JP3223428B2 - Chest radiographic imaging position determination device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for determining the position of a radiographic image of a chest, and more particularly to an apparatus for automatically determining the position of a radiographic image of the human chest during radiography.

[0002]

2. Description of the Related Art Radiation images such as X-ray images are widely used for medical purposes, and the following methods are available for obtaining such radiation images as electric image signals. That is, the radiation that has passed through the human body as a subject is absorbed by a certain kind of phosphor, and then the phosphor is excited by light or heat energy, so that the radiation energy that the phosphor has accumulated by the absorption is absorbed. Radiation is emitted as fluorescence, and this fluorescence is detected by a photoelectric conversion element to electrically obtain radiation image information. After digitizing the radiation image signal, the radiation image signal is subjected to image processing such as gradation processing and spatial frequency processing. The data is output to a CRT or the like for visualization (see Japanese Patent Application Laid-Open No. 63-189853).

By the way, in the processing stage before outputting a digital radiation image signal as a visible image as described above, a gradation image and a spatial frequency process are performed as described above to obtain a visible image suitable for image interpretation. Although it is necessary, even if the image data is of the same part of the human body, if the image is taken while changing the photographing position (the direction of the human body as the subject), the density of the region of interest changes in each reproduced image. Sometimes.

For example, in order to diagnose the thoracic vertebra, when the chest of the human body is photographed from the front, the thoracic vertebra, which is the region of interest, overlaps the mediastinum, which is difficult to transmit radiation. It will overlap with the lung field that is easy to penetrate.
Therefore, when such image data is processed under the same image processing conditions and then reproduced, the thoracic vertebrae portion has a relatively low density in the front image and a relatively high density in the side image.

[0005] Therefore, a method and apparatus for performing optimal image processing for each position by automatically judging the position of the chest radiographic image,
Conventionally, various proposals have been made. As this method and apparatus, there are a method using a density histogram of image data and a method using a signal level distribution (hereinafter, referred to as profile information) along a predetermined image direction in the image data. (Japanese Unexamined Patent Publication No. Sho 6)
JP-A-3-262128, JP-A-63-262132, JP-A-63-262134, JP-A-63-262134
Japanese Patent Application Laid-Open No. 63-262129, Japanese Patent Application Laid-Open No. 63-262135, and the like, which determine the degree of separation of bimodal peaks (see Japanese Patent Application Laid-Open Nos. 63-262129 and 63-262135). No. 106642, JP-A-63-2
No. 62130, JP-A-63-262131,
JP-A-63-262137, JP-A-63-262
No. 138 publication).

Further, a method for obtaining a representative value or a characteristic value from a density histogram (see Japanese Patent Application Laid-Open No. 63-106641, etc.), a method for obtaining a variance of the density histogram (see Japanese Patent Application Laid-Open No. 63-262133, etc.), There has been proposed a method for calculating the sum of signal values near the center of profile information in the left-right direction (see JP-A-63-262136).

[0007]

However, according to the above method, even when a density histogram is used,
Even when profile information is used, a large amount of calculation for automatically judging the body position in which the image has been taken poses a problem. For example, a standard pattern of front image data and a standard pattern of side image data are obtained from many pieces of image data, and the similarity between the standard pattern and subsequently input data is checked (Japanese Patent Laid-Open No. 63-26163).
In Japanese Unexamined Patent Publication No. 2130 and Japanese Unexamined Patent Application Publication No. 63-262137, a complicated method such as multiple regression analysis must be used to obtain a standard pattern, which increases the amount of calculation.

When the photographing position is determined using the density histogram, the photographing position is determined by grasping the difference in shape of the density histogram. Since the shape difference between the density histogram of the image data and the density histogram is small, erroneous determination is increased. Furthermore, when the photographing position is determined using the profile information, if there is a lateral displacement of the human body within the photographing range, the profile pattern based on the image data changes even if the photographing position is the same. Therefore, there is a problem that it becomes impossible to secure the accuracy of determining the photographing position.

SUMMARY OF THE INVENTION The present invention has been made in consideration of the above problems, and has as its object to provide an apparatus capable of accurately determining the photographing position of a subject in a chest radiographic image with a small amount of calculation.

[0010]

SUMMARY OF THE INVENTION Therefore, an apparatus for determining a radiographic position of a chest radiographic image according to the present invention determines the radiographic position of a radiographic image of the chest formed corresponding to the amount of radiation transmitted through the human chest. And is configured as shown in FIG. In FIG. 1, a spine position detecting means detects a spine position in an image based on an image signal of a chest radiographic image, and a spine characteristic analyzing means detects a spine linearity or a spine based on a spine position detection result. Blank in image
Analyze the distance to the shaded part . Then, the photographing posture determining means is configured to analyze the linearity of the spine analyzed by the spine characteristic analyzing means .
Alternatively, it is determined whether the imaging position in the chest radiographic image is the front or the side based on the distance from the spine to the plain portion in the image .

[0011]

In other words, when viewed from the front (or the back), the spine of the human chest extends linearly in the middle of the approximate body, and an image substantially symmetric about the spine can be obtained. On the other hand, when viewed from the side, the spine is located closer to the back than the center of the body, so that there is no symmetry about the spine, and the spine is curved. The thoracic radiation image radiographing position determining apparatus according to the present invention detects the vertebral column position in the image so that the radiographing position can be determined using such basic properties of the vertebral column, and based on the result, the vertebral column position is determined . Linear or spine
By analyzing the distance from the pillar to the transparent portion in the image , and comparing the analysis result with the basic properties, it is determined whether the photographing position is the front or the side.

[0012]

Embodiments of the present invention will be described below. FIG. 2 showing an embodiment of the present invention is a medical radiation image recording and reading apparatus including the apparatus for determining the position of a chest radiographic image according to the present invention, in which a human chest M is photographed and an image of the photographed chest M is taken. The following description will be made on the assumption that the region of interest in the chest is diagnosed by reproducing the region of interest.

Here, the radiation source 1 is controlled by the radiation control device 2 to emit radiation (generally, X-rays) toward the human chest M. The recording and reading device 3 includes a radiation image conversion panel 4 on a surface facing the radiation source 1 with the chest M interposed therebetween.
The energy according to the radiation transmittance distribution of the chest M with respect to the irradiation radiation dose from the camera is accumulated in the stimulable phosphor layer, and a latent image of the chest M is formed there.

The conversion panel 4 is provided with a stimulable phosphor layer on a support by vapor deposition of a stimulable phosphor or application of a stimulable phosphor paint. The layer is shielded or covered by a protective member to prevent adverse effects and damage due to the environment. As the stimulable phosphor material, for example, materials disclosed in JP-A-61-72091 or JP-A-59-75200 are used.

A light beam generator (gas laser, solid-state laser, semiconductor laser, etc.) 5 generates a light beam whose emission intensity is controlled, and the light beam reaches a scanner 6 via various optical systems. , Where it is deflected,
To deflect the light path, and is guided to the conversion panel 4 as stimulating excitation scanning light. The light collector 8 has a light collecting end made of an optical fiber positioned close to the conversion panel 4 on which the stimulating excitation light is scanned, and is proportional to the latent image energy from the conversion panel 4 scanned by the light beam. The photostimulable light having the light emission intensity is received. Reference numeral 9 denotes a filter that passes only light in the photostimulated emission wavelength region from the light introduced from the light collector 8, and the light that has passed through the filter 9 enters the photomultiplier 10 and corresponds to the incident light. The current signal is photoelectrically converted into a current signal.

The output current from the photomultiplier 10 is converted into a voltage signal by a current / voltage converter 11, amplified by an amplifier 12, and then converted by an A / D converter 13 into digital data (a digital signal proportional to the radiation transmission amount). (Radiation image signal). The digital image signal proportional to the radiation transmission amount of each part of the subject is sequentially image-processed in the image processing device 14, and the image signal after the image processing is transmitted to the printer 17 via the interface 16. I have.

Reference numeral 15 denotes a CPU for controlling image processing in the image processing device 14, which performs various image processing (eg, spatial frequency processing, enlargement, reduction, etc.) on digital radiation image data output from the A / D converter 13. , Movement, rotation, statistical processing, etc.) in the image processing device 14 and output to the printer 17 in a form suitable for diagnosis, so that the printer 17 can obtain a hard copy of the chest radiation image. Here, in order to set an appropriate processing condition in the image processing, it is determined whether the imaging position in the chest radiographic image is the front or the side, and the image processing condition is determined based on the determination result. I have to.

Note that what is connected via the interface 16 may be a monitor such as a CRT or a storage device (filing system) such as a semiconductor storage device. Reference numeral 18 denotes a read gain adjusting circuit, which adjusts the light beam intensity of the light beam generator 5, the gain of the photomultiplier 10 by adjusting the power supply voltage of the photomultiplier high-voltage power supply 19, and the current / voltage converter. The gain of the amplifier 11 and the amplifier 12 and the adjustment of the input dynamic range of the A / D converter 13 are adjusted, and the reading gain of the radiation image signal is adjusted comprehensively.

The portion of the image processing device 14 relating to the photographing position determination according to the present invention is specifically configured as shown in FIG. The digital radiation image signal of the chest M photoelectrically read from the stimulable phosphor layer of the conversion panel 4 is first converted into a spine position detecting section 21 as a spine position detecting means.
Here, the position of the spine in the chest radiographic image is detected based on the signal level distribution (profile information) in the direction crossing the spine.

The detection result of the spine position in the spine position detection unit 21 is output to a characteristic analysis unit 22 as a spine characteristic analysis unit, where the analysis of the detection result of the spine position is performed. The analysis is for analyzing characteristics of a spine in a chest image known in advance, and the linearity of the spine is analyzed.
Or, analyze the distance from the spine to the blank part in the image
You.

A body position determining section 23 as a body position determining means determines the photographing position based on the analysis result of the characteristic analyzing section 22 as to whether the chest radiation image is a front image or a side image. Then, the image processing condition determination unit 24 determines the conditions of the image processing such as the gradation processing based on the determined photographing postures, in accordance with each photographing posture. The image processing unit 25 performs image processing such as gradation processing on the chest radiation image signal in accordance with the condition determined by the image processing condition determination unit 24, and outputs a signal of the processing result to the printer.

The spinal column position detector 2 other than the gradation processor 23
1, characteristic analysis unit 22, posture determination unit 23, image processing condition determination unit 2
The image signal used in step 4 does not need to be performed using the total number of pixels read by the recording / reading device 3. The process of determining body posture and determining gradation processing conditions is simplified by using an image signal thinned out from the original image signal. You may make it.
Further, a posture determination and an image processing condition determination are performed based on an image signal obtained by “read ahead” as disclosed in Japanese Patent Application Laid-Open No. 58-67240, and “main reading” is performed according to the determined processing condition. The obtained image signal may be processed. The `` read-ahead '' is to read an image signal by stimulating emission with excitation light having a lower energy than the `` read-ahead '' prior to the `` read-ahead '' provided for observation reading, which is obtained by the `` read-ahead '' In response to the determination of the image processing condition based on the obtained image signal, the image signal obtained by the “real reading” is subjected to image processing.

Next, the determination of the photographing position according to the present invention performed in the image processing apparatus 14 whose detailed configuration is shown in FIG. 3 will be described in detail. First, in the spine position detection unit 21,
The position of the spine in the chest radiation image is detected by detecting the signal level distribution (profile information) in the direction crossing the spine. In this embodiment, it is assumed that the vertical direction of the image always coincides with the vertical direction of the human body.

The detection of the spine position using projection as profile information is performed as follows. For example, as shown in FIG. 4 or 5, a plurality of band-shaped regions are set in the vertical direction in a direction crossing the spinal column, and the integrated value (average value) of the signals in the pixel rows arranged in the vertical direction is set for each band-shaped region. A projection is created as profile information indicating how it changes in the horizontal direction, and in each of the projections created for each band-shaped area, the horizontal position in the pixel row where the signal level is the minimum point is detected.

That is, in a chest radiographic image, generally, the bone portion, particularly the spine (thoracic vertebra), has the lowest radiation transmittance, so that the signal level distribution in the direction across the spine is the smallest. Can be considered as the point of intersection with the spine. When a plurality of minimum points are detected, the minimum point closest to the center in the horizontal direction may be specified as a point intersecting the spine. In the case where a plurality of band-shaped regions are set in the vertical direction as described above, a region from the neck to approximately the diaphragm may be set as the region setting region.

As described above, if the vertebral column position is detected as the minimum point of projection for each of a plurality of band-shaped regions set in the vertical direction, the line connecting the detection points in each band-shaped region indicates the vertebral column. Therefore, two-dimensional position setting of the spine in each image is performed by directly connecting the spine points obtained discretely by a straight line or by approximating by a quadratic or cubic function.

In the case where a profile is created as a signal level distribution of one horizontal pixel row, a profile is created at a plurality of locations in the vertical direction in the same manner as described above (see FIGS. 6 and 7). Of the vertebral column is determined as the minimum point in the profile, and a process of connecting a plurality of the determined vertebral points is performed to detect the vertebral column position in each image.

When the position of the spine is detected as described above, the characteristic analysis unit 22 analyzes the detection result, and based on the analysis result, the position determination unit 23 determines whether the photographing position is front or side. Is determined. First, the case of analyzing the linearity of the spine is described. Normally, when the chest is photographed from the front, the spine is almost straight, but it is detected because the spine is curved in the side view. If the spine is substantially straight, it can be determined that the image is a front image, and if it is curved, it can be determined that the image is a side image.

The linearity is determined, for example, by referring to FIGS.
As shown in, the upper end position and the lower end position of a part of the spine detected by the spine position detection unit 21 are linearly connected, and the area of the area surrounded by a line connecting the upper and lower ends with a straight line and the spine line is defined as It can be determined based on whether the area is equal to or greater than a predetermined determination level. If the detected spine is substantially straight, the area is smaller, and the area is larger as the curvature is larger, so when the area is equal to or greater than the predetermined determination level, the chest image is taken from the side, When it is lower than the determination level, it can be determined that the vehicle is in front.

In addition, the linearity can be determined by the method described above.
As shown in FIG. 10 and FIG. 11, the calculation can be performed by obtaining a straight line passing through a plurality of points above the detected plurality of spine points, and calculating a deviation between the straight line and the spine. Also in this case, the shift is small when the chest image is the front, and the shift is large when the chest image is on the side, so that the photographing position can be determined based on the shift amount. The deviation can be detected by integrating the distance from each spine point to the straight line, or by calculating the area of a region surrounded by the spine line and a straight line passing through the plurality of upper points.

Further, when connecting a plurality of vertebral column points obtained by approximation by a quadratic or cubic function, the degree of curvature of the vertebral column is determined by the coefficient of the approximate function expression, and the degree of curvature is large. also as possible out be determined to be a side image when.

[0032]

On the other hand, from the vertebral column to the blank part in the image
By analyzing the distance, the photographing position (front or side)
Surface) can be determined. That is, in the chest front image,
The vertebral column is located in the approximate center of the human body, and in the side view, the vertebral column is located not on the center but on the side closer to the back, so in the case of the front, radiation does not pass through the human body from the vertebral column.
The distance to the plain part is relatively large ( when the plain part of the flank is symmetric except for the plain part of the neck and shoulders), while in the case of the side, the back side As a result, the distance from the spine to the bare part becomes shorter. Therefore, it is also possible to determine whether the photographing position is the front or the side by analyzing the distance from the spine to the plain part (the positional relationship between the spine and the plain part). .

The analysis of whether or not the spine is close to the spine is performed, for example, in the lower half of the image, in order to exclude the neck and the shoulder in the front image. The distance to the plain portion is obtained at a plurality of locations, and the shortest one is detected from these. Then, the shortest distance is compared with a predetermined value, and when the shortest distance is equal to or less than the predetermined value, the shortest distance is regarded as being detected as the distance between the straight back portion of the back side and the spinal column, and the photographing position is regarded as the side position. If the shortest distance is a relatively large value corresponding to the distance from the center to the flanks in the front image, it is determined that the image is a front image (see FIGS. 12 and 13).

In addition to the above-described method, the analysis of whether or not the spine is close to the hollow part is performed by the average of the horizontal coordinate values of the spine and the horizontal direction of the boundary (edge) between the hollow part and the human body. It is also possible to compare with the average of the coordinate values and make a determination based on the deviation between them. Here, the boundary portion between the transparent portion and the human body can be obtained by calculating the differential value of the profile in the horizontal axis direction, and the points located on both sides in the peak value of the differential value as the boundary portion. Further, in specifying the missing portion, a histogram of the chest image signal is created as shown in FIG. 14, and the peak portion near the maximum signal of this histogram is recognized as corresponding to the signal of the missing portion. By allowing a portion having a predetermined signal width near the maximum signal of the histogram to be recognized as a signal value corresponding to a blank portion, a blank region in an image can be specified.

[0036]

[0037]

[0038]

As described above, the position of the vertebral column is detected in the image, and the linearity of the vertebral column is determined based on the detection result .
By analyzing the distance to a directly irradiated portion of the Luo image state, and are not able to discriminate the photographing posture, it is possible to relatively easily determine the photographing posture without requiring complex operations. Also,
Based on the spine in the human chest, its linearity and omission
Since the distance to the portion is determined, it is hardly affected by the body type and the photographing conditions, and accurate body position determination is possible.

However, in consideration of the unique cases shown in FIGS. 15 to 17 , there is a possibility of erroneous diagnosis if the analysis result of the linearity of the spine which is easy to calculate is used. Solution of the distance between the vertebral column and the plain part in the image, although it is complicated
It is desired to perform the analysis.

The example shown in FIG. 15 shows a case where a patient with scoliosis is photographed from the front. In this case, although the degree of curvature of the spine in the side image is smaller, the linearity of the spine is analyzed to determine the photographing position. In such a case, it is difficult to optimally set a threshold value for determining the degree of curvature, which may lead to erroneous determination .

In some cases, as shown in FIG. 16 , in the chest side image, the spine is photographed such that the spine is located substantially at the center of the image. obtaining Ri mower when taken shifted to. Furthermore,
As shown in FIG. 17 , it is assumed that one lung has been removed or reduced, and in this case, the symmetry of the left and right regions with the spine as an axis is impaired . Is considered , the distance between the spine and the clear part in the image
It would most desirably be analyzing.

[0043]

In the analysis of the characteristics of the spine, the photographing position is set to either the front or the side based on the analysis result.
Instead of making an off-line determination, fuzzy inference or the like may be used to set the likelihood of being a front image or a side image as a numerical value. When a definite discrimination result cannot be obtained from the certainty, a simple analysis ( analysis of linearity) to a more complicated analysis (abstract portion)
Or so as to shift to analysis) of the distance to the straight line
The determination result (the likelihood of being a front image or a side image) based on the gender and the distance to the blank portion may be comprehensively determined to finally determine the photographing position.

Furthermore, the final position determination result is not clearly shown as either the front or the side, but is ambiguous expression (membership value in fuzzy inference).
Or an unknown determination result may be provided in addition to the front or side surface. As described above, when the imaging position of the chest radiographic image is determined to be either the front or the side, the image processing condition determination unit 24 uses the determined imaging position as at least information, preferably the signal level of the region of interest. The image processing conditions such as gradation processing and spatial frequency processing are determined with reference to the image processing section.
The image processing in 25 is performed.

It should be noted that a chest radiographic image signal that has been subjected to image processing based on processing conditions set based on the imaging position determination is:
The hard copy may be immediately performed by the printer 17 as described above. However, the hard copy may be reproduced on a CRT, or may be temporarily stored in a filing system and read out when necessary to be hard copied or displayed on the CRT. May be.

When storing the chest radiographic image in the filing system, it is preferable to store the information of the imaging position in correspondence with each chest image signal, and the determined result is not necessarily the determination of the image processing condition. The image signal does not need to be intended, and the image signal before image processing may be stored together with the information of the photographing position. Further, in the present embodiment, the chest radiation image signal photoelectrically read from the stimulable phosphor layer is configured to perform gradation processing. However, the present invention is not limited to image reading using a stimulable phosphor. Alternatively, a configuration using other two-dimensional radiation detectors may be used. For example, on a silver halide film that records a chest radiographic image,
A configuration may be adopted in which light from a light source such as a laser or a fluorescent lamp is applied to obtain transmitted light through the silver halide film, and the transmitted light is photoelectrically converted to obtain a chest radiation image signal.

[0048]

As described above, according to the chest radiation image photographing position determining apparatus according to the present invention, the photographing position is determined by utilizing the characteristics of the spine (linearity or the distance to the unpenetrated portion), which can be easily detected. It is possible to easily determine whether the front is the side or the side, and it is possible to accurately determine the posture without being affected by the body type and the imaging conditions.For example, the image processing conditions are determined based on the determination of the body position In this case, there is an effect that the diagnostic performance for medical use can be improved by setting the optimum processing conditions.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a basic configuration of the present invention.

FIG. 2 is an overall system block diagram showing one embodiment of the present invention.

FIG. 3 is a block diagram illustrating a configuration of an image processing unit according to the embodiment.

FIG. 4 is a diagram showing detection of a spinal column position based on profile information.

FIG. 5 is a diagram showing spine position detection based on profile information.

FIG. 6 is a diagram showing detection of a spinal column position based on profile information.

FIG. 7 is a diagram showing detection of a spinal column position based on profile information.

FIG. 8 is a diagram showing linearity analysis of a spine based on area comparison.

FIG. 9 is a diagram showing linearity analysis of a spinal column based on area comparison.

FIG. 10 is a diagram showing a linearity analysis of a spinal column based on a deviation from linear approximation.

FIG. 11 is a diagram showing a linearity analysis of a spinal column based on a deviation from linear approximation.

FIG. 12 is a diagram showing an analysis of a distance between a plain part and a spine.

FIG. 13 is a diagram showing an analysis of a distance between a hollow part and a spine.

FIG. 14 is a diagram for explaining identification of a blank portion using a histogram.

FIG. 15 shows the state of the spine in the case of scoliosis.
FIG.

FIG. 16 when the spine position is set to the center in the side image
FIG.

FIG. 17 shows a state in which one lung has been removed in the front image .
FIG.

──────────────────────────────────────────────────続 き Continued on the front page (58) Fields surveyed (Int. Cl. ⁷ , DB name) A61B 6/00-6/14 G06T 1/00

Claims (2)

(57) [Claims] An apparatus for determining the position of a thoracic radiation image formed in correspondence with the amount of radiation transmitted through a human chest, wherein a position of a spine in the image is detected based on an image signal of the thoracic radiation image. the basis of the spine position detecting means for, based on the detection result of the spine position by該脊column position detecting means, a spine characteristics analyzing means for analyzing the linearity of the spinal column, the linearity of the spine analyzed by該脊column characterization means And a photographing position discriminating means for discriminating whether the photographing position in the chest radiographic image is the front or the side, and a photographing posture discriminating apparatus for a chest radiographic image. 2. An apparatus for judging the position of a radiographic image of a chest formed corresponding to the amount of radiation transmitted through the human chest, wherein a position of a spine in the image is detected based on an image signal of the radiographic image of the chest. A vertebral column position detecting means, a vertebral column characteristic analyzing means for analyzing a distance from the vertebral column to a transparent portion in an image based on a result of the vertebral column position detection by the vertebral column position detecting means, In the image from the spine
Imaging position determination means for determining whether the imaging position in the chest radiographic image is front or side based on the distance to the shadowed portion ; and radiographing position determination of the radiographic image of the chest characterized by comprising: apparatus.