JP2006087934A - Image processing apparatus and image processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform image processing suitable for diagnosis corresponding to a change in photographing region and photographing condition. <P>SOLUTION: This image processing apparatus includes: a histogram creating means for creating a histogram showing the frequency of each interval of a radiation image signal; an object histogram separating means for separating an object histogram corresponding to an object area from the histogram; a threshold determining means for determining a predetermined proportion value of the sum of frequencies in the object histogram as the threshold of frequency; a desired image signal area determining means for determining a desired image signal frequency on the basis of comparison between the frequency of each interval of the image signal in the object histogram and the threshold of the frequency; an image processing condition determining means for determining an image processing condition for the radiation image on the basis of a characteristic amount of the image signal included in the desired image signal area; and an image processing means for processing the radiation image according to the image processing condition. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an image processing apparatus and an image processing method.

  Radiation images such as X-ray images are often used for disease diagnosis, etc. In order to obtain this X-ray image, X-rays transmitted through the subject are irradiated onto a phosphor layer (phosphor screen), thereby Conventionally, so-called radiographs, in which visible light is generated and developed by irradiating a film using a silver salt with the visible light in the same manner as ordinary photographs, have been used.

  However, in recent years, a method for taking out an image directly from a phosphor layer without using a film coated with silver salt has been devised.

  In this method, the radiation that has passed through the subject is absorbed by the phosphor, and then the phosphor is excited by light or thermal energy, for example, so that the radiation energy accumulated by the phosphor is absorbed as fluorescence. There is a method of radiating, photoelectrically converting the fluorescence, and further A / D converting to obtain a digital image signal (see, for example, Patent Document 1).

  The radiographic image signal thus obtained is output as it is or visualized as it is or after being subjected to image processing.

  In addition, a silver salt film on which a radiographic image is recorded is irradiated with light from a light source such as a laser or a fluorescent lamp to obtain transmitted light of the silver salt film, and the transmitted light is photoelectrically converted to obtain a radiographic image signal. There is also a method.

  By the way, when the radiological image signal obtained as described above is reproduced, the purpose of making the concentration of the region of interest (image part necessary for diagnosis in medical use) constant in the reproduced image, and the structure and lesion of the human body In order to output the shadow of the image more easily, various image processing such as gradation processing and spatial frequency processing is performed, and then output to a CRT or the like for visualization and used for diagnosis.

  Conventionally, as a method of determining the image processing condition, there is a method of inputting information related to an imaging region and an imaging method of a subject and determining an image processing condition suitable for the imaging region and the imaging method based on the information ( For example, see Patent Document 2.)

  In addition, there is a method of determining gradation processing conditions for the spine, heart, and lung field based on the histogram shape of the image signal, particularly for chest radiographic images (see, for example, Patent Document 3).

  Further, there is a method of determining a signal range based on a characteristic value of the cumulative histogram and performing gradation processing conditions so that the range corresponds to a predetermined output signal range. For example, a signal value at which a frequency cumulative value is 50% A signal range having a certain width before and after the center is determined (for example, see Patent Document 4).

Furthermore, there has been a method of determining image processing conditions with the point at which the frequency of the histogram falls to 5% of the maximum frequency as the maximum value / minimum value of the desired signal range (see, for example, Patent Document 5).
Japanese Patent Laid-Open No. 55-12144 JP-A-61-68031 JP 55-116339 A JP 63-31641 A JP-A-55-88740

  However, the conventional image processing condition determination method has the following problems.

  The method disclosed in Patent Document 2 has a configuration in which image processing cannot be performed unless information on an imaging region is given, and the user needs to input the imaging region using a menu button or the like, and the operation is complicated. In addition, if the computer automatically recognizes the site to be imaged, a complex algorithm for site recognition is required, and a processing algorithm must be defined in advance for all of the sites. There is a problem that the memory capacity to be increased.

  In addition, the method disclosed in Patent Document 3 has a problem in that since the histogram shape varies depending on the part, it cannot be used for other parts and is not versatile.

  In addition, in the method disclosed in Patent Document 4, there are cases where a part that is unnecessary for diagnosis depending on the part is included in the image, and there is a case where the part is not included. There is a problem that a desired image signal area is not necessarily selected due to an influence of an image portion which is not necessary for the image processing.

  Further, in the method disclosed in Patent Document 5, a signal value having the maximum frequency may be a signal of an image part unnecessary for diagnosis, and is unstable as a reference for determining a desired image signal range. There was a problem.

  For example, in a radiological image of the front of the abdomen or the front of the lumbar vertebra, a histogram shape as shown in FIG. 8 is generally used, and an image portion of a portion necessary for diagnosis in either the method using the cumulative histogram value or the method based on the maximum frequency However, in the radiographic images of the chest, head, limb bones, etc., there are regions where radiation is not necessary for diagnosis (regions where radiation does not pass through the subject), so the histogram shape is As shown in FIG. 9, there is a concern that the method based on the maximum frequency may not be able to specify an image portion necessary for diagnosis. Furthermore, in a radiographic image of a cervical spine, as shown in FIG. 10, since there are unexposed areas of radiation and image portions (such as jaws and shoulders) unnecessary for diagnosis on the low signal side, a cumulative histogram is present. Neither the method using the value nor the method based on the maximum frequency can accurately identify the image portion necessary for the diagnosis, and thus cannot set an appropriate image processing.

  The present invention has been made in view of the above-mentioned problems. Appropriate image processing conditions can be determined by a simple algorithm for a wide range of radiographic images, and a wide range of radiographic images can be obtained by image processing based on such image processing conditions. The purpose is to make it easy to see.

  The above object is achieved by the following invention.

(1) Histogram creation means for creating a histogram indicating the frequency for each predetermined section width of the image signal of the radiation image formed corresponding to the amount of radiation transmitted through the subject;
Subject histogram separation means for separating and creating a subject histogram corresponding to the subject area from the histogram created by the histogram creation means;
Threshold value determining means for obtaining a sum of frequencies in the subject histogram separated and created by the subject histogram separating means, and determining a predetermined ratio value of the sum as a frequency threshold value;
A desired image signal region is determined based on a comparison between a frequency for each predetermined section width of the image signal in the subject histogram separated and created by the subject histogram separation unit and a frequency threshold value determined by the threshold value determination unit. A desired image signal region determining means;
Image processing condition determining means for determining an image processing condition for the radiation image based on a feature amount of an image signal included in the desired image signal area determined by the desired image signal area determining means;
An image processing apparatus comprising: an image processing unit configured to perform image processing on the radiation image based on the image processing condition determined by the image processing condition determining unit.

  (2) The desired image signal region determining means includes a frequency for each predetermined section width of the image signal in the subject histogram separated and created by the subject histogram separating means, and a frequency threshold determined by the threshold determining means. The image processing apparatus according to (1), wherein at least one of a maximum value and a minimum value of a desired image signal region is determined based on the comparison of the above.

  (3) The histogram creating means includes an irradiation field identifying means for identifying a radiation field based on the image signal of the radiation image, and the image signal in the radiation field identified by the irradiation field identifying means. The image processing apparatus according to any one of (1) and (2), wherein a histogram is created using

(4) a histogram creating step for creating a histogram indicating the frequency of each predetermined section width of the image signal of the radiation image formed corresponding to the amount of radiation transmitted through the subject;
A subject histogram separation step of separating and creating a subject histogram corresponding to a subject region from the histogram created in the histogram creation step;
A threshold determination step of calculating a sum of frequencies in the subject histogram separated and created in the subject histogram separation step, and determining a predetermined ratio value of the sum as a frequency threshold;
A desired image signal region is determined based on a comparison between the frequency for each predetermined section width of the image signal in the subject histogram separated and created in the subject histogram separation step and the frequency threshold value determined in the threshold value determination step. A desired image signal region determination step;
An image processing step of performing image processing on the radiographic image under an image processing condition determined based on a feature amount of an image signal included in the desired image signal region determined in the desired image signal region determining step. Image processing method.

  According to the invention described in (1) or (4), it is possible to determine the desired image signal area by excluding the signal of the radiation missing area, and without being affected by the fluctuation of the frequency in the desired image signal area. The desired image signal area can be determined, so that an appropriate image processing condition can be determined, and the boundary of the desired image signal area is detected based on a comparison between the threshold value and the frequency on the subject histogram. Therefore, there is an effect that the desired image signal region can be stably determined regardless of the histogram shape.

  Embodiments of the present invention will be described below.

  FIG. 2 showing an embodiment is a radiographic image recording / reading apparatus for medical use including an image processing condition determination apparatus for radiographic images and an image processing apparatus according to the present invention, in which each part of the human body (chest, abdomen) The following description will be made on the assumption that M is photographed and the photographed image of each part of the human body is reproduced to diagnose the human body structure and lesion.

  Here, the radiation generation source 1 is controlled by the radiation control device 2 to irradiate the subject with radiation (generally, X-rays). The recording / reading apparatus 3 includes a radiation image conversion panel 4 on a surface facing the radiation source 1 with the subject interposed therebetween. The conversion panel 4 has a radiation transmittance distribution of each part of the human body with respect to the radiation dose from the radiation source 1. The energy thus stored is accumulated in the photostimulable phosphor layer, and a latent image of each part of the human body is formed there.

  The conversion panel 4 is provided with a photostimulable phosphor layer on a support by vapor phase deposition of photostimulable phosphor or application of photostimulable phosphor paint, and the photostimulable phosphor layer is an environment. In order to block the adverse effects and damage caused by the above, the protective member is shielded or covered. As the photostimulable phosphor material, for example, a material disclosed in Japanese Patent Application Laid-Open No. 61-72091 or Japanese Patent Application Laid-Open No. 59-75200 is used.

  A light beam generation unit (gas laser, solid state laser, semiconductor laser, etc.) 5 generates a light beam whose emission intensity is controlled, and the light beam reaches the scanner 6 via various optical systems and deflects there. In addition, the optical path is deflected by the reflecting mirror 7 and guided to the conversion panel 4 as stimulated excitation scanning light.

  The condensing body 8 is located near the conversion panel 4 where the stimulating excitation light is scanned, and a condensing end made of an optical fiber or a sheet-like light guide member is positioned. From the conversion panel 4 scanned with the light beam, A stimulated light emission having a light emission intensity proportional to the latent image energy is received. Reference numeral 9 denotes a filter that passes only light in the stimulated 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 is converted into the incident light. It is photoelectrically converted into a corresponding current signal.

  The output current from the photomultiplier 10 is converted into a voltage signal by the current / voltage converter 11, amplified by the amplifier 12, and then converted into digital data (digital radiation image signal) by the A / D converter 13. . The digital image signal proportional to the amount of radiation transmitted through each part of the subject is sequentially 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. Yes.

  Reference numeral 15 denotes a CPU that controls image processing in the image processing apparatus 14, and performs various image processing including gradation processing on digital radiographic image data output from the A / D converter 13 (for example, spatial frequency processing, dynamic Range compression, enlargement, reduction, movement, rotation, statistical processing, etc.) are performed in the image processing device 14 and output to the printer 17 after having a shape suitable for diagnosis. The printer 17 makes a hard copy of the radiation image of each part of the human body. To be obtained.

  Note that a monitor such as a CRT may be connected via the interface 16, and further, a storage device (filing system) such as a semiconductor storage device may be used.

  Reference numeral 18 denotes a read gain adjustment circuit. The read gain adjustment circuit 18 adjusts the light beam intensity of the light beam generator 5, adjusts the gain of the photomultiplier 10 by adjusting the power supply voltage of the high voltage power supply 19 for the photomultiplier, and current / voltage. The gain adjustment of the converter 11 and the amplifier 12 and 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 part related to the image processing condition determination and the image processing according to the present invention of the image processing device 14 is specifically shown in the block diagram showing the basic configuration of the device according to the invention of FIG. 1 and the block diagram of FIG. The digital radiation image signal of each part M of the human body, which is photoelectrically read from the photostimulable phosphor layer of the conversion panel 4, is first identified by the irradiation field identification unit 21 (irradiation field identification means). That is, when an irradiation field stop is performed at the time of radiography, the radiation image includes a region where the radiation stop is performed (region where radiation is shielded and not irradiated). The irradiation field area is identified as preprocessing so that the signal of the area where the irradiation field stop has been performed can be eliminated in the creation of a histogram described later.

  The irradiation field region is identified using a method as disclosed in, for example, Japanese Patent Laid-Open No. 5-7579. Specifically, after thinning out the digital radiographic image signal, the image area is divided into a plurality of small areas, and for each small area, a variance value of the image signal included in the small area is obtained. Then, the rows and columns of the small areas including the predetermined number or more of the small areas (small areas having a wide variation range of the included image signal) having the variance value equal to or larger than the predetermined value are set as candidates for defining the outline of the irradiation field. Further, based on an image signal in a small region outside the small region set as the irradiation field contour candidate, the correctness of the result of the contour identification is determined, and finally the irradiation field region is determined based on the determination result. Is identified.

  When the identification of the irradiation field region is completed, the histogram creation unit 22 (histogram creation means) creates a histogram indicating the frequency of each image signal based on the image signal in the irradiation field region. The histogram creation in the histogram creation unit 22 may be relatively coarse. For example, the interval width of the image signal may be about 10 for a 10-bit image, that is, an image having 1024 gradations.

  Next, the subject histogram separation unit 23 (subject histogram separation means) separates and creates a histogram of the subject portion from the histogram created by the histogram creation unit 22. That is, even within the irradiation field region, the subject portion is divided into a portion where the subject does not exist and the radiation is directly irradiated to the conversion panel 4 (hereinafter, referred to as a blank region). Therefore, only the image signal of the subject portion is extracted by excluding the image signal of the unclear region.

  For example, when the high-signal portion (the side with a large amount of radiation transmission) has a sharp peak in the histogram of the subject portion, for example, the peak portion (statistical amount) is not included. Created by removing the signal as a signal of the missing area.

  As a method for detecting a peak on a histogram, an image signal region having a frequency exceeding a predetermined frequency (for example, 1/2 of the maximum frequency) is regarded as a “mountain” of the histogram, and the most frequent point in that region is determined. It can be detected as a peak. Further, the peak may be detected by subjecting the histogram curve as it is or after performing a smoothing process and then performing a differentiation process.

  Further, as disclosed in Japanese Patent Laid-Open No. 63-262141, the maximum value S2 of the image signal area of the subject is obtained based on the statistics of the whole histogram, and the signal area exceeding the maximum value S2 is determined as a blank area. Processing to remove as a corresponding signal may be performed. The detection of the maximum value S2 based on the statistic is performed by dividing the image signal on the histogram into two regions based on an arbitrary signal value (a value between the intermediate value and the maximum value), and averaging the frequency included in each region Based on the value, class separation indicating the difference in frequency included in each region is calculated, and the region on the high signal side is identified as a missing region based on the classification of the signal region when the separation becomes the largest. What is necessary is just composition.

  When the subject histogram (histogram consisting only of the signal value of the subject portion) is created as described above, the frequency threshold value used for specifying the desired image signal region is then determined by the frequency threshold value determination unit 24 (threshold value determination unit). It is determined from the subject histogram.

  The threshold is determined as a predetermined ratio value of the sum of the frequencies of the subject histogram (a value corresponding to the area from the minimum value Smin to the maximum value S2 of the image signal on the subject histogram). The predetermined ratio is preferably 0.01% to 1.0%, and more preferably about 0.1%.

  The desired image signal region determination unit 25 (desired image signal region determination means) determines a desired image signal region based on the subject histogram and the threshold value.

  The desired image signal area is characterized as a signal area sandwiched between a minimum value ThresL and a maximum value ThresH, and the threshold value and the frequency on the subject histogram are sequentially set from a predetermined start signal value toward the low signal side. In comparison, a signal value whose frequency is initially less than the threshold is searched, and the signal value is set as the minimum value ThresL (see FIGS. 4 to 6).

  On the other hand, the maximum value ThresH, like the minimum value ThresL, sequentially compares the threshold value with the frequency on the subject histogram from the predetermined start signal value toward the high signal side. A signal value that is less than the threshold value is searched for, and the signal value is set as the maximum value ThresH, or the maximum value S2 of the image signal corresponding to the subject shown on the subject histogram is determined as the maximum value ThresH.

  Here, as a start signal value that is a signal value for starting the comparison between the threshold value and the frequency on the histogram, the frequency peaks at a position closest to the maximum value S2 on the lower signal side than the maximum value S2. It is preferable to use a signal value or a signal value at a predetermined ratio from the minimum value Smin in the signal region between the minimum value Smin and the maximum value S2, and the predetermined ratio is preferably 60 to 80%. It is.

  According to the above configuration, the desired signal region is desired even if the histogram shape changes variously as shown in FIGS. 4 to 6 depending on the presence / absence or size of an image portion unnecessary for diagnosis, the characteristics of the subject, imaging conditions, and the like. It can be obtained stably as an image signal area, and an image suitable for diagnosis can be obtained by appropriately determining the image part necessary for diagnosis for any part of the human body or radiographic image of any imaging condition. An image can be obtained.

  Specifically, for example, as shown in FIG. 6, even if there is a frequency valley in the desired image signal area on the histogram, the subject histogram indicates that the valley portion is erroneously selected as the boundary of the desired image signal area. This can be avoided by setting a threshold value based on the total sum of the frequencies. Further, in the histogram as shown in FIG. 5 in which an image unnecessary for diagnosis is included on the low signal side (histogram of a radiation image of the cervical spine, etc.), the comparison between the threshold value and the frequency on the histogram is started from the high signal side. Therefore, it is possible to reliably eliminate the image area of the portion unnecessary for the diagnosis on the low signal side.

  In the above embodiment, it is not necessary for the user to input information on the imaging region, and automatic recognition of the imaging region requiring a complicated algorithm is not required, so that image processing can be executed at high speed. Furthermore, since it is not necessary to store an image processing condition determination algorithm for each imaging region, the amount of memory used can be reduced.

  If the subject histogram is rough so that the interval width of the image signal is about 10, the minimum value ThresL and the maximum value ThresH can be obtained without being affected by noise.

  When the desired image signal area is determined as an area sandwiched between the minimum value ThresL and the maximum value ThresH as described above, image processing is performed based on the feature amount of the image signal included in the next desired image signal area. A condition, specifically, a conversion table for gradation conversion is determined.

  The gradation conversion table is determined by first determining a reference signal value (feature amount) from the desired image signal area in the reference signal value determining unit 26 (reference signal value determining means). In the reference signal value determination unit 26, an accumulated frequency from the minimum value ThresL to the maximum value ThresH is obtained, and a signal value at a point where the accumulated frequency becomes a predetermined ratio of the whole is set as a reference signal value.

  In determining the reference signal value based on the cumulative frequency from the minimum value ThresL to the maximum value ThresH, in order to accurately determine the reference signal value, the interval width of the image signal is set to 1 for each signal value. It is preferable to obtain the frequency for each.

  The predetermined ratio is preferably 0% to 30%, and more preferably about 10%. This is because, when the subject is a human body, normal bones are included in any part, and stable gradation conversion can be performed by using the bone part as a reference. This is because the reference signal value is determined closer to the low signal.

  Here, for simplicity, a signal value at a predetermined ratio from the minimum value ThresL between the minimum value ThresL and the maximum value S2 may be used as the reference signal value.

  When the reference signal value is determined, the gradation processing condition determination unit 27 (image processing condition determination unit, gradation conversion table setting unit) determines the reference signal value based on the data stored in the basic gradation characteristic storage unit 28 and the reference signal value. A gradation conversion table is set.

  The basic gradation characteristic storage unit 28 (basic characteristic curve storage means) stores in advance a basic characteristic curve indicating basic characteristics of gradation conversion in a coordinate system for gradation conversion composed of input signal values and output signal values. The gradation processing condition determination unit 27 rotates or translates the basic characteristic curve so that the reference signal value is converted into a predetermined output signal value (output density or luminance) set in advance. The desired gradation conversion table is obtained by moving. Here, the correspondence between the predetermined output signal value and the output density or luminance is determined based on the specific characteristics of the image output apparatus (printer, CRT, etc.) to be used.

  The basic gradation storage unit may store a relational expression between an input signal value and an output signal value representing the basic characteristic curve, or each of input signal value to output signal value corresponding to the basic characteristic curve. A lookup table indicating correspondence may be stored.

  A plurality of basic characteristic curves are stored in the basic gradation characteristic storage unit 28, and the plurality of basic characteristic curves are based on the characteristics of the image signal included in the desired image signal area obtained by the histogram processing. It is also possible to select and read out the most appropriate curve from among the above, move the read basic characteristic curve on the coordinate system, and set a gradation conversion table used for gradation processing of the radiation image. .

  When the gradation conversion table is set based on the feature amount (reference signal value) of the image signal included in the desired image signal area, the gradation processing unit 29 (image processing means) obtains the digital radiation obtained by the recording / reading device 3. The image signal is converted by the gradation conversion table to perform gradation processing, and the gradation-processed image signal is output to the printer 17 via the interface 16 to visualize the radiation image.

  In the above-described embodiment, the configuration in which the conversion table for gradation processing is determined as the determination of the image processing condition based on the image signal included in the desired image signal area has been described. Alternatively, the dynamic range compression processing conditions may be determined based on the image signal included in the desired image signal region.

  Next, a more preferred embodiment will be described along the image processing described above according to the flowchart of FIG.

  In the flowchart of FIG. 7, first, in S <b> 1 (irradiation field identifying means), irradiation field recognition processing is executed in preparation for the case where the irradiation field stop is performed at the time of imaging.

  Next, in S2, a boundary signal S2 (here, S2 = ThresH) between the subject area and the blank area is determined. Such a determination is performed using the method using class separation based on the whole histogram described above (referred to as discriminant analysis method).

  In S3 (subject histogram separating means), a rough histogram (subject histogram) is created between the minimum value Smin of the image signal and the signal ThresH corresponding to the maximum value of the subject area. The coarse histogram is, for example, a histogram in which the interval width of the image signal is about 10.

  In S4, the peak portion of the frequency is detected by threshold processing based on the maximum frequency in the subject histogram (detection of a signal range having a frequency of 1/2 or more of the maximum frequency).

  In S5, it is determined whether or not there is a sharp peak in the predetermined low signal part. Here, when it is determined that there is a sharp peak in the low signal part, the process proceeds to S6, and the signal value (peak position) at the position where the frequency is peaked in the low signal part is half the peak frequency before and after the peak position. A value obtained by adding signal widths (half widths) indicating the above frequencies is set as a cutoff value Cutoff.

  On the other hand, when it is determined in S5 that there is no sharp peak in the predetermined low signal part, the process proceeds to S7, and the minimum value Smin of the image signal is set as the cut-off value Cutoff.

  The cut-off value Cutoff set in S6 or S7 defines a signal range for obtaining a desired image signal area, as will be described later. If the shooting conditions are inappropriate and the low signal is saturated, S6 By this processing, the saturated low signal is excluded from the signal range for obtaining the desired image signal region, and it is possible to avoid that the inadequate shooting condition adversely affects the determination of the desired image signal region.

  In S8, the minimum value ThresL of the signal region necessary for diagnosis is set by threshold processing between the cut-off value Cutoff and the signal ThresH corresponding to the maximum value of the subject region.

  Specifically, the sum of frequencies from the cut-off value Cutoff on the rough histogram to the signal ThresH corresponding to the maximum value of the subject area is obtained, and a value of 1/1000 of the sum is used as the frequency. A threshold is set (threshold determining means). Then, the frequency of each signal value and the threshold value are sequentially compared from the peak position closest to the signal value ThresH on the lower signal side than the signal value ThresH toward the lower signal side. A signal that is equal to or less than the threshold is searched, and the signal is set as the minimum value ThresL of the desired image signal region (see FIGS. 4 to 6). Thereby, the desired image signal area is determined as an area from the minimum value ThresL to the signal value ThresH (desired image signal area determining means).

  Next, in S9, a cumulative histogram is created between the signals ThresH and ThresL (desired image signal region) with the section width of the image signal set to 1.

  In S10, the signal value Sstd taking 10% of the generated cumulative histogram value is set as a reference density signal value, and the reference signal value Sstd is converted into a predetermined output signal value (output density or luminance). A tone conversion table is set (image processing condition determining means), and tone conversion of the radiation image is performed based on the tone conversion table (image processing means).

  In the above embodiment, gradation processing is described as an example of image processing. However, in addition to gradation processing, for example, the conditions of spatial frequency processing as disclosed in Japanese Patent Publication No. 62-62376 are as described above. It may be set based on the image signal in the desired image signal area determined in this way. That is, by setting an emphasis parameter (enhancement coefficient β in Japanese Patent Publication No. 62-62376) so that the emphasis degree becomes relatively large with respect to the image signal of the desired image signal area, a portion necessary for diagnosis can be obtained. It can be emphasized selectively for easy viewing.

  Further, the content of the dynamic range compression processing may be changed based on the determined image signal of the desired image signal region. For example, since the dynamic range information (feature amount) of the subject is obtained by determining the desired image signal area, the degree of compression of the dynamic range can be changed based on such information. Alternatively, the reference value Sx or Sy for determining the degree of correction of the dynamic range compression correction data in Japanese Patent Application Laid-Open No. 2-292679 may be determined based on the image signal in the desired image signal area.

  Further, the radiographic image signal that has been subjected to image processing (gradation processing, etc.) may be immediately hard-copied by the printer 17 as described above, but may be reproduced on the CRT simultaneously or independently, or It may be temporarily stored in the filing system, read out when necessary, hard copied, or displayed on the CRT.

  In the above embodiment, all steps S1 to S10 are executed using the radiographic image itself that is finally the object of observation interpretation. However, as disclosed in JP-A-58-67240, observation is performed. When a method of performing a “pre-reading” operation for determining a reading condition for “main reading” prior to a reading operation for obtaining a radiographic image signal for interpretation (referred to as “main reading”), for example, S1 to S1 The first half steps such as S2 or S1 to S7 are executed by using the “prefetch” signal, and the read gain of “main read” is determined based on the intermediate result to perform “main read”, and then the second half. This step may be executed using a “real reading” signal.

  In the above embodiment, S1 to S10 do not have to be executed by using the total amount of information of the radiation plug image signal to be subjected to observation interpretation. For example, an image signal reduced by pixel thinning processing may be used. From the viewpoint of improving processing speed and saving memory capacity, it is preferable. In that case, the effective pixel size of the reduced image is preferably 0.4 mm to 10.0 mm, and more preferably 1.0 mm to 6.0 mm.

  When storing a radiographic image in a filing system, a processed radiographic image signal that has been subjected to image processing (gradation processing, etc.) may be stored. A tone conversion table) and the like, and gradation processing may be performed at the time of reading.

  Further, in the present embodiment, the radiation image signal photoelectrically read from the photostimulable phosphor layer is configured to perform gradation processing, but is not limited to image reading using the photostimulable phosphor, Other two-dimensional radiation detectors or one-dimensional radiation detectors may be used. For example, a silver salt film on which a radiographic image is recorded is irradiated with light from a light source such as a laser or a fluorescent lamp to form a silver salt. A configuration may be employed in which the transmitted light of the film is obtained and a radiation image signal is obtained by photoelectrically converting the transmitted light.

  In other embodiments, the following may be used.

  After creating a histogram for the entire image, a subject histogram consisting only of the image signal of the subject area excluding the image signal corresponding to the radiation missing area is separated and created based on the histogram, and a predetermined ratio value of the sum of the frequencies in the subject histogram Is determined as a frequency threshold, and a histogram of the radiation image signal is created based on the subject histogram and the threshold, and the frequency threshold is determined from the histogram. Then, a desired image signal area (an image signal area corresponding to the desired image) is determined based on a comparison between the frequency on the histogram and the threshold value, and based on the feature amount of the image signal included in the desired image signal area. An image processing condition suitable for visualizing the desired image may be determined.

  In addition, when a radiographic image is composed of a subject region and a radiation missing region, first, a histogram for the entire image is created, and then the subject region excluding the image signal corresponding to the radiation missing region based on the histogram is created. A histogram consisting only of image signals is separated and created, and the threshold value and the desired image signal area are determined based on the histogram corresponding to the subject area, so that the image signal of the radiation missing area is the threshold value and the desired image signal area. It may be so arranged that it does not affect the determination.

  Further, the sum of the frequencies of the histogram is calculated, a frequency threshold for determining the frequency on the subject histogram is determined based on the sum, a desired image signal region is determined based on the threshold, and the frequency on the subject histogram is determined. It is also possible to determine at least one of the maximum value and the minimum value of the desired image signal region by comparing the size of the desired image signal region with the threshold value, and to specify the desired image signal region according to the determination.

  Further, the frequency on the subject histogram is sequentially compared with the threshold value from the predetermined start signal value toward the low signal side on the subject histogram, and the signal value whose frequency is less than the threshold value is first set as the minimum value of the desired image signal region. On the other hand, similarly, the frequency on the subject histogram is sequentially compared with the threshold value from the predetermined start signal value toward the high signal side on the subject histogram, and the signal value with the frequency less than the threshold value is first set in the desired image signal region. The maximum value may be set, or the maximum value of the image signal corresponding to the subject area may be set as the maximum value of the desired image signal area, and the signal area including the predetermined start signal value may be set as the desired image signal area. .

  Further, the predetermined start signal value is set to a peak position having a frequency closest to the maximum value of the image signal corresponding to the subject area or a predetermined ratio position of the image signal width corresponding to the subject area, and the image signal corresponding to the subject area The desired image signal area is determined in step 1, and when the radiation field is reduced and radiation imaging is performed, the irradiation field area is identified and a histogram is created using the image signal in the irradiation field area. By doing so, the determination of the threshold value and the desired image signal area may be avoided from being influenced by the image signal of the area where the irradiation field stop has been performed.

  Further, a gradation conversion table may be determined based on the feature amount of the image signal included in the determined desired image signal region so that the radiation image is visualized with an appropriate gradation corresponding to the feature amount. In a configuration for determining a gradation conversion table, a basic characteristic curve indicating characteristics of gradation conversion is stored in a coordinate system for gradation conversion composed of input signal values and output signal values, and obtained from a subject histogram. The basic characteristic curve is deformed so that the reference signal value of the desired image signal area is converted into a predetermined output signal value, and an appropriate gradation conversion table (that is, the input signal value in the gradation conversion process) A lookup table indicating individual correspondence with output signal values may be set.

  Further, the radiographic image signal is image-processed according to the image processing conditions determined according to the desired image signal region determined based on the subject histogram as described above, so that a radiographic image signal suitable for visualization of the desired image is obtained. Also good.

  By taking such a configuration, the following effects can be obtained.

  The desired image signal area can be determined by excluding the signal of the radiation missing area, and the desired image signal area can be determined without being affected by the frequency fluctuation in the desired image signal area. Image processing conditions can be determined, and the boundary of the desired image signal area is detected based on a comparison between the threshold value and the frequency on the subject histogram. There is an effect that the signal region can be determined.

  In addition, since the comparison of the frequency is sequentially performed with the relatively high signal side of the desired image signal region as the start signal value toward the low signal side, unnecessary image portions on the low signal side can be surely excluded and the desired image signal region There is an effect that can be determined.

  In addition, even for radiation images that have been subjected to irradiation field reduction, a histogram is created using image signals in the irradiation field, and the desired image signal area is inappropriately affected by the image signal in the irradiation field restriction area. There is an effect that it is possible to avoid being determined.

  Further, there is an effect that a conversion table for gradation conversion is determined as the image processing condition, and gradation processing suitable for visualization of a desired image can be performed.

  In addition, since the basic characteristics of gradation conversion stored in advance are deformed based on the feature amount of the image signal in the desired image signal area, the gradation conversion condition corresponding to the desired image signal area is determined. There is an effect that it is possible to easily determine the optimum gradation processing condition for the desired image.

  In addition, image processing is performed under an image processing condition that is optimal for a desired image, and particularly in a radiographic image for medical use, there is an effect that a site necessary for diagnosis can be visualized more easily.

FIG. 2 is a block diagram showing a basic configuration of an apparatus according to the invention of claim 1. 1 is an overall system schematic diagram showing an embodiment of the present invention. The block diagram which shows the structure of the image process part in embodiment. The diagram which shows the mode of the minimum value ThresL determination of the desired image signal area | region. The diagram which shows the mode of the minimum value ThresL determination of the desired image signal area | region. The diagram which shows the mode of the minimum value ThresL determination of the desired image signal area | region. The flowchart which shows more preferable embodiment of a gradation process. The diagram for demonstrating the problem in a conventional apparatus. The diagram for demonstrating the problem in a conventional apparatus. The diagram for demonstrating the problem in a conventional apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Radiation generation source 3 Recording / reading apparatus 4 Conversion panel 5 Light beam generation part 10 Photomultiplier 14 Image processing apparatus 15 CPU
17 Printer 21 Irradiation Field Identification Unit 22 Histogram Creation Unit 23 Subject Histogram Separation Unit 24 Frequency Threshold Determination Unit 25 Desired Image Signal Area Determination Unit 26 Reference Signal Value Determination Unit 27 Gradation Processing Condition Determination Unit 28 Basic Gradation Characteristic Storage Unit 29 Floor Adjustment processing

Claims (4)

A histogram creating means for creating a histogram indicating a frequency for each predetermined section width of the image signal of the radiation image formed corresponding to the amount of radiation transmitted through the subject;
Subject histogram separation means for separating and creating a subject histogram corresponding to the subject area from the histogram created by the histogram creation means;
Threshold value determining means for obtaining a sum of frequencies in the subject histogram separated and created by the subject histogram separating means, and determining a predetermined ratio value of the sum as a frequency threshold value;
A desired image signal region is determined based on a comparison between a frequency for each predetermined section width of the image signal in the subject histogram separated and created by the subject histogram separation unit and a frequency threshold value determined by the threshold value determination unit. A desired image signal region determining means;
Image processing condition determining means for determining an image processing condition for the radiation image based on a feature amount of an image signal included in the desired image signal area determined by the desired image signal area determining means;
An image processing apparatus comprising: an image processing unit configured to perform image processing on the radiation image based on the image processing condition determined by the image processing condition determining unit. The desired image signal region determining means compares the frequency for each predetermined section width of the image signal in the subject histogram separated and created by the subject histogram separating means with the frequency threshold determined by the threshold determining means. The image processing apparatus according to claim 1, wherein at least one of a maximum value and a minimum value of the desired image signal area is determined based on the determination result. The histogram creating means includes an irradiation field identifying means for identifying a radiation field area based on an image signal of the radiation image, and using the image signal in the radiation field identified by the irradiation field identifying means. The image processing apparatus according to claim 1, wherein a histogram is created. A histogram creating step for creating a histogram indicating a frequency for each predetermined section width of an image signal of a radiographic image formed corresponding to the amount of radiation transmitted through the subject;
A subject histogram separation step of separating and creating a subject histogram corresponding to a subject region from the histogram created in the histogram creation step;
A threshold determination step of calculating a sum of frequencies in the subject histogram separated and created in the subject histogram separation step, and determining a predetermined ratio value of the sum as a frequency threshold;
A desired image signal region is determined based on a comparison between the frequency for each predetermined section width of the image signal in the subject histogram separated and created in the subject histogram separation step and the frequency threshold value determined in the threshold value determination step. A desired image signal region determination step;
An image processing step of performing image processing on the radiographic image under an image processing condition determined based on a feature amount of an image signal included in the desired image signal region determined in the desired image signal region determining step. Image processing method.

Images