JPH1032756A - Method and device for processing gradation of image information - Google Patents

Abstract

PROBLEM TO BE SOLVED: To execute gradation processing with optimum density without causing black solid or white void at a local area in a ROI set on a radiograph with simple operation in a gradation processor for image information. SOLUTION: In a whole image P' displayed on a display means 20, a desired local area P1' is set by a local area setting means 30, and a characteristics value calculating means 51 of a gradation change means 50 calculates a simple average value Q of local image information S1 corresponding to the local area P1'. When the emphasis degree of gradation desirable for an operator is inputted from an input means 40, corresponding to the inputted emphasis degree, the inclination of the gradation converting curve (gradation processing condition) of a gradation processor 10 is enlarged so that a point GC on the curve corresponding to the simple average value Q of the local image information S1 becomes a fixed point.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a gradation processing method and apparatus for image information, and more particularly, to gradation processing for an entire radiation image of a subject and further performing gradation processing for a local area in the radiation image. The present invention relates to a gradation processing method and an apparatus for performing the following.

[0002]

2. Description of the Related Art There are various methods for reading a radiographic image recorded on an X-ray film or a stimulable phosphor sheet to obtain image information, subjecting the image information to appropriate gradation processing, and then reproducing it as a visible image. (For example, JP-B-61-5193, JP-A-55-12429, JP-A-56-11395,
No. 55-163472, No. 56-104645, No. 55-116340).

When performing such gradation processing, for example, in the field of medical images, it is necessary to select processing suitable for the purpose of diagnosis. However, the selection of such processing is familiar with image processing technology. It is actually difficult to select the most suitable image processing conditions for the purpose of diagnosis, even if all of them are left to a doctor or the like who does not have it.

Therefore, the applicant of the present application can input optimum information on a radiographic image based on this input information only by inputting information on a subject imaging method and an imaging part without any special knowledge on image processing. By automatically obtaining the conditions and performing gradation processing on the radiographic image based on the gradation processing conditions, a reproduced image subjected to gradation processing suitable for a diagnosis purpose can be obtained with a simple operation. A radiation image reproducing apparatus has been proposed (JP-A-59-28144).

The above gradation processing conditions include a gradient G of the gradation curve which specifies a gradation conversion curve representing the gradation processing shown in FIG.
A, the type GT of the gradation curve, the fixed point (reference point) GC of the gradation curve, and the shift amount GS of the gradation curve.

Further, the applicant of the present application has proposed a desired local area (for example, soft tissue, air contained in tissue,
Japanese Patent Application Laid-Open No. 7-92587 discloses a technique for setting a region of interest (hereinafter referred to as an ROI) and further performing a gradation process in order to make it easier to observe at least a local region in the ROI. I have.

In this technique, an operator such as a doctor sets a desired local area on a whole image displayed on a display means by using a keyboard or the like, and further sets a type of a part (soft tissue, air contained in tissue, air contained in tissue, or the like). Bone, calcareous, etc.), the optimum image processing condition G based on the input type
C and GS are automatically determined. Specifically, when the type of the selected and input part is a soft tissue, first, the density DP of the set local region in the ROI is determined, and the density DS (preset) is excellent in suitability for observation and interpretation of the soft tissue. Difference (DP
The original GS is changed to GS ′ (= GS + DP−DS) and the original GC is changed to GC ′ (= DS) so as to shift the gradation conversion curve by −DS).

By performing gradation processing on the local area in the ROI based on the newly set gradation processing conditions, the local area, which is the region of interest, is displayed on the display means at a gradation suitable for observation and interpretation. can do.

[0009]

The technique disclosed in Japanese Patent Application Laid-Open No. 7-92587 is a process of merely shifting the density of a local region in an ROI with respect to an original image (entire image). When the region is a tumor shadow or the like whose density change is gradual, it is not enough to interpret details such as the height of the tumor shadow.

Therefore, by changing the gradient GA of the gradation conversion curve, the degree of gradation emphasis can be increased, and an image more suitable for image interpretation can be obtained. For example, by changing the slope GA from 1.0 to 1.2, the degree of emphasis of the gradation of the image in the local region is increased.

However, if the gradient GA is simply changed from 1.0 to 1.2, the overall density of the local area becomes high or low, and the image of the desired essential part may be crushed black or white. There is.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and provides a gradation processing of image information capable of performing an optimum gradation processing by a simple operation on a local region in an ROI set on a radiographic image. It is an object to provide a method and an apparatus.

[0013]

According to the present invention, there is provided a method and apparatus for gradation processing of image information, wherein a gradient of a gradation conversion curve is maintained while keeping characteristic values such as an average density value of a local region set on a radiographic image constant. It is characterized in that GA is changed.

In other words, the image information gradation processing method and apparatus of the present invention perform gradation processing based on a predetermined gradation conversion curve and display the image information carrying a radiation image of a subject on predetermined display means. Setting a desired local region on the displayed radiographic image, and performing a gradation process on at least the set local region so as to have a further desired gradation, Determine a predetermined characteristic value related to the density or luminance of the local area,
The gradient of the gradation conversion curve is changed while maintaining the predetermined characteristic value constant, and gradation processing is performed on the at least local region based on the gradation conversion curve with the changed slope. It is assumed that.

Here, the predetermined characteristic value may be, for example, an average value of density or luminance in a local area, a median value of density or luminance in a local area, or a predetermined value in a cumulative histogram of density or luminance in a local area. , Or the average value of the density or luminance of the remaining portion of the local region after removing the direct radiation portion directly irradiated with radiation, or the like. The same applies to the following inventions.

The desired local area may be automatically detected and set.

For example, when the desired local region is a tumor shadow or a microcalcification shadow which is a characteristic form of breast cancer in a breast radiographic image (mammography), these shadows are automatically processed by iris filter processing or morphological processing. Can be detected.

Here, the iris filter processing is an abnormal shadow detection processing proposed by the present applicant in Japanese Patent Application No. Hei 7-316679, which automatically detects a tumor shadow. Calculate the gradient vector of the density value of
The evaluation value corresponding to the degree of concentration of the gradient vector is subjected to threshold processing, and the tumor shadow can be accurately separated and detected from blood vessels, mammary glands, and the like.

On the other hand, the morphology processing is a microcalcification detection processing for automatically detecting microcalcification shadows proposed in Japanese Patent Application No. Hei 7-316679, and specifically, is applied to a radiographic image. On the other hand, the minimum value processing and the maximum value processing are performed using structural elements of a predetermined size and shape, and the difference between the processed image and the original image is obtained, thereby being affected by complicated background information. Calcified shadows can be extracted without distortion.

Further, the gradation processing apparatus for image information according to the present invention comprises:
Image information carrying a radiation image of a subject, gradation processing means for performing gradation processing based on a predetermined gradation conversion curve, display means for displaying the gradation-processed image information as a visible image, Local area setting means for setting a desired local area on the displayed radiographic image; input means for inputting a desired gradation for the local area; and A gradation changing means for changing a gradient of the gradation conversion curve, wherein the gradation processing means further performs a further gradation process on at least the local region based on the gradation conversion curve having the changed inclination. In the gradation processing apparatus for image information described above, the gradation changing means includes characteristic value calculating means for calculating a predetermined characteristic value relating to the density or luminance of the local region, and the determined predetermined characteristic It is characterized in that it is the one that changes the inclination of the gradation conversion curve while maintaining constant.

[0021]

According to the method and the apparatus for gradation processing of image information of the present invention, the gradient GA of the gradation conversion curve can be obtained while keeping a predetermined characteristic value such as an average density value of a local area set on a radiographic image constant. The change does not increase or decrease the overall density of the local region, and accordingly appropriately changes the degree of gradation enhancement while appropriately maintaining the desired image density of the essential part. Can be.

[0022]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, specific embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of a gradation processing apparatus for image information according to the present invention. The illustrated gradation processing apparatus 100 includes image information S that carries a radiation image P of a subject.
To a gradation processing means 10 for performing gradation processing based on a predetermined gradation conversion curve, and a CRT display for displaying image information S 'subjected to gradation processing by the gradation processing means 10 as a visible image P'. Etc., and the displayed radiation image P ′
A local area setting means 30 for setting a desired local area P1 'above, an input means 40 for inputting a desired gray scale for the set local area P1', and a gray scale input to the input means 40 And a gradation changing means 50 for changing the gradient of the gradation conversion curve in the gradation processing means 10.

The gradation changing means 50 further includes a local area P1 '.
Value calculating means for obtaining a predetermined characteristic value Q related to the density of
51, and the gradient GA of the gradation conversion curve of the gradation processing means 10 is changed while maintaining the obtained predetermined characteristic value Q constant. The image information S is again subjected to gradation processing according to the gradation conversion curve, and the display means 20 is caused to display the visible image P ″.

As shown in FIG. 2, the gradation processing conditions for defining the gradation conversion curve of the gradation processing means 10 are GA indicating the gradient of the gradation curve, GT indicating the type of the gradation curve, and gradation. It comprises a fixed point GC of the tone curve and a shift amount GS of the tone curve. Here, the contrast decreases when the gradient GA of the gradation curve is reduced, and the contrast improves when the gradient GA is increased.
The tone curve type GT is a preset curve pattern. For example, the A type represents a straight tone. The fixed point GC of the gradation curve represents a fixed point when changing the gradient GA of the gradation curve, and the shift amount GS represents a movement amount of moving the gradation curve in parallel.

The local area setting means 30 is a means for setting an ROI at a desired portion on the radiation image P 'displayed on the display means 20, and includes a mouse 31 and a keyboard (not shown). Have.

The input means 40 is a means for the operator to input a value or the like corresponding to the gradient GA of the gradation conversion curve. If the local area setting means 30 has a keyboard, this input means The unit 40 may also serve as the unit 40.

The characteristic value calculating means 51 of the gradation changing means 50 determines whether the local area P in the ROI set by the local area setting means 30 is correct.
1 ′ is extracted from the image information (hereinafter, referred to as entire image information) S,
The average value S1mean (= ΣSij / N; Sij of the signal values (density values) of the extracted local image information S1; Sij represents the signal value of each pixel included in the ROI, and N represents the number of pixels included in the ROI. ) Is calculated as the characteristic value Q. Then, the tone changing means 50 rotates the point (Q, GC) on the original tone curve so that the value of the point GC on the tone curve corresponding to the calculated characteristic value Q does not change. As a center,
The gradient of the gradation curve of the gradation processing unit 10 is changed so as to conform to the gradient GA of the gradation conversion curve input from the input unit 40.

FIG. 3A shows a radiation image P used in the present embodiment, and is specifically a mammography in which a breast is a subject.

Hereinafter, the operation of the gradation processing apparatus according to the present embodiment will be described.

First, a radiation image P as shown in FIG.
The whole image information S read by the predetermined reading means from the stimulable phosphor sheet on which is stored is input to the gradation processing means 10 of the gradation processing apparatus 100.

The whole image information S input to the gradation processing means 10 is converted into processed whole image information S 'according to a gradation conversion curve set in advance according to the type of the input image and the like, and is displayed. Output to 20. In the embodiment shown in FIG. 1, only the gradation processing is performed on the input image information S, but this does not prevent the use of various kinds of signal processing such as frequency processing and enlargement / reduction. .

The display means 20 displays the inputted processed whole image signal S 'as a visible image (whole image) P' on its display surface as shown in FIG. An image interpreter such as a doctor observes the entire image P '.

Here, when a doctor or the like who observes the whole image P 'finds a part recognized as an abnormal shadow P2' in the image of the subject, the doctor or the like uses the mouse 31 (or a keyboard (not shown)) of the local area setting means 30. Then, the ROI is set by designating the local area P1 'including the abnormal shadow P2', and a desired enhancement degree for the local area P1 'is input from the input means 40. That is, since the doctor or the like wants to display the abnormal shadow P2 'in a high-gradation image more suitable for diagnosis for the local area P1', the display means 20
Of the local area P1 'displayed at a higher level than the gray level of the local area P1'. Specifically, for example, the degree of emphasis is set based on the gradation of the displayed local area P1 '(= 1.0).
Enter 1.2 to improve by 20%. Note that the degree of gradation enhancement determines the slope GA of the gradation conversion curve.

The gradation changing means 50 sets the local area P
The local image information S1 corresponding to 1 'is extracted from the entire image information S, and the average value S1mean of the density values of the extracted local image information S1 is calculated as the characteristic value Q. Then, the gradation changing means 50 keeps the value of the point GC on the gradation curve corresponding to the calculated characteristic value Q unchanged, that is, moves the point (Q, GC) on the original gradation curve. As a point, the gradient GA of the gradation conversion curve input from the input means 40
To the slope of the gradation curve of the gradation processing means 10 so that
Multiply by 1.2.

By the above operation, as shown by the broken line in FIG. 1, the gradation conversion curve of the gradation processing means 10 increases the degree of gradation emphasis around the average density value Q of the local area P1 '.
The image information S is subjected to gradation processing in accordance with the gradation conversion curve in which the degree of gradation enhancement is increased, and the display means 20 is displayed.
As shown in FIG. 3 (C), an entire image P ″ including a local area P1 ″ having a gradation desired by a doctor or the like is displayed.

Here, at least a local area P1 "of the displayed whole image P" has an average density Q
Is equal to the average density value of the local area P1 'in the original whole image P', so that the displayed image does not become black or fly white. Therefore, the display means 20 displays an image having a density suitable for observation and a desired gradation.

As described above, according to the gradation processing apparatus of the present embodiment, when the gradation of a local region in the whole image is set to a desired gradation, it is determined that the density of at least the local region is inappropriate. In order to automatically set the rotation center of the gradation conversion curve so that it does not occur, it is necessary for a doctor who does not have specialized knowledge about image processing to set the rotation center GC through trial and error many times. There is no.

In the image information gradation processing apparatus of the present embodiment, the characteristic value calculating means 51 determines the average value S1mean of the local image information S1 in the ROI set by the local area setting means 30.
Has been described as the characteristic value Q, but the characteristic value Q is not limited to such a simple average value. For example, the median value S1med of the local image information S1
Also, a percentage value of a predetermined cumulative frequency in the cumulative histogram of the local image information S1 (equivalent to a median value if the value is 50%) S1per, or a direct radiation portion of the local image information S1 to which radiation is directly applied is removed. An average value of the image information of the remaining portion (subject portion) can be applied. Further, the maximum value Smax of the image information of the remaining portion of the whole image information S from which the radiation portion has been directly removed is obtained, and the average value of the image information of the remaining portion of the local image information S1 from which the portion equal to or more than the maximum value Smax has been removed. Can also be applied.

In this embodiment, the whole image is processed and displayed according to the new gradation processing condition after the gradation processing condition is changed. May be processed and displayed under the original gradation processing condition, and only the local region may be processed and displayed under the new gradation processing condition. in this case,
The local region processed according to the new gradation processing condition may be displayed so as to be superimposed on the original whole image or separately displayed on another display means different from the display means on which the whole image is displayed. You may make it.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a gradation processing apparatus for image information according to the present invention.

FIG. 2 is a graph showing a gradation conversion curve set by gradation processing conditions.

3A is a diagram illustrating a radiation image processed by the gradation processing apparatus illustrated in FIG. 1, FIG. 3B is a diagram illustrating an entire image processed under preset gradation processing conditions, and FIG. Diagram showing the whole image processed by changing the key gradient

[Explanation of symbols]

 10 gradation processing means 20 display means 30 local area setting means 31 mouse 40 input means 50 gradation changing means 51 characteristic value calculation means 100 gradation processing device S image information GA, GT, GC, GS gradation processing conditions

Claims (10)

[Claims] An image information carrying a radiation image of a subject is displayed on a predetermined display means after being subjected to gradation processing based on a predetermined gradation conversion curve, and a desired image is displayed on the displayed radiation image. A gradation processing method for image information in which a local region to be set is set and gradation processing is performed so that at least the set local region has a further desired gradation, wherein a predetermined characteristic value relating to density or luminance of the local region is provided. The gradient of the gradation conversion curve is changed in accordance with the desired gradation while maintaining the predetermined characteristic value constant. Based on the gradation conversion curve with the changed slope, And performing a gradation process on at least the local region. 2. The gradation processing method for image information according to claim 1, wherein said predetermined characteristic value is an average value of density or luminance in said local area. 3. The gradation processing method for image information according to claim 1, wherein the predetermined characteristic value is a median value of density or luminance in the local area. 4. The gradation processing of image information according to claim 1, wherein said predetermined characteristic value is a percentage value of a predetermined cumulative frequency in a cumulative histogram of density or luminance in said local area. Method. 5. The apparatus according to claim 1, wherein the predetermined characteristic value is an average value of density or luminance of a remaining portion of the local region after removing a direct radiation portion directly irradiated with radiation. Image information gradation processing method. 6. A gradation processing means for performing gradation processing on image information carrying a radiation image of a subject based on a predetermined gradation conversion curve, and displaying the gradation-processed image information as a visible image. Display means, and a local area setting means for setting a desired local area on the displayed radiographic image,
Input means for inputting a desired gradation for the local region;
A gradation changing means for changing a slope of the gradation conversion curve based on the gradation inputted to the input means, wherein the gradation processing means performs the processing based on the gradation conversion curve having the changed slope. An image information gradation processing apparatus configured to further perform gradation processing on at least the local region, wherein the gradation change unit includes a characteristic value calculation unit that obtains a predetermined characteristic value related to density or luminance of the local region. A gradation processing apparatus for image information, wherein the inclination of the gradation conversion curve is changed while maintaining the obtained predetermined characteristic value constant. 7. The image information gradation processing apparatus according to claim 6, wherein the predetermined characteristic value is an average value of density or luminance in the local area. 8. The gradation processing apparatus for image information according to claim 6, wherein said predetermined characteristic value is a median value of density or luminance in said local area. 9. The gradation processing of image information according to claim 6, wherein the predetermined characteristic value is a percentage value of a predetermined cumulative frequency in a cumulative histogram of density or luminance in the local region. apparatus. 10. The characteristic value according to claim 6, wherein the predetermined characteristic value is an average value of the density or luminance of a remaining portion of the local region after removing a direct radiation portion directly irradiated with radiation. Image information gradation processing device.