normalization processing look-up table is generated by an operation using an inverse function of the function indicating the characteristic curve CC of the imaging panel 41. It is needless to say that the normalization processing may be performed by arithmetic processing without using the normalization processing lookup table.

As a result of this normalization process, as shown in FIG. 10, the radiation dose R is lower than the doses R1 and R2 at which the radiation can obtain image data of the desired reference values S1 and S2.
Since the image data of the desired reference values S1 and S2 can be obtained even with a to Rb, the exposure of the subject can be reduced, and at the same time, the variation in the signal distribution due to the difference in the body type of the subject is corrected. be able to.

Next, gradation processing is performed using the normalized image data DTreg obtained by the normalization processing. In the gradation processing, for example, a gradation conversion curve as shown in FIG. 11 is used, and the reference value S of the normalized image data DTreg is used.
1 and S2 are parameter values of levels S1 ′ and S2 ′, and the normalized image data DTreg is output image data DTou.
is converted to t. The levels S1 'and S2' correspond to predetermined luminance or photographic density in the output image.

The gradation conversion curve is represented by the normalized image data DTr.
It is preferable that the function is continuous over the entire signal region of eg, and that its differential function is also continuous. Further, it is preferable that the sign of the differential coefficient is constant over the entire signal region.

Further, since the preferred shape of the gradation conversion curve and the levels S1 'and S2' differ depending on the photographing part, photographing position, photographing condition, photographing method, etc., the gradation transformation curve is created for each image. Also, for example, Japanese Patent Publication 5-2
As shown in No. 6138, a plurality of basic tone conversion curves are stored in advance, and one of the basic tone conversion curves is read out and rotated and translated to obtain a desired tone conversion curve. Can be easily obtained. The image processing unit 26 is provided with a gradation processing lookup table corresponding to a plurality of basic gradation curves, and an image obtained by referring to the gradation processing lookup table based on the normalized image data DTreg. By correcting the data according to the rotation and parallel movement of the basic gradation conversion curve, it is possible to obtain output image data DTout that has been subjected to gradation conversion. In the gradation conversion processing, not only two reference values S1 and S2 but also one reference value or three or more reference values may be used.

Here, the selection of the basic gradation curve and the rotation and translation of the basic gradation curve are performed based on the photographing part, photographing position, photographing conditions, photographing method and the like. When such information is input as management information using the input device 27, by using this management information, the basic gradation curve can be easily selected and the rotation direction of the basic gradation curve can be selected. And the amount of translation can be determined. Further, the levels of the reference values S1 and S2 may be changed based on an imaging part, an imaging position, imaging conditions, and an imaging method.

Further, the selection of the basic gradation curve and the rotation or translation of the basic gradation curve may be performed based on the information on the type of the image display device and the type of the external device for outputting the image. This is because the preferable gradation may differ depending on the image output method.

Next, the frequency emphasis processing and the dynamic range compression processing will be described. In the frequency emphasizing process, for example, in order to control the sharpness by the non-sharp mask process shown in Expression (1), the function F is calculated as follows.
3 and JP-B-62-62376.

Soua = Sorg + F (Sorg−Sus) (1) Here, Soua is the image data after the processing, Sorg is the image data before the frequency enhancement processing, and Sus is the average of the image data before the frequency enhancement processing. This is unsharp data obtained by a conversion process or the like.

In this frequency emphasizing process, for example, F (So
rg-Sus) is β × (Sorg-Sus),
When β (emphasis coefficient) is equal to the reference values T1 and T2 as shown in FIG.
Varies approximately linearly between

As shown by the solid line in FIG. 13, when values "A" and "B" are set to emphasize low luminance, β from the reference value T1 to the value "A" is maximized. Therefore, the value is minimized up to the value “B” −the reference value T2. The value "A"-
Up to the value “B”, β changes almost linearly. When emphasizing high luminance, as shown by the broken line, β from the reference value T1 to the value “A” is minimized, and the value “B” −the reference value T2
Up to the maximum. Further, from the value “A” to the value “B”, β
Is changed almost linearly. It should be noted that although not shown, β of the value “A” to the value “B” is maximized when the medium luminance is emphasized. Thus, in the frequency emphasis processing, the sharpness of an arbitrary luminance portion can be controlled by the function F.

Further, the method of the frequency emphasis processing is not limited to the above-described unsharp mask processing.
For example, a technique such as the multi-resolution method disclosed in Japanese Patent No. 645 may be used. In the frequency emphasizing process, the frequency band to be emphasized and the degree of emphasis are set based on the photographing site, photographing position, photographing condition, photographing method, and the like, similarly to the selection of the basic gradation curve in the gradation process. .

In the dynamic range compression processing, a function G is determined by the method disclosed in Japanese Patent Publication No. 266318 in order to perform control so that the density is within a legible range by the compression processing represented by the equation (2).

Stb = Sorg + G (Sus) (2) Note that Stb is image data after processing, Sorg is image data before dynamic range compression processing, and Sus is image data before dynamic range compression processing. This is the unsharp data obtained by the above method.

Here, as shown in FIG. 14 (a), G (Sus) is low, including the case where the non-sharp data Sus has a characteristic that G (Sus) increases when the level becomes lower than the level "La". It is assumed that the density of the density region is high, and FIG.
The image data Sorg shown in (b) is image data Stb in which the dynamic range on the low density side is compressed as shown in FIG. G (Sus) is shown in FIG.
As shown in (d), the unsharp data Sus has the level “L”.
In the case where G (Sus) is smaller than “b”, the density of the high-density area is determined to be high, and the image data Sorg shown in FIG.
As shown in FIG. 4E, the dynamic range on the high density side is compressed. In the dynamic range compression process as well, the correction frequency band and the degree of correction are set based on the imaging site, imaging body position, imaging conditions, imaging method, and the like.

Here, reference values T1 and T2 and values "A" and "B" or levels "La" and "Lb" which are processing conditions in the frequency emphasizing process and the dynamic range compressing process.
Is determined by a method similar to the method of determining the representative values D1 and D2.

As described above, according to the above-described embodiment, even when the ratio between the high-density region and the low-density region greatly changes, stable gradation processing is performed without depending on the distribution of image data. Therefore, a radiation image having a density and contrast suitable for diagnosis or the like can always be obtained stably.

The image data used to determine the processing conditions may be, for example, image data that has been subjected to thinning processing. In this case, since the data amount is reduced, the processing speed can be improved and the memory capacity and the like can be reduced. In this case, the effective pixel size of the reduced image obtained by thinning is 0.4 mm to 10.0 mm, preferably 1.
It is preferable to use image data thinned to be 0 mm to 6.0 mm.

[0130]

As described above, claims 1 to 1 are as described above.
According to the invention described in Item 6, by analyzing the inside of the subject area and dividing the subject area into smaller areas, it is possible to output each area with an appropriate density.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a configuration of an image processing apparatus.

FIG. 2 is a diagram illustrating a configuration of an imaging panel.

FIG. 3 is a diagram showing a configuration when a radiation image reader using a stimulable phosphor is used.

FIG. 4 is a diagram showing a configuration of a controller.

FIG. 5 is a diagram illustrating a configuration of an image processing apparatus.

FIG. 6 is a diagram showing a local region of a radiographic image generated based on radiation that has passed through the breast.

FIG. 7 is a diagram showing a radiation image generated based on radiation that has passed through the breast.

FIG. 8 is an image diagram of a typical mammogram photographing direction.

FIG. 9 is a diagram showing level conversion.

FIG. 10 is a diagram showing a normalization process.

FIG. 11 is a diagram showing gradation conversion characteristics.

FIG. 12 is a diagram illustrating a relationship between an enhancement coefficient and image data.

FIG. 13 is a diagram illustrating a relationship between an enhancement coefficient and image data.

FIG. 14 is a diagram illustrating a dynamic range compression process.

[Explanation of symbols]

 26 Image processing unit 100 Breast region recognition means 101 Breast region analysis means 101a Local region threshold value determination means 102 Region of interest determination means 103 Reference signal value determination means 103a Reference signal value automatic selection means 103b Manual reference signal value determination area selection means 104 Image Processing means

Claims (18)

[Claims] 1. An image processing method for processing a radiation image generated based on radiation that has passed through a breast, comprising the steps of: recognizing a breast region and analyzing the recognized breast region; An image processing method characterized by classifying into characteristic regions, determining a reference signal value based on the classification result, and performing image processing based on the reference signal value. 2. The image processing method according to claim 1, wherein the analysis in the breast region determines a threshold value in a local region. 3. The image processing method according to claim 1, wherein the reference signal value is determined based on pixels around a mammary gland region. 4. The image processing method according to claim 1, wherein the reference signal value is determined based on a pectoral muscle region. 5. The image processing method according to claim 1, wherein the reference signal value is determined based on a fat area. 6. The image processing method according to claim 1, wherein the reference signal value is determined based on a vicinity of a breast margin. 7. The method according to claim 1, wherein an area in which one or more reference signal values are determined among the reference signal values is automatically determined according to a feature of an image. Item 2. The image processing method according to item 1. 8. The image processing method according to claim 3, wherein one or more reference signal value determination areas among the reference signal values can be manually determined. 9. The method according to claim 3, wherein an area for determining the reference signal value is determined in advance, and the reference signal value is determined from the determined area for determining the reference signal value. The image processing method described in the above. 10. An image processing apparatus for processing a radiation image generated based on radiation passing through a breast, comprising: a breast region recognizing unit for recognizing a breast region; and a breast region recognized by the breast region recognizing unit. A breast region classification unit that classifies the inside of the breast region into a characteristic region by analyzing, and a reference signal value determination unit that determines a reference signal value based on a result of classification by the breast region classification unit, An image processing apparatus for performing image processing based on the reference signal value. 11. The image processing apparatus according to claim 10, wherein said intra-breast region classification means includes a local area threshold value determination means for determining a threshold value in a local area. 12. The image processing apparatus according to claim 10, wherein the reference signal value is determined based on pixels around a mammary gland region. 13. An apparatus according to claim 10, wherein said reference signal value is determined based on a pectoral muscle region. 14. The image processing apparatus according to claim 10, wherein said reference signal value is determined based on a fat area. 15. The image processing apparatus according to claim 10, wherein the reference signal value is determined based on a vicinity of a breast margin. 16. An apparatus according to claim 1, further comprising a reference signal value automatic selecting means for automatically determining one or more reference signal value determining regions in accordance with image characteristics among said reference signal values. The image processing apparatus according to claim 10. 17. The apparatus according to claim 12, further comprising a manual reference signal value determining area selecting means for manually determining one or more reference signal value determining areas among the reference signal values. The image processing apparatus according to claim 1. 18. A reference signal value determining area storing means for storing an area for determining the reference signal value in advance, wherein the reference signal value is determined by the reference signal value determining area storing means. The image processing apparatus according to claim 12.