JPH08131428A - Digital x-ray radiographing device - Google Patents

Abstract

PURPOSE: To provide a digital X-ray radiographing device capable of estimating at least one of a radiographing part, a radiographing direction and a radiographing method from radioscopic image data, deciding and radiographing a radiographing condition based on an estimated result, or deciding the arithmetic processing content of the image data based on the estimated result and displaying processed image data. CONSTITUTION: Processing step 21 to estimate at least one of the radiographing part, the radiographing direction and the radiographing method based on the radioscopic image data in the digital radiographing device, and processing step 22 to decide radiographing condition control based on the estimated result and the radioscopic image data, or processing step (not shown in figure) to decide the arithmetic processing content of the image data are provided.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention estimates at least one of a photographing region, a photographing direction, and a photographing method from fluoroscopic image data, determines photographing conditions based on the estimation result, and photographs the estimated condition. The present invention relates to a digital X-ray imaging apparatus capable of determining the content of arithmetic processing of image data based on the above.

[0002]

2. Description of the Related Art In a conventional inspection using a digital X-ray imaging apparatus, an operator observes a fluoroscopic image by inserting a collimator and a density compensation filter (hereinafter abbreviated as a compensation filter) in a necessary area for observation. I adjusted the brightness so that it would be suitable for shooting. Further, when observing with a display device, gradation conversion may be performed manually for observing.

[0003]

In the above-mentioned conventional digital X-ray imaging apparatus, the operator must perform various operations such as determination of the imaging position or insertion of a catheter at the time of imaging, and the collimator / compensation filter, X-ray There is a problem in that operations such as determination of conditions impose a heavy burden on the operator and greatly affect the examination efficiency. Further, in the case of displaying an image and making a diagnosis, an operation of manually performing gradation conversion to make the diagnosis easy is also a heavy burden on the operator and adversely affects the diagnosis efficiency.

An object of the present invention is to estimate the imaging condition by estimating at least one of the imaging region, the imaging direction, and the imaging method, and automatically insert the collimator / compensation filter and the X-ray condition according to the imaging condition. An object of the present invention is to provide a digital X-ray imaging apparatus that can reduce the burden on the operator during imaging and improve examination efficiency by making settings.

Another object of the present invention is to estimate at least one of a photographing region, a photographing direction and a photographing method,
An object of the present invention is to provide a digital X-ray imaging apparatus capable of determining the arithmetic processing content of image data based on the estimation result and displaying the image data processed based on the processing content.

[0006]

In order to achieve the above object, the present invention estimates at least one of an imaging region, an imaging direction, and an imaging method based on fluoroscopic image data of a digital X-ray imaging apparatus. Processing step (21 in FIG. 2)
And a processing step (22 in FIG. 2) for determining imaging condition control (see FIGS. 6 and 7) based on the estimation result and the fluoroscopic image data. Further, it is characterized in that a processing step for deciding the arithmetic processing content of the image data is provided based on the estimation result, and the image data automatically processed according to the decided content is displayed.

[0007]

According to the present invention, based on the fluoroscopic image data,
Processing steps are provided for estimating at least one of the imaging region, the imaging direction, and the imaging method. In this step, for example, input data is created and input based on the fluoroscopic image data to an estimation neural network created by learning images of the imaged region / direction / method to be estimated in advance, and the imaged region / direction / method is estimated. . In addition to the estimation using only the neural network, the estimation can be performed in combination with the information such as the position and angle of the bed and the arm that can be fetched from the device.

Further, according to the present invention, by providing the processing step for determining the control of the photographing condition based on the above-mentioned estimation result, for example, the photographing is carried out by the device including the filter and the collimator which are composed of four vertically and horizontally. When the part is estimated to be the front of the head, the filter uses the upper and left three sheets and does not rotate, the collimator covers the upper and left and right three sheets except the head, and the lower one covers the neck, and the right front of the heart. Oblique position (RAO: Ri
If it is estimated to be ght Anterior Oblique), the filter is used by rotating one on the top or the right, the collimator covers four parts other than the heart with the top, bottom, left, and right, etc. Common basic control for each imaging site, direction, method Can be set in advance. The X-ray control can also use a preset imaging region, direction, voltage, current, exposure time for each method, and the like. Further, the photographing condition control for each image can be determined based on the estimation result, the basic control, and the fluoroscopic image. For example, when the estimation result is the front of the head, it is determined by the above basic control that the three upper, left, and right filters are used and the filter is not rotated.
The compensation filter insertion area can be detected from the perspective image using a predetermined threshold value, and the position covering the compensation filter insertion area can be calculated without rotating the three filters. If the estimation result is RAO of the heart, it is decided that the filter is used by rotating the upper or right one, so the position of the filter covering the compensation filter insertion area occurring in the upper right is calculated, and the upper or right filter is calculated. It can be decided to use a filter having a smaller amount of movement / rotation. By estimating the imaged region, direction, and method in this way, the basic control of the device characteristic of the imaged region, direction, and method is limited, and the imaging condition control for each captured image is determined based on the basic control. By using this method, unnecessary calculation does not occur, and shooting control can be performed at high speed.

Further, according to the present invention, based on the position of the compensation filter and the X-ray attenuation characteristic of the compensation filter measured in advance for the first perspective image data in which the compensation filter is already inserted up to an arbitrary position. Since the conversion process for creating the second perspective image data in the state in which the compensation filter is not inserted is provided, even if the compensation filter is inserted in an inappropriate position, the compensation filter is not returned to the initial position. The insertion area of the compensation filter can be detected.

Further, according to the present invention, based on at least the result of estimating one of the imaging region, direction and method,
It is possible to determine the content of arithmetic processing for image data. The arithmetic processing referred to here is general image processing such as filter processing / gradation conversion processing, smoothing processing / edge emphasis processing / enlargement / reduction processing, etc. A suitable method can be selected and processed / displayed. For example, in the filter process, setting of a filter suitable for the imaged region, in gradation conversion, setting of a conversion curve suitable for the imaged region and the image capturing method, and in smoothing process, setting of a smoothing size suitable for the imaged region.

[0011]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing an example of a digital X-ray imaging apparatus according to the present invention. This device is used for the subject 1
An X-ray tube 2 disposed opposite to each other with an image in between, an image intensifier (hereinafter abbreviated as II) 3, which is an X-ray detector, between the subject 1 and the X-ray tube 2. Placed X
Line compensation filter 4, collimator 5, television camera 6, analog-digital converter (hereinafter abbreviated as A / D converter) 7, arithmetic unit 8, image memory 9, digital-analog converter (hereinafter D / A). (Abbreviated as converter)
10, a monitor 11 which is a display device, an X-ray controller 12,
A high voltage generator 13, an X-ray compensation filter controller 14,
Collimator control device 15, console 16, central processing unit 1
It is composed of 7. The outline of the operation of the above-described device of the present invention will be described.

A predetermined high voltage (X
The tube voltage), the tube current, the pulse width, and the sequence of imaging are defined, the high voltage generator 13 generates a voltage / current according to the sequence, and the X-ray tube 2 generates an X-ray. The exposed X-rays pass through the subject 1 and the I.V. I. At 3, it is converted into a visible light image. The visible light image is formed on the television camera 6, and the output signal from the television camera 6 is digitally converted as image data by the A / D converter 7. The image data is calculated by the calculation device 8 as necessary and stored in the image memory 9. The stored image data is converted into an analog signal by the D / A converter 10 and displayed on the monitor 11.

FIG. 2 is a flow chart showing the photographing procedure of the present invention. In the following, at least one of the imaging region, the imaging direction, and the imaging method is estimated based on the fluoroscopic image data, the imaging control condition of the X-ray compensation filter 4 is set, and the X-ray compensation filter controller 14 controls the X-ray compensation. An example of automatically controlling the filter 4 and capturing an image will be described.

[Step 21]: Imaging site / imaging direction /
Imaging method estimation Prepare a large number of image data of the imaging site, imaging direction, and imaging method to be estimated in advance, calculate the data necessary for learning the neural network from each image data, and learn the imaging site, imaging direction, and imaging method Estimation neural network 3
1 (hereinafter abbreviated as estimated NN) is created. Estimated N. N. An example of 31 is shown in FIGS. FIG.
The example in (a) is based on the assumption of cardiovascular imaging. The structure of the neural network is a three-layer hierarchical type, and learning is performed by the back propagation method. Learning is performed by creating learning data to be input to the input layer from the teacher image data. The learning data is created, for example, by dividing the image into a plurality of blocks, and normalizing the average luminance value / luminance standard deviation value of each block, and the average luminance value deviation / average luminance standard deviation value between blocks. In addition, an image may have a different luminance distribution due to differences in shooting conditions or the like and may not be suitable for learning. Therefore, when creating learning data, fitting conversion may be performed so as to obtain a predetermined luminance distribution. Alternatively, the image may be edge-emphasized to use the feature amount representing the edge shape or position as learning data, or the image data may be binarized to use the area, shape, or variation of the brightness value representing the feature as learning data. The estimated N.V. learned using the learning data created in this way. N. Input data 32 created from perspective image data is input to 31 and an estimation result 33 is output. The estimated N.V. of FIG. N. Numeral 31 designates all the imaging parts, the imaging directions, which are targeted by one neural network.
This is a configuration for estimating the shooting method pattern, but FIG.
Is a configuration in which multiple neural nets are used to perform stepwise estimation. For example, the estimated N. N. In (1), four heads, heart, abdomen, and limbs are estimated.
N. In (2), the front or side is estimated. In this way, if the group for each stage is appropriate, the method of estimating in stages with a plurality of neural nets is a method of estimating all the imaging parts, imaging directions, and imaging method patterns of interest with one neural network. Compared with this, the scale of each neural network and the entire network becomes smaller, so that learning efficiency can be improved.

[Step 22]: Shooting control setting [Step 22a]: Basic control setting FIG. 4 shows an example in which the compensation filter 4 is composed of four blades. Each of the blades A41 to D44 moves back and forth and rotates left and right based on the center of the image. Of course, the number of blades of the compensation filter 4 is not limited to four, and the shape of each blade is not limited to the rectangular shape shown in FIG. 4 and may be any shape. Step according to the compensating filter 4 configuration and collimator 5 configuration for each device
Based on the estimation result 33 of p21, the shooting control is determined in the order of the detailed control for determining the basic shooting control for each image. For example, as shown in FIG. 5, the knowledge base 5 of the basic control may be used in accordance with the compensating filter 4 configuration and the collimator 5 configuration for each device.
1 is created in advance, and that information is used for detailed shooting control settings 2
3 or directly to the control device. For example, when it is estimated that the head is the front of the head, the compensation filter 4 uses the blade A41, the blade B42, and the blade C43, and the blade A41 has an angle of 0 °.
The blade B42 is fixed at 90 ° and the blade C43 is fixed at 180 °, and moves only forward and backward. For X-ray conditions, refer to basic control information of voltage 70 to 75 kV, current 400 to 600 mA, and exposure time 60 to 80 ms. When the heart RAO is estimated, the compensating filter 4 uses the blade A41 or the blade B42, and the angles are 0 ° to 60 ° for the blade A41 and 3 for the blade B42.
Rotate / move back and forth in the range of 0 ° to 90 °, X-ray condition is
For example, the basic control information of voltage 70 to 85 kV, current 400 to 600 mA, and exposure time 5 ms is referred to.

[Step 22b]: Detailed control setting The detailed control is determined with reference to the basic control information determined in step 22a. Here, the method of determining the insertion position of the compensation filter will be mainly described. This example is presumed to be cardiac RAO,
The basic control of the compensation filter is blade A41 or blade B42.
Is used.

As shown in FIG. 6, a compensation filter insertion area (hereinafter abbreviated as a filter area) 62 using a predetermined threshold brightness value S inside a circular image area 67 of a perspective image.
To calculate. Here, the image area 67 does not necessarily have to be circular. Next, the compensation filter insertion position is determined from the filter area 62, and the vane A41 or the vane B42 is moved to the position 64 after insertion of the compensation filter. The compensation filter insertion position, the current position of the blade A41, and the current position of the blade B42 are respectively line segments 63, 65, 66 as shown in the figure.
Is used to represent the distance from the image center 61 by the length and the angle by the direction. For example, the current position 6 of the blade A41
5 is a distance of 200 and an angle of 0 °. As a method of determining which of the blade A41 and the blade B42 is used, a blade having a small amount of movement from the current position of each blade to the compensation filter insertion position 63 is selected. However, if the determined blade exceeds the rotation range of the basic control blade, the other blade shall be used. This control is not necessary when the movable range of each blade is not limited, but it is necessary because the movable range of the blade normally exists.

The method for determining the compensation filter insertion position is, for example, as shown in FIG. 7A, the image center 61 of the filter area 62.
The approximate straight line 72 of the edge point sequence 71 on the side is calculated, the position 73 of the straight line that maintains the inclination of the approximate straight line 72 and covers the entire filter region 62 is calculated, and the distance between the straight line 73 and the image center 61 is calculated.
The angle is the compensation filter insertion position 74. Edge point sequence 7
Calculation of the approximate straight line 72 from 1

[0020]

[Equation 1]

Is used. Another determination method is shown in FIG.
As shown in, the inertial main axis 75 is calculated using the coordinate points in the filter area 62, and the straight line position 76 that keeps the inclination of the main axis 75 and covers the entire filter area 62 is calculated.
The distance and angle with respect to the image center 61 of are the compensation filter insertion positions 77. Calculation of the principal axis of inertia 75 from the coordinate points of the filter area 62

[0022]

[Equation 2]

Is used. As shown in FIG. 6, when each blade does not cover the image region 67 when the perspective image data is captured (hereinafter referred to as the initial state), the captured image data is used as it is to determine the filter region 62. You can However, at the time of actual photographing, the wings often fall arbitrarily on the image area 67 of the fluoroscopic image data. Since the brightness of the winged portion is changed, an appropriate filter area 6 is obtained using this perspective image data.
You cannot decide 2. In order to determine the filter area 62, it is necessary to take in the fluoroscopic image data in the initial state, so it is conceivable to return each blade to the outside of the image area 67 each time. However, if there is no significant difference between the actual filter insertion position and the current filter position, the movement of the filter becomes troublesome twice and the photographing time becomes long. As a method of solving this, the position of each blade and the X-ray attenuation characteristic or the luminance attenuation rate at each position are measured in advance, an inverse conversion table thereof is created, and this inverse conversion table and the position of each blade (from the image center 61 The distance and angle of 1) are used to convert the perspective image data to create the initial state perspective image data. By using the created perspective image data in the initial state, the filter region 62 and the compensation filter insertion position 63 can be determined. Also, the collimator can determine the insertion position under the constraint of the basic control information as in the compensation filter. As the X-ray condition, the average brightness value of the entire fluoroscopic image data or a predetermined area is calculated, and the voltage / current is determined under the constraint of the basic control condition so that the brightness value becomes a predetermined reference brightness value. The area for calculating the average luminance value may be determined for each estimated imaging region, imaging direction, and imaging method.

[Step 23]: Device control / photographing Detailed control determined in Step 22b or Step 22
The information on the basic control determined in a is sent to the X-ray controller 12, the compensation filter controller 14, and the collimator controller 15 and controlled by each controller to capture an image.

In the above-mentioned example, the three-layer hierarchical neural network is used for estimating the imaged part, the imaged direction, and the imaging method, but the structure of the neural network is not limited to three layers. Further, as shown in FIG. 8, device information 81 such as bed position and arm angle is displayed.
Can be taken in and estimated. For example, subject 1
Since it is possible to estimate the rough body shape from the height and age of the person, by incorporating the position of the head of the subject 1 in the estimation, it is possible to roughly estimate the movement amount of the bed and the arm position based on the position of the head. The part to be imaged can be estimated. Also, the shooting direction can be estimated by taking in the arm angle. This is estimated N. N. By capturing in accordance with 31, it is possible to more efficiently estimate the part / imaging direction. In addition to the estimation using the neural network, it is also possible to estimate the imaging region / direction only by combining the above-described subject information and device information.

In addition to the above embodiment, an embodiment will be described in which the contents of arithmetic processing for image data are determined based on the estimation result and displayed. The contents of the arithmetic processing are general image processing involving display, for example, filter processing, gradation conversion processing, smoothing processing, edge emphasis processing, and enlargement / reduction processing. For example, in the case of gradation conversion processing, since the gradation conversion curves suitable for the double contrast image / filled image of the stomach are different, the N.D. N. It is possible to perform automatic gradation conversion of the image data using the gradation conversion table that is estimated in step 31 and is prepared in advance for each imaging region or imaging method based on the estimation result and then displayed. The estimation in this embodiment can also be estimated and processed from image data after photographing other than the fluoroscopic image data described in the previous embodiment.

[0027]

As described above, according to the present invention, at least one of the imaging region, the imaging direction, and the imaging method is estimated to determine the imaging condition, and the collimator / compensation filter of the collimator / compensation filter is selected according to the imaging condition. Since automatic insertion and automatic switching of X-ray conditions can be performed, it is possible to provide a digital X-ray imaging apparatus that reduces the burden on the operator during imaging and improves examination efficiency. Further, it is possible to determine the arithmetic processing contents of the image data based on the above estimation result, perform image processing suitable for each image data and display it, so that the burden on the observer is reduced and the diagnostic efficiency is improved. A digital X-ray imaging apparatus capable of performing the above is provided.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a digital X-ray imaging apparatus according to an embodiment of the present invention.

FIG. 2 is a flowchart showing digital X-ray imaging processing according to an embodiment of the present invention.

FIG. 3 is a diagram illustrating an imaging region / imaging direction / imaging method estimation process according to an embodiment of the present invention.

FIG. 4 is a diagram showing a compensation filter according to an embodiment of the present invention.

FIG. 5 is a diagram illustrating basic control conditions for digital X-ray photography according to an embodiment of the present invention.

FIG. 6 is a diagram illustrating insertion of a compensation filter according to an embodiment of the present invention.

FIG. 7 is a diagram illustrating a method of determining the insertion position of a compensation filter according to an embodiment of the present invention.

FIG. 8 is a diagram illustrating an imaging region / imaging direction / imaging method estimation process according to an embodiment of the present invention.

[Explanation of symbols]

1: subject, 2: X-ray tube, 3: I.D. I. 4: X-ray compensation filter, 6: TV camera, 7: A / D converter,
8: arithmetic processing device, 9: image memory, 10: D / A converter, 11: monitor, 31: estimation neural network, 6
2: Compensation filter insertion area, 63, 74, 77: Compensation filter insertion position, 71: Edge point sequence, 72: Approximate straight line,
75: principal axis of inertia.

Claims (11)

[Claims] 1. An X-ray source, a concentration compensation filter arranged on the front surface of the X-ray source for attenuating X-rays, and a transmission from the object which is arranged so as to face the X-ray source with an object interposed therebetween. An X-ray detector for converting an X-ray image into a visible light image, a television camera for converting the visible light image into an electric signal, an analog-digital converter for converting the electric signal into a digital signal, and the digital signal An image memory for storing as image data, an arithmetic processing device for reading the digital signal and performing arithmetic processing, an analog-digital converter for converting a digital signal processed by the arithmetic processing device into an analog signal, and the analog signal A digital X-ray imaging apparatus having a display device for displaying an image of at least one of an imaging region, an imaging direction, and an imaging method based on fluoroscopic image data. And a processing step of estimating,
A digital X-ray imaging apparatus having either a processing step for determining control of imaging conditions based on an estimation result and the fluoroscopic image data or a processing step for determining contents of arithmetic processing in the arithmetic processing apparatus. . 2. The digital X-ray imaging apparatus according to claim 1, wherein the control of the imaging condition includes control of the density compensation filter. 3. The process according to claim 1, wherein the process of estimating at least one of the imaging region, the imaging direction, and the imaging method is based on the fluoroscopic image data and device information. Digital X-ray equipment. 4. The digital X-ray imaging apparatus according to claim 1, further comprising a neural network used in a process of estimating at least one of the imaging region, the imaging direction, and the imaging method. 5. The digital X-ray imaging apparatus according to claim 1, wherein the arithmetic processing content is filter processing. 6. The digital X-ray imaging apparatus according to claim 1, wherein the arithmetic processing content is gradation conversion processing. 7. The control of the density compensation filter is performed for the first perspective image data in which the density compensation filter is inserted up to an arbitrary position, the density compensation filter position and the density compensation filter measured in advance. A process of creating second perspective image data in a state in which the density compensation filter is not inserted based on the X-ray attenuation characteristic, and determining an insertion position of the density compensation filter based on the second perspective image data. The digital radiography apparatus according to claim 2, further comprising steps. 8. The digital X-ray imaging apparatus according to claim 3, wherein the information of the apparatus includes a bed position and an X-ray source arm position. 9. The digital X-ray imaging apparatus according to claim 4, wherein the input value to the neural network includes a value based on an average luminance value of each block obtained by dividing the fluoroscopic image data into a plurality of blocks. . 10. The processing step of determining the insertion position of the density compensation filter calculates an insertion area of the density compensation filter from the perspective image data, and approximates an edge point sequence of the insertion area located on the center side of the image. 8. The digital X-ray imaging apparatus according to claim 7, wherein a straight line is calculated, and the insertion position of the density compensation filter is determined based on the inclination of the approximate straight line. 11. The processing step of determining the insertion position of the density compensation filter calculates an insertion area of the density compensation filter from the perspective image data, calculates a principal axis of inertia of the insertion area, and uses the principal axis as a reference. 8. The digital X-ray imaging apparatus according to claim 7, wherein the insertion position of the density compensation filter is determined.