JP2013103023A - Image diagnostic support device and image diagnostic support method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enhance the capability to deal with image diagnostic while securing the eligibility and efficiency of diagnosis.SOLUTION: The image diagnostic support device for supporting a diagnosis using a medical image acquired from a subject includes: a weighted image generation means S2 which generates a weighted image with an abnormal shadow candidate region being weighted by performing processing on an original image; a composite image forming means S3 which forms a composite image in which a tissue shadow of the original image and the abnormal shadow candidate region of the weighted image are displayed by superimposing the original image with the abnormal shadow candidate region of the weighted image; and an image display means S4 which displays the composite image. When a composition parameter change means S5 which changes a composition parameter of the composite image forming means S3 is further provided, reading operation can be further facilitated since a composition mode can be adjusted.

Description

  The present invention relates to an image diagnosis support apparatus and an image diagnosis support method, and more particularly, to a support technology suitable for performing diagnosis by reading a captured image in mammography (mammography) examination.

  In general, with the development of medical examination equipment, in recent years various diagnostic images, for example, X-ray images taken with a CR device or CT device, NMR images taken with an MRI device, etc. are read and diagnosed. Image diagnosis to be performed is frequently performed. In such image diagnosis, it is difficult to secure the interpretation doctor because it is greatly affected by the proficiency of the interpretation technique of the doctor, etc., especially when a large amount of image diagnosis is required, such as examination data May increase the burden on the interpreting doctor and may lead to misdiagnosis due to fatigue or carelessness. Therefore, various devices for supporting image diagnosis, for example, computer-aided diagnosis devices such as CAD (Computer-Aided Diagnosis or Detection) have been developed. In this type of support device, a device that automatically detects a lesion site (for example, an abnormal shadow candidate region such as a tumor shadow or a microcalcification region, which will be described later) from an image, and identifies and displays the position is realized. Yes. There is also known a method for performing category determination corresponding to the degree of risk according to the nature of the lesion site.

  Conventionally, regarding a specific image display mode, various methods for facilitating visual diagnosis of an image by detecting an abnormal shadow candidate region by a computer and emphasizing it in some display mode have been proposed. For example, in the following Patent Document 1, an abnormal shadow candidate such as a mass shadow or a microcalcification cluster is detected by analysis of image data, and an emphasis process for emphasizing the abnormal shadow candidate area, or an abnormal shadow candidate area is deleted / There is disclosed a method of performing deletion / attenuation processing to attenuate and displaying a plurality of types of processed images together with an original image in some cases. Moreover, in patent document 2, the position of the element which may comprise the signs of micro calcification is calculated | required, The method which can grasp | ascertain the distribution of micro calcification easily by displaying the image which emphasized these elements Is disclosed. Further, in Patent Document 3, a first local region including a calcification cluster in which candidate points for micro calcification shadows are continuous is designated and enlarged, and one or more candidates are displayed in the displayed first local region. A method of enlarging and displaying a second local region including a point is disclosed. Further, in Patent Document 4, a calcified pixel of an original image is specified and its position information is extracted, a new pixel is calculated from the original image according to a reduction ratio, and it is determined that there is a calcified pixel within the new pixel range. Then, the pixel value of the new pixel is calculated based on the pixel value of the calcified pixel included in the new pixel range, and a reduced image is generated from the pixel value of each new pixel, thereby detecting the detected lesion site. A method for making information visible even in a reduced image is disclosed.

JP 2005-102784 A Japanese Patent Laying-Open No. 2005-169122 JP 2007-313196 A JP 2009-136457 A

  However, although various conventional methods as described above have been proposed, problems such as a rapid increase in examination images and a shortage of interpretation doctors in recent years have not been solved. This is because, in any method, complicated detection processing is required to detect abnormal shadow sites and lesion sites and specify their positions, and as a result, the cost of installing a support system increases. This is because it is practically difficult to secure a large number of support devices.

  Therefore, the present invention solves the above-described problems, and the problem is to realize a configuration that can easily install and introduce an image diagnosis support apparatus or support method having good support characteristics for image diagnosis, The aim is to enhance the ability of diagnostic imaging while ensuring eligibility and efficiency.

  In view of such circumstances, an image diagnosis support apparatus according to the present invention is an image diagnosis support apparatus for supporting diagnosis using a medical image acquired from a subject, and performs processing on an original image of the medical image. The enhanced image generating means for generating an enhanced image in which the abnormal shadow candidate region is enhanced by the superimposing image, and the original image and the enhanced image are superimposed on each other to cause both the tissue shadow of the original image and the abnormal shadow candidate region of the enhanced image. A composite image forming unit that forms a composite image to be displayed and an image display unit that displays the composite image are provided. Here, a medical image is an image used for diagnosis, which is an image acquired using all electromagnetic waves (including light) and radiation in a broad sense that means particle beams, and acoustic waves such as ultrasonic waves. The image acquired using is included.

  In the present invention, it is preferable that the emphasized image is one in which the abnormal shadow candidate region is represented as an island-like region in a substantially uniform background. This facilitates the formation of a composite image, and facilitates the understanding of the visibility and positional relationship between abnormal shadow candidate regions and tissue shadows in the formed composite image.

  In the present invention, it is preferable that the composite image forming unit forms the composite image by adding the original image and the emphasized image. According to this, a composite image can be formed by a relatively simple calculation process regardless of the image mode. Here, the addition processing includes arithmetic processing such as simple addition, weighted addition, simple average, and weighted average of pixel values of corresponding pixels of the original image and the emphasized image.

  In the present invention, it is preferable that the composite image forming unit forms the composite image by making a part of the emphasized image other than the abnormal shadow candidate region transparent or semi-transparent and overlaying the original image. . According to this, although it is necessary to distinguish and process the partial area in the enhanced image, the display mode of the tissue shadow in the range corresponding to the partial area in the original image and the display of the abnormal shadow candidate area of the enhanced image Since the mode can be set independently of each other, the visibility when grasping the mode and positioning of the abnormal shadow in the tissue can be improved. In this case, the transparency (pixel value) of a partial region of the emphasized image is arbitrary, but it is preferable to perform complete transparency. Also in this case, the weighting of the original image and the emphasized image can be changed. Furthermore, it is desirable that the partial area is an area serving as a background of the abnormal shadow candidate area or all areas other than the abnormal shadow candidate area.

  In the present invention, it is preferable that the image display means is configured to be able to switch and display the original image and the composite image or to display them in parallel. As a result, the comparison between the original image and the synthesized image is facilitated, and the interpretation work based on the original image can be performed more easily.

  In the present invention, it further includes a synthesis parameter changing unit that changes a synthesis parameter related to a superposition mode of the original image and the emphasized image by the synthesized image forming unit in response to a specific operation input or automatically. Is preferred. In this case, it is preferable that the composite image forming unit forms the composite image with a new composite parameter every time the composite parameter is changed. Since the visibility of the composite image changes by changing the composite parameter, it is possible to further facilitate the positioning of the abnormal shadow candidate region in the tissue shadow and the reading of the surrounding tissue shadow. In particular, for abnormal shadow candidate areas that are difficult to visually recognize, such as microcalcification areas, the degree of emphasis of abnormal shadow candidate areas and the visibility of surrounding tissue shadows can be changed by changing the composite parameters by changing the composite parameters. Therefore, it is easy to prevent oversight in image diagnosis. Also, when changing the composition ratio corresponding to the contribution ratio between the original image and the emphasized image in the composition parameters, the ratio of the display intensity of the tissue shadow and the abnormal shadow candidate area can be adjusted. Both oversight prevention and confirmation of tissue shadow can be facilitated.

  In the present invention, it is preferable to further include an abnormal area enlarging unit that expands the abnormal shadow candidate area by performing an expansion process on the abnormal shadow candidate area. This makes it easier to visually recognize the abnormal shadow candidate region on the composite image and avoids the disappearance of the abnormal shadow candidate region when the composite image is displayed in a reduced size.

  Next, an image diagnosis support method according to the present invention is an image diagnosis support method for supporting diagnosis using a medical image acquired from a subject, and an abnormal shadow candidate region is emphasized by processing an original image. An enhanced image generating step in which the enhanced image is generated, and a composite image in which the tissue shadow of the original image and the abnormal shadow candidate region of the enhanced image are displayed together by superimposing the original image and the enhanced image A composite image forming step in which the composite image is formed, and an image display step in which the composite image is displayed.

  In the present invention, it is preferable that the enhanced image is one in which the abnormal shadow candidate region is represented as an island region in a substantially uniform background.

  In the present invention, it is preferable that in the composite image forming step, the composite image is formed by adding the original image and the emphasized image.

  Furthermore, in the composite image forming step, it is preferable that the composite image is formed by making a part of the emphasized image other than the abnormal shadow candidate region transparent or semi-transparent and overlaying the original image.

  In the image display step, it is preferable that the original image and the composite image can be switched and displayed or can be displayed in parallel.

  The present invention further includes a synthesis parameter changing step in which a synthesis parameter relating to a superposition mode of the original image and the emphasized image in the synthesized image forming step is changed according to a specific operation input or automatically. In the composite image forming step, it is preferable that the composite image is formed with a new composite parameter every time the composite parameter is changed.

  In the present invention, it is preferable that the method further includes an abnormal region expansion step in which an expansion process is performed on the abnormal shadow candidate region to expand the abnormal shadow candidate region.

  According to the present invention, by superimposing the original image and the emphasized image, it is possible to easily provide a composite image in which both the tissue shadow of the original image and the abnormal shadow candidate region of the emphasized image are displayed. Can be installed easily, and the positional relationship between the tissue and abnormal shadow candidate areas can be grasped quickly and clearly at the time of interpretation, enabling mass processing while ensuring the accuracy and efficiency of image diagnosis. It is possible to achieve an excellent effect of.

It is a schematic block diagram which shows the main structures and operations of the diagnostic imaging support apparatus and support method according to the embodiment of the present invention. It is a schematic external view showing an external appearance example of the system configuration of the embodiment. It is a schematic block diagram which shows the function structural example of the local computer of the embodiment. It is explanatory drawing (a)-(c) which shows the formation method of the synthesized image of the embodiment typically. It is explanatory drawing (a)-(c) which shows typically the change of the display mode by the synthesis rate adjustment of the synthesized image of the embodiment. It is explanatory drawing (a)-(e) which shows typically the formation method of the reduction image of the embodiment. It is a schematic flowchart which shows the example of the process sequence of the system configuration | structure of the embodiment. It is a screen display figure which shows the example of the basic setting screen of the embodiment. It is a screen display figure which shows the example of the display setting screen of the embodiment. It is a screen display figure which shows the example of the main display screen of the embodiment. It is a photograph which shows the example of the actual original image (MLO-L) of the embodiment. It is a photograph which shows the example of the actual original image (MLO-R) of the embodiment. It is a photograph which shows the example of the actual original image (CC-L) of the embodiment. It is a photograph which shows the example of the actual original image (CC-R) of the embodiment. It is a photograph which shows the example of the actual emphasis image (MLO-L) of the embodiment. It is a photograph which shows the example of the actual original image (MLO-R) of the embodiment. It is a photograph which shows the example of the actual original image (CC-L) of the embodiment. It is a photograph which shows the example of the actual original image (CC-R) of the embodiment. It is a photograph which shows the example of the actual synthesized image (MLO-L) of the embodiment. It is a photograph which shows the example of the actual original image (MLO-R) of the embodiment. It is a photograph which shows the example of the actual original image (CC-L) of the embodiment. It is a photograph which shows the example of the actual original image (CC-R) of the embodiment.

  Next, embodiments of the present invention will be described in detail with reference to the accompanying drawings. FIG. 1 is a schematic configuration diagram showing the relationship of the function realizing means and the processing procedure showing the main configuration of the image diagnosis support apparatus and image diagnosis support method of this embodiment.

  The original image capture / display function (or step) S1 is a function (or step) that captures image data of the original image and displays the original image by display means such as a display. The original image refers to an X-ray image, an ultrasonic image, an MRI image, or the like obtained by photographing at least a part of the body of a subject (a subject of an acquired image, for example, a person who has been examined). In this function (step) S1, original image data, which is typically in the form of digital data, is input from outside the apparatus or read out from recording means in the apparatus, and an appropriate value is determined according to the original image data. The display means displays the original image. This function (step) S1 may constitute part of the image display means (image display step) in that it displays an original image.

  Next, the enhanced image generation function (or step) S2 based on the original image performs various filter processes and conversion processes on the original image. These processes may include various processes known as image processing filters, including smoothing filters such as a moving average filter and a Gaussian filter, as long as an abnormal shadow candidate region can be emphasized. Particularly, the binarization filter, the median filter, the top hat transform, the variable ring filter (triple ring filter), the wavelet transform, the mask are suitable for the present embodiment aiming to emphasize the microcalcification region. Processing. The binarization filter includes various types of binarization using a threshold, for example, processing using a P tile method or a discriminant analysis method. This function (or step) S2 constitutes the above-described enhanced image generation means (or enhanced image generation step).

  In the case of the present embodiment, the above-mentioned emphasized image is an emphasis on an abnormal shadow candidate region that is suspected to be a tumor shadow or a microcalcification region in the case of tumor diagnosis. Various types of abnormal sites such as cysts and fibrous tissue are emphasized with respect to the original image. Here, it is more preferable that the body tissue shadow as a background other than the abnormal shadow candidate region is weakened. In particular, it is desirable that tissue shadows other than the abnormal shadow candidate regions are completely erased and only the abnormal shadow candidate regions are left in an island shape. In general, the above-described image processing removes noise components unrelated to the abnormal shadow candidate area, and performs an emphasis process to distinguish the abnormal shadow candidate area from the tissue shadow that is the background. Such a process is desirably performed by appropriately combining a plurality of processes among the above-described suitable filter processes or conversion processes. Preferred combinations in this case include, for example, a combination in which the median filter and the binarization filter are executed in this order, and a combination in which the top hat transform and the wavelet transform are executed in this order. In particular, by performing binarization processing in the final stage, the abnormal shadow candidate region and the background can be clearly distinguished and displayed on the image. In particular, image processing for emphasizing the above-described microcalcification region is well known, and when such processing is performed, the tissue shadow is removed and a bright spot distribution corresponding to the microcalcification region is obtained.

  As an example of the enhanced image, for example, an abnormal shadow candidate area is expressed with high brightness, and a background other than the abnormal shadow candidate area is expressed with low brightness. In this case, it is preferable that the brightness difference between the abnormal shadow candidate region and the other background is large, and the brightness difference is preferably 30% or more, particularly 50% or more of the gradation width of the image. Further, the background may be made uniform using mask processing. Furthermore, it is good also as an image which distinguished the said abnormal shadow candidate area | region and the said background by binary by the binarization process.

  Next, in the present embodiment, in the superimposing function (or step) S3 of the original image and the enhanced image, the original image and the enhanced image are superimposed to form a composite image. Here, the composite image basically only needs to display both the tissue shadow of the original image and the abnormal shadow candidate region of the enhanced image. The superimposing process of the original image and the emphasized image can be performed by various methods that satisfy the basic condition as a result of the above-described composite image. This function (or step) S3 constitutes the above-described composite image forming means (or composite image forming step).

  The above-described superimposing process of the original image and the emphasized image may be, for example, an image obtained by superimposing the entire original image and the entire emphasized image in some appropriate manner (addition mode described later), Further, a partial area other than the abnormal shadow candidate area of the enhanced image, for example, the background described above as a transparent or semi-transparent state and the original image superimposed thereon (transmission mode described later) may be used. The former means that the pixel value (for example, lightness) of the original image and the pixel value (for example, lightness) of the enhanced image are added in a manner that is unified by some standard. Thereby, since processing can be performed regardless of the image mode, image composition can be executed by a relatively simple arithmetic processing. On the other hand, the latter means that the pixel value (for example, brightness) of the partial area (background, etc.) is reduced or added to the pixel value (for example, brightness) of the original image as 0. Thus, although it is necessary to distinguish and process the partial area (background, etc.), the abnormal shadow candidate area and the partial area can be set independently of each other, so the abnormal shadow candidate area and the surrounding tissue shadow can be set. The visibility of each can be improved. In any case, the ratio of the contribution degree of the original image and the emphasized image (overlapping weighting) to the display of the composite image can be set to an appropriate ratio, and this ratio is expressed by a composite ratio described later. Here, when the ratio is fixed, it is preferable that the original image and the emphasized image be 50% each.

  In the present embodiment, the description has been made on the assumption that the image to be handled is basically a grayscale image having a predetermined gradation value for each pixel, and the pixel value = lightness value as described above. Can be dealt with by treating the brightness of the pixel value in the same manner as described above. Further, by setting a predetermined color scale in the color image and using the hue instead of the gradation, it is possible to deal with the case by handling the hue of the pixel value in the same manner as the above gradation. For example, replace the hue with a numerical value, and form a color scale in the order of red, yellow, green, blue, purple from the larger pixel value, and handle these numerical values in the same way as the above brightness, depending on the hue Image composition can also be performed. Furthermore, as an image of the present invention, in a binarized image in which each pixel takes a black and white binary value, an image accompanied by a pseudo gradation expression such as a random dither method or an error diffusion method is used instead of binarization by a threshold value. An aspect may be sufficient.

  The composite image formed by the superimposition function (or step) S3 of the original image and the emphasized image is displayed by display means such as a display monitor (display) in the composite image display function (or step) S4. The display mode by the display unit may be at least a mode in which both the tissue shadow of the original image and the abnormal shadow candidate region of the emphasized image can be visually recognized in the composite image. Therefore, as long as such a mode is used, even if the composite image to be displayed is a gray scale (monochrome gradation display) or a color display, it is a binarized image with pseudo gradation expression. It doesn't matter. This function (or step) S4 constitutes the image display means (or image display step).

  It is desirable that the image display means of this embodiment can display both the original image and the composite image. This function may display the original image and the composite image simultaneously in parallel, or may selectively display the original image and the composite image by switching. In the latter case, it is desirable that the image is switched in accordance with the operation of the viewer (for example, an interpreter), but the original image and the composite image may be switched automatically. For example, you may make it display an original image and a synthesized image alternately with a fixed period.

  In the present embodiment, as a more preferable embodiment, a synthesis parameter changing function (or step) S5 is provided. In this case, the synthesis parameter is used in the above-described superposition (image synthesis) processing of the original image and the emphasized image, and is a variable relating to the superposition mode of the image when forming the composite image, and preferably the composite image It depends on the contributions of the original image and the emphasized image to the display (for example, the composition ratio which is the contribution ratio). Here, for example, when the addition mode is selected as the composition mode, the original image and the emphasized image are appropriately weighted and added to form a composite image, and the composition ratio is set according to the numerical value representing the weight at the time of the addition Is done. In the above-described example, when the original image and the emphasized image are combined with a contribution of 50%, the combination rate is 50% (or 0.5).

  In the above addition mode, when the weighted average of the pixel value of the original image and the pixel value of the enhanced image with an arbitrary weighting is used as the pixel value of the composite image, the pixel value Ps of the composite image is the pixel value of the original image. Using Po and the weight coefficient Wo, and the pixel value Pe and the weight coefficient We of the enhanced image, the following equation 1 is obtained.

  Further, the synthesis rate S at this time is given by the following equation 2, for example. In this case, when the synthesis rate S is 0, it matches the original image, and when the synthesis rate S is 1, it matches the enhanced image. Therefore, as the composition rate S increases, the contribution (influence) of the original image to the display of the composite image decreases, and the contribution (influence) of the emphasized image increases.

  Although in the addition mode as described above, it is also possible to simply add the original image and the emphasized image. In this case, the pixel value Ps of the composite image can be obtained from the pixel value Po of the original image and the pixel value Pe of the emphasized image by the following equation (3). Here, Mo is the contribution of the original image, Me is the contribution of the enhanced image, and Pmax is the maximum pixel value. That is, when the pixel value Ps of the synthesized image obtained by the upper equation is equal to or less than Pmax, the pixel value is used as it is, and when the pixel value Ps of the synthesized image obtained by the upper equation is equal to or larger than Pmax. , The pixel value Ps is set to Pmax as shown in the lower equation.

  In this case, although not particularly limited, if Mo = Me = 1, the pixel value Ps of the original image and the pixel value Pe of the enhanced image are within the range where the pixel value Ps of the composite image does not exceed the maximum value Pmax. The simple addition value of If Mo + Me = 1, the pixel value Ps is a weighted average by appropriate weighting of the pixel values Po and Pe in a range not exceeding the maximum value Pmax. In this case, the synthesis rate S is expressed by, for example, the following formula 4, and when the synthesis rate S is 0, it matches the original image, and when the synthesis rate is 1, it matches the enhanced image. Therefore, as the composition rate S increases (closer to 1), the contribution (influence) of the original image in the composite image decreases and the contribution (influence) of the enhanced image increases. However, if Mo + Me is set to be larger than 1, the composite sensitivity is increased by that amount, and a composite image in which the tissue is emphasized is obtained. On the other hand, the original pixel values of the original image and the emphasized image are the same. In a large region, the maximum value Pmax is likely to be reached, and gradation saturation (so-called white stripes or black blur) is likely to occur.

  The above synthesis rate can be defined even when other overlay methods are employed. For example, when the transmission mode is selected as the composition mode, the background is made transparent in the emphasized image, the emphasized image is displayed as it is in the abnormal shadow candidate region, and the original image is overlaid on the original image. The image is displayed as it is. Even in this case, if the intensity ratio between the portion of the original image displayed in the background and the portion of the emphasized image displayed in the abnormal shadow candidate region is the composition ratio S, the composite image is obtained at an appropriate composition ratio S. Can be formed.

  The function (or step) S5 described above constitutes the synthesis parameter changing means (or the synthesis parameter changing step). This means or process may be realized by changing the synthesis parameter such as the synthesis rate S according to the operation of the image interpreter as described above. However, the synthesis parameter such as the synthesis rate S is automatically changed over time. It may be realized by changing. For example, over time, the composition rate S is gradually increased from 0 to 1, conversely, the composition rate S is gradually decreased from 1 to 0, or the composition rate S is increased or decreased once each time. It may be repeated several times. Of course, only one of the increase process and the decrease process may be repeated a plurality of times.

  Since the above synthesis parameter only needs to relate to the superposition mode, in addition to the synthesis rate S indicating the contribution ratio of the original image and the emphasized image to the display of the composite image, it varies depending on the contribution ratio. Various other variables can also be used. Also, whether the synthesis mode is the addition mode or the transmission mode, the type of the addition operation expression in the addition mode (not limited to the above formulas 1 and 3, but other formulas can be used), and the transmission mode. The background transparency in the mode can also be used as a synthesis parameter.

  In the present embodiment, image diagnosis and recording work (step) S6 are performed by an interpreting doctor or the like. This operation (step) S6 can be performed by displaying and observing only the composite image, but generally it is preferably performed by using the original image and the composite image together. When performing image diagnosis using both images in combination, the above-mentioned image display means displays the original image and the synthesized image simultaneously in parallel, or synthesizes the original image automatically or in response to an operation by a radiographer or the like. Switch images to display. When both images are used together, it is desirable for accurate diagnosis to check the position of the abnormal shadow candidate region in the tissue shadow with the composite image and then read the shadow of the position with the original image.

  FIG. 2 is a schematic external view schematically showing an external appearance example of the configuration of the image diagnosis support apparatus of the present embodiment. In the case of the illustrated example, a computer for image diagnosis is arranged at the work place of the interpreter, the computer main body 1, the input means 2 such as the keyboard 2a and the mouse 2b connected to the computer main body 1, and various monitors (displays). The display unit 3 includes a diagnostic recording display unit 3a and image display units 3b and 3c. The computer main body 1 includes an information communication unit 1a for exchanging information with the outside.

  In addition to the image diagnostic computer, the image data 5A of the original image and the image information 5B corresponding to the image data 5A are stored in association with each other, and the diagnostic record 5C is stored with the image data 5A and the image information 5B. A server system 4 is provided for configuring the examination database 5D formed by recording in association with each other. The server system 4 includes an information communication unit 4a for exchanging information with the outside, an information recording unit 4b for storing the image data 5A, image information 5B, and diagnostic record 5C, and the computer via the information communication unit 4a. A server control unit 4c that performs a server function for an external local computer such as the main body 1 and manages data exchange between the local computer and the information recording unit 4b, for example, is provided.

  Between the information communication unit 1a of the computer main body 1 and the information communication unit 4a of the server system 4 via the dedicated line, the internal network, the Internet, etc. by wired communication, wireless communication, optical communication method, etc. Transfer of image data 5A, image information 5B, and diagnostic record 5C is performed. When the computer main body 1 is operated, the server system 4 can be accessed, and one or a plurality of sets of image data 5A and image information 5B that are associated with each other according to the operation content (for example, multiple sets of information included in the examination database). You can select and call. In addition, the diagnostic record 5C input to the computer main body 1 can be recorded in the information recording unit 4b of the server system 4 in association with the image data 5A and the image information 5B, and a part of the examination database 5D can be reconfigured. The illustrated example is only an example of the system configuration, and various configurations can be arbitrarily formed. For example, at least a part of data management processing or image processing, specifically, setting processing of an image incorporation format into data, or (loading) image processing such as triple ring filter, wavelet transform, or the like You may perform by an external server etc. (for example, cloud computing).

  The image data 5 </ b> A and the image information 5 </ b> B called up as described above are displayed by the display unit 3 in a state that can be visually recognized by a radiogram interpreter. Here, the image information 5B includes subject name, age and other subject information, photographing date and time, photographing location, photographing device and other image acquisition information, image format, gradation information, image size and other image content information, and other information. Information (for example, various information defined in the DICOM standard of the medical image system) can be mentioned.

  The image information 5B is displayed on the display unit 3 by the diagnostic recording display unit 3a. The diagnostic record display means 3a is used to access and read the examination database 5D stored in the information recording unit 4b of the server system 4 and to store (additional) the diagnostic result 5C in the information recording unit 4b. The operation screen is also displayed. When the specific image data 5A and image information 5B are read from the examination database 5D, the specific image information 5B is displayed on the diagnostic recording display means 3a, and the corresponding image data 5A is displayed on the display means for image display. 3b and 3c.

  The illustrated example shows a state in which mammography examination data is displayed. The image data 5A includes, for example, left and right internal and external obliquely captured images MLO-L and MLO-R (MLO; Medio-lateral Oblique), and left and right head-to-tail captured images CC-L and CC-R (CC; Cranio-caudal) is included, and these are displayed on the image display display means 3b and 3c. In the display mode shown in FIG. 2, left and right inside / outside obliquely captured images MLO-L and MLO-R among the four images are displayed on one image display display means 3b, and left and right head and tail are displayed. The direction-captured images CC-L and CC-R are displayed on the other image display display means 3c. However, the present invention is not limited to such a display mode. Images may be displayed, or only one or two images may be displayed, and other images may be displayed by switching the display. Further, the image display display means may be prepared for the number of images for a specific subject, and each image may be displayed on each display means one by one. Further, the type and number of images included in the image data 5A are not particularly limited to the above example, and other types of images and other numbers of images may be used.

  A diagnostic recording area for recording the diagnostic result 5C is prepared on the display screen of the diagnostic recording display means 3a on which the image information 5B is displayed, and the diagnostic recording 5C can be input by designating the diagnostic recording area. ing. In addition, a navigation area 3n may be prepared in a part of the display screen. The navigation area 3n displays the original image, the emphasized image, or the composite image based on the image data 5A. Each image is reduced and displayed in accordance with the navigation area 3n. In the example shown in the figure, two images are displayed simultaneously in parallel, but a single image may be displayed, or three or more images may be displayed in parallel. It is not something. When a composite image is displayed in the navigation area 3n, it is preferable to display a reduced image process (application of an abnormal area enlargement unit or a process in an abnormal area enlargement process) described later.

  FIG. 3 is a schematic configuration diagram schematically showing the configuration of the image diagnostic computer of the same configuration example. The computer main body 1 includes a CPU (Central Processing Unit) 10 which is an arithmetic processing means, a system bus 11 connected to the CPU 10, and a ROM (Read Only) connected to the system bus 11 via appropriate input / output means. A memory 12 including a memory, a memory 13 including a random access memory (RAM), and an image generating unit 14 including an image processing circuit configured on a main board or a graphic board. Further, if necessary, recording storage means 15 and 16 for locally storing image data, diagnostic data, and the like are provided. Further, the information communication unit 1a, the input unit 2, and the display unit 3 are connected.

  The memory 12 stores an image diagnosis support program for performing the overall processing of the present embodiment. The image diagnosis support program is read out and executed by arithmetic processing by the CPU 10 and the like. Functions (steps) S1 to S5 shown and a series of processes shown in FIG. The image diagnosis support program includes an image processing program, and the CPU 10 and the image generation unit 14 operate according to the image processing program to generate an emphasized image and a composite image.

  Next, in the present embodiment having the above-described configuration, the overall effect based on the functions of the emphasized image generating unit, the synthesized image forming unit, and the synthesized parameter changing unit illustrated in FIG. 1 will be described. FIG. 4 is an explanatory diagram schematically showing an example of (a) an original image, (b) an emphasized image, and (c) a composite image. In the original image shown in FIG. 4A, the abnormal shadow candidate AS is displayed in a state included in the tissue shadow TS. On the other hand, in the emphasized image shown in FIG. 4B, the abnormal shadow candidate AS is displayed as an abnormal shadow candidate AS ′ in an enhanced form, but the display intensity of the tissue shadow TS other than the abnormal shadow candidate AS is weakened. Yes. In the illustrated example, the tissue shadow TS of the original image is substantially erased in the enhanced image, and the abnormal shadow candidate AS ′ is represented as an island-shaped region in a uniform background. Here, the formation method of the abnormal shadow candidate AS ′ in the enhanced image by the enhancement process of the abnormal shadow candidate AS in the original image is the difference between the pixel value of the abnormal shadow candidate AS and the pixel value (lightness) of the other background portion. The pixel value (brightness) of the abnormal shadow candidate AS may be changed so as to increase, or the region where the abnormal shadow candidate AS exists in the original image is expanded to expand the abnormal shadow in the enhanced image. The area where the candidate AS ′ exists may be expanded, or both may be performed together.

  In the composite image formed as described above, the abnormal shadow candidate AS ′ emphasized as described above is represented in the tissue shadow TS, so that it is difficult to overlook the abnormal shadow candidate AS ′. can get. Further, by observing the composite image, it is possible to observe the position of the abnormal shadow candidate AS ′ in the tissue shadow TS and the mode of the tissue shadow TS around the abnormal shadow candidate AS ′. Therefore, when the original image is observed after viewing the composite image, it is possible to quickly and reliably grasp the position of the abnormal shadow candidate AS in the original image corresponding to the specific abnormal shadow candidate AS ′. In addition, since the tissue aspect around the abnormal shadow candidate AS can be easily grasped, the diagnosis for the abnormal shadow candidate AS can be performed quickly and accurately. In particular, since the microcalcification region is easily overlooked, the effect of preventing oversight of the diagnosis target is high.

  The original image (a), the emphasized image (b), and the composite image (c) shown in FIG. 4 are changed for convenience of explanation by changing the line width of the tissue shadow TS and the size of the abnormal shadow candidate according to the display intensity. Although they are drawn as line drawings, these are merely schematic and are different from actual images. Thus, actual image examples are shown in FIGS. 11 to 14 are examples of original images, and FIGS. 15 to 18 are examples of emphasized images. Here, the emphasized image is obtained by first applying the median filter, performing top-hat conversion, applying the median filter again, and finally performing binarization processing. FIGS. 19 to 22 show synthesized images formed by synthesizing the original image and the emphasized image in the addition mode with a synthesis rate of 50% (0.5).

  FIG. 5 is an explanatory diagram schematically showing a display mode when the composition rate S of the composite image is increased or decreased. When the synthesis rate S is changed by the above-described synthesis parameter changing means and the synthesis rate S is sequentially increased in the order of FIGS. 5A, 5B, and 5C, the display mode of the tissue shadow TS is gradually increased. The display mode of the abnormal shadow candidate AS ′ is gradually strengthened. At this time, since the composition rate S can be changed by the operation of the input unit 2 or the like, the display mode of the composite image (weighting of the original image and the emphasized image) can be adjusted according to the preference of the interpreter. Thus, oversight of the abnormal shadow candidate AS ′ can be prevented, and the positioning of the abnormal shadow candidate AS ′ in the tissue shadow TS and the reading of the state of the surrounding tissue shadow TS can be facilitated. For example, if importance is placed on preventing oversight of the abnormal shadow candidate AS ′, the composition rate S may be increased. If importance is placed on the positioning of the abnormal shadow candidate AS ′ in the tissue shadow TS, the composition rate S is set to an intermediate value. If the emphasis is on grasping the state of the tissue shadow TS around the abnormal shadow candidate AS ′, the composition rate S may be reduced.

  In addition, since the image reader basically makes a diagnosis based on the original image, it is generally necessary to display not only the synthesized image but also the original image. In this case, the original image and the composite image can be displayed in parallel by providing a plurality of screens in the same display means, by providing a plurality of display means, or by switching and displaying in the same display screen. Both images may be visible. However, when the composition rate S is changed by the composition parameter changing unit as described above, a period for setting the composition rate S to 0 or a value close to 0 is provided. It is also possible to switch between the original image and the composite image. In addition, as described above, the composition rate S can be automatically increased or decreased with the passage of time, but at this time, the composition rate S may be temporarily set to 0 or a value close to 0.

  Next, in this embodiment, as an example of a method for forming a composite image, a description will be given of a composite image formation example suitable for improving the visibility of abnormal shadow candidates or displaying as a reduced image. FIGS. 6A to 6E are explanatory diagrams for explaining the composite image forming method of this example. Here, FIG. 6A is an original image and shows an example in which abnormal shadow candidates AS are arranged in the tissue shadow TS. FIG. 6B is an enhanced image similar to the above, in which the abnormal shadow candidate AS ′ is displayed as the tissue shadow TS is weakened and enhanced. FIG. 6C illustrates a corrected enhanced image in which a corrected abnormal shadow candidate AS ″ is displayed by expanding the abnormal shadow candidate area by further expanding the abnormal shadow candidate AS ′ in the enhanced image of FIG. In this example, as an abnormal area enlarging means, a function or step for performing an expansion process on an area where the abnormal shadow candidate AS ′ is present is provided for the enhanced image on which the abnormal shadow candidate AS ′ is displayed. The abnormal shadow candidate AS ′ is converted into a corrected abnormal shadow candidate AS ″ that occupies a region enlarged on the image display than the region where the abnormal shadow candidate AS ′ was present.

  In this example, the corrected composite image shown in FIG. 6D is formed by superimposing (combining) the above-described correction emphasized image with the original image. As the synthesis method (superposition method) at this time, basically, the various methods described above can be adopted. In this corrected composite image, a corrected abnormal shadow candidate AS ″ is displayed in the tissue shadow TS. In the corrected composite image based on the correction enhanced image configured in this way, there is a corrected abnormal shadow candidate AS ″. Since the area is enlarged (widened) as compared with the area where the abnormal shadow candidate AS ′ of the composite image based on the emphasized image is relatively present, even if the area is reduced as shown in FIG. It is possible to prevent the shadow candidate AS ″ from disappearing during display or being difficult to be visually recognized. Regardless of whether or not the reduced candidate display is performed, the corrected composite shadow image AS ″ has the visibility of the corrected abnormal shadow candidate AS ″. For further improvement, when the corrected composite image is used for image diagnosis, an abnormal shadow is overlooked compared to a case where a composite image formed without applying the abnormal region enlargement means (abnormal region enlargement process) is used. There is an advantage that the teeth can be further reliably prevented. Here, the abnormal region enlarging means or process may be applied to the enhanced image as described above, but may be applied to the synthesized image after being synthesized with the original image.

  In the above example, the dilation process is performed to change the abnormal shadow candidate AS ′ to the corrected abnormal shadow candidate AS ″. However, the dilation process may be repeated a plurality of times, and the opening process is performed to obtain a noise removal function. In addition, in the opening process, the number of expansion processes may be increased more than the number of contraction processes, and the modified composite image is displayed in the same manner as a normal composite image for diagnosis. 2, it may be displayed in the navigation area 3n of the diagnostic record display means 3a shown in Fig. 2. In this case, the navigation area 3n is displayed as the diagnostic record display means 3a as in the illustrated example. In addition, since it is necessary to perform other information display (display of the image information 5B and the diagnostic record 5C), the display area is restricted, so that it is often necessary to perform reduced display. Because of the function of the navigation area 3a, it is necessary to clearly indicate the positional relationship between the abnormal shadow candidate areas, so that the modified composite image in which the existence area of the abnormal shadow candidate AS ″ is relatively enlarged on the screen as described above. Is extremely effective as a display in the navigation area 3n.

  Next, a more specific embodiment of the image diagnosis support apparatus and image diagnosis support method will be described with reference to FIGS. FIG. 7 is a schematic flowchart of the present embodiment (the above-described image diagnosis support program), FIG. 8 is an explanatory diagram illustrating an example of a basic setting screen for image processing, and FIG. 9 is an explanatory diagram illustrating an example of a display setting screen. As shown in FIG. 7, in the present embodiment, when the above-described diagnostic imaging support program is started (SS1), an authentication process is first executed (SS2). For example, an interpreter is authenticated by an ID or a password. Is called. Subsequently, initialization processing is executed in the computer main body 1 which is a local computer (SS3), and processing including reading of initial information from the server system 4 is executed. An initial screen (not shown) is displayed.

  Next, through various operations on the initial screen, the examination database 5D of the server system 4 is accessed, and desired data (the image data 5A and the image information 5B) are read (SS4). Thereafter, analysis of DICOM information or the like of the data is executed (SS5), and predetermined information (for example, image information 5B, image data 5A in the navigation area 3n as necessary) is displayed on the display screen of the diagnostic recording display means 3a. ) Is displayed (SS6). Next, it is confirmed whether or not the condition setting relating to image processing and image display has been completed (SS7). If the condition setting has been completed, the process is started. If the condition setting has not been completed, the default condition is set. The process proceeds to the setting routine (SS8).

  The image processing basic setting screen (image processing setting screen) shown in FIG. 8 is one of the setting screens performed in the condition setting routine, and sets the display mode of the original image (changes the processing setting of the histogram processing). A pre-processing setting unit A shown in the upper part of the figure for performing) and an emphasis processing setting part B shown in the lower part of the figure for setting the mode of the emphasis processing when generating the emphasized image from the original image. ing. In the pre-processing setting unit A, for example, in the above example, a histogram (a graph indicating a distribution of pixel values such as a luminance distribution and a hue distribution) is displayed for each of the acquired images, and in response to an input operation on the display screen. Thus, the processing mode of the histogram processing can be set. In the illustrated example, the acquired image is displayed in each row of the table on the left side of the diagram, and a histogram H1 of the grayscale image displayed in the selected row (in the illustrated example, the second row is indicated by hatching) is illustrated. Displayed in the graph on the right. In the graph, the tone curve TC is also displayed together with the histogram H1.

  Here, when the check boxes for gradation width adjustment, free curve adjustment, and level correction are selected, the respective adjustment operations can be performed. For example, when the free curve adjustment is selected, the histogram H1 can be arbitrarily adjusted and set by operating the tone curve TC in the graph on the right side of the figure, and changed to the histogram H2. When the OK button is pressed, the above selection is made. The definition file name of the histogram processing is input to the processing definition column in the table on the left side of the figure in the line. As a setting method of the histogram processing, not only the adjustment of the tone curve TC described above but also a default processing setting can be input by using the input field above the graph. It is also possible to perform the above-described gradation width adjustment and level correction by operating the illustrated triangular knob on the ruler at the bottom of the figure.

  In the table on the left side of the figure, a plurality of acquired images included in the accessed database are all displayed for each row, and the above-described histogram processing can be set for each image. In addition, a setting line for buttons A and B formed at the upper right of the display screen shown in FIG. 10 is also provided at the bottom of the table, and histogram processing executed when each button A and B is selected. Can be set in advance. The histogram processing set for each acquired image by the pre-processing setting unit A is automatically executed, and the image that has been subjected to histogram processing is used as the original image. Note that the pre-processing of the acquired image is not limited to the histogram processing in the illustrated example, and various processes for obtaining an original image can be executed.

  In the emphasis processing setting unit B, a table provided with rows (rows having the same numbers) corresponding to the respective rows of the table of the above-described preprocessing setting unit A is formed, and in each row, an original image corresponding to the above-described acquired image The contents of the processing steps of the emphasis processing applied to the can be set. In the lower part of the table, the same setting lines for buttons A and B as described above are also prepared. In each row, the content of enhancement processing for the original image is set in column order.

  In the table of the emphasis process setting unit B, a plurality of process steps (step 1, step 2,...) Are provided in order. In each processing step, various processing contents for the original image can be input by selecting a choice prepared in advance. The contents of the processing step include median filter, top hat transform, wavelet transform, variable ring filter (triple ring filter in the illustrated example), binarization processing, and the like. Further, it is configured so that the presence or absence of mask processing can also be set. Here, the mask processing is to divide the range including the abnormal shadow candidate region of the emphasized image and the other range (partial region such as the background), and at least a part of the region collectively has the same pixel value. This is a process for removing noise in regions other than the abnormal shadow candidate region, such as setting to, performing contraction processing, or performing smoothing processing.

  FIG. 9 shows an example of a setting screen for performing various settings of display modes used for image diagnosis. In this setting screen, a basic display setting unit C in the upper part of the figure and a composite display setting part D in the lower part of the figure are provided. In the basic display setting unit C, the number of screens of the system (the number of display monitors, for example, 3 in the example shown in FIG. 2) and the screen type (for example, the left and right display monitors are set as the operation screen) It can be set by selecting from the pull-down list or the like. Moreover, the screen position of each display monitor can be individually set by inputting the X and Y coordinate values of the screen.

  The basic display setting unit C can also set the display mode of the initial display screen. In the illustrated example, the initial screen layout mode (MLO only, CC only, MLO and CC order, CC and MLO order), and the type of composite mode when forming a composite image (the above addition mode and transmission mode) Selection) and display mode modes (for example, the main VIEW which is the image display screen shown in FIG. 10 and the shadow VIEW which makes the buttons and tabs arranged around the window of the display screen semi-transparent display) Selection etc.) can be set respectively. With Shadow VIEW, the transmittance can be set with a slide bar or the like. In addition, it is possible to set whether to display multiple images as thumbnails on a part of the screen, whether to display various images temporarily, and how many seconds the display time is set with a check box etc. It has become.

  In the above-described composite display setting unit D, the initial composite rate S can be set for each of the addition mode and the transmission mode using a slide bar or the like, for example, as shown in the drawing. Also, whether to enable the composition rate adjustment function, whether to enable automatic variation of the composition rate, and whether to perform reciprocal variation (alternating repetition of increase and decrease of composition rate) in the case of automatic variation Whether or not can be set by a check box or the like. In the case of automatic change, it is also possible to set a width for changing the synthesis rate by using a slide bar or the like.

  In the setting screen shown in FIGS. 8 and 9, various setting contents can be saved (save button), canceled (return button), and setting data can be output (setting output button). It is. The setting contents may be configured to be stored only in the local computer, or may be configured to be stored as part of the image information 5B or the examination database 5D.

  Returning to FIG. 7 again, the explanation of the diagnostic imaging support program will be continued. First, the acquired image is processed according to the setting contents of the pre-processing setting unit A set as described above to form an original image. Based on the original image, the setting contents of the above-described enhancement processing setting unit B are changed. In response, enhancement processing is executed, and an enhanced image is generated (SS9). Thereafter, a composite image is formed according to the setting contents of the basic display setting unit C and the composite display setting unit D (SS10), and the composite image is, for example, the display means 3 (image display display means 3b and 3c) of the system. Is displayed (SS11).

  FIG. 10 shows an example of a main VIEW display screen for displaying various images. In the example shown in FIG. 2, this display screen is displayed on the image display display means 3b and 3c. In this display screen, an image feed button, a subject advance button, an image return button, a subject return button, etc. are provided in the upper operation section provided at the top of the window to send or return a set of subjects' captured images. Image selection operation unit, MLO image only display button, CC image only display button, MLO and CC image display button, CC and MLO image display button, display mode selection, A button and B button, etc. The displayed display mode selection operation unit, an end button, and the like are arranged.

  In addition, an image switching operation unit in which an original image button for displaying an original image, an enhancement button for displaying an emphasized image, a combination button for displaying a composite image, and the like are arranged in a lower area of the upper operation unit And a synthesis rate adjustment operation unit including a slide bar for adjusting the synthesis rate S is provided. In addition, a transparent button for selecting a synthesis mode, an addition button, a symbol button for displaying a symbol indicating the position of an abnormal shadow candidate, an arrow button for displaying an arrow, and the like are additionally prepared. In addition, an image display area for displaying various images is formed below the upper operation unit, and a plurality of images including the displayed images are displayed as thumbnails in a lower part below the image display area. Is displayed. In the illustrated example, any one of the original image, the emphasized image, and the composite image can be selected and displayed in the image display area. Note that any one or more of the original image, the emphasized image, and the composite image may be displayed. Furthermore, a plurality of similar images (for example, two images of MLO-L and MLO-R) may be displayed simultaneously.

  Here, in the example shown in FIG. 2, by selecting an appropriate part of the modified composite image displayed in a reduced size in the navigation area 3n of the diagnostic recording display means 3a (performing a click operation etc. with the mouse 2b), An enlarged image of the original image or the synthesized image may be displayed on the image display area shown in FIG. 10 of the image display display means 3b, 3c.

  In the display screen shown in FIG. 10, the types of the original image, the enhanced image, and the synthesized image are appropriately selected by operating the original image button, the emphasizing button, and the synthesizing button during the interpretation and diagnosis work by the interpreter. It can be switched and displayed (SS12). At this time, a plurality of types of images may be displayed concurrently, or a plurality of the same types of images may be displayed simultaneously. Furthermore, it is also possible to change and display the composite rate of the composite image appropriately by operating the composite rate adjustment operation unit (SS13). Thereafter, when the interpreter inputs the diagnostic result 5C, the diagnostic result 5C is accepted (SS14), and the diagnostic result is stored in the information recording means in the local computer 1, the examination database 5D, etc. according to the storing operation of the interpreter (SS15). .

  Note that the image diagnosis support apparatus and image diagnosis support method of the present invention are not limited to the illustrated examples described above, and it is needless to say that various changes can be made without departing from the scope of the present invention. For example, in the above-described embodiments and examples, description has been made on the premise of image diagnosis of mammography examination. However, the present invention is not limited to such a case, and various diagnoses are made based on medical images acquired from various subjects. It can be widely applied to

S1 ... Original image capture / display, S2 ... Emphasized image generation means, S3 ... Composite image formation means, S4 ... Image display means, S5 ... Composite parameter change means, 1 ... Computer main body, 1a ... Information communication section, 2 ... Input means, 3 ... display means, 4 ... server system, 5A ... image data, 5B ... image information, 5C ... diagnostic record, 5D ... examination database

Claims (8)

An image diagnosis support apparatus for supporting diagnosis using a medical image acquired from a subject,
Enhanced image generating means for generating an enhanced image by emphasizing an abnormal shadow candidate region by performing processing on the original image of the medical image;
Composite image forming means for forming a composite image in which the tissue shadow of the original image and the abnormal shadow candidate region of the enhanced image are displayed together by superimposing the original image and the enhanced image;
Image display means for displaying the composite image;
An image diagnosis support apparatus comprising:   The image diagnosis support apparatus according to claim 1, wherein the enhanced image is an abnormal shadow candidate region represented as an island region in a substantially uniform background.   The image diagnosis support apparatus according to claim 1, wherein the composite image forming unit forms the composite image by an addition process of the original image and the emphasized image.   The composite image forming unit forms the composite image by making a part of the emphasized image other than the abnormal shadow candidate region transparent or semi-transparent and superimposing the original image. The image diagnosis support apparatus according to 1 or 2.   5. The image diagnosis support apparatus according to claim 1, wherein the image display unit is configured to be able to switch and display the original image and the composite image or to display them in parallel. .   It further comprises synthesis parameter changing means for changing a synthesis parameter related to a superposition mode of the original image and the emphasized image by the synthesized image forming means in response to a specific operation input or automatically. The diagnostic imaging support apparatus according to any one of claims 1 to 4.   The image diagnosis support according to any one of claims 1 to 6, further comprising an abnormal area enlarging unit that expands the abnormal shadow candidate area by performing an expansion process on the abnormal shadow candidate area. apparatus. An image diagnosis support method for supporting diagnosis using a medical image acquired from a subject,
An enhanced image generation step in which an enhanced image is generated by emphasizing an abnormal shadow candidate region by processing the original image;
A composite image forming step for forming a composite image in which the tissue shadow of the original image and the abnormal shadow candidate region of the emphasized image are displayed together by superimposing the original image and the enhanced image;
An image display step in which the composite image is displayed;
An image diagnosis support method comprising: