JP5001248B2 - Diagnostic imaging support device - Google Patents

Description

  The present invention relates to an image diagnosis support apparatus that extracts a shadow or the like as a lesion candidate from a medical image using computer image processing and displays the extracted shadow as a lesion candidate in an identifiable manner.

  A conventional image diagnosis support apparatus is disclosed in Patent Document 1. The procedure disclosed here is to first convert a medical image including a CT image, MR image, ultrasonic image, and a difference image between a past image and a current image into a multi-value by a multi-value conversion means. Find the center coordinates of the shadow for the image. Various image processes are performed on the medical image and the multi-valued image on the basis of the central coordinates to discriminate what seems to be a lesion candidate. In this way, when automatically determining a lesion candidate from a medical image, shadows of different sizes and shapes can be handled in a unified manner, and computer calculations can be completed in a short time.

  This diagnostic imaging support apparatus is capable of detecting cancers and polyps that have occurred in various organs such as the stomach, large intestine, and bronchi as abnormal shadows.

International Publication WO2002 / 02002 Publication

  However, in the above prior art, there are cases where it is difficult to discriminate the cancer that has formed on the wall of the organ as an abnormal shadow, and therefore, establishment of a more accurate discrimination method has been desired.

  In addition, although the above-described conventional shadow determination processing is performed by selecting only a medical image having the target organ, the operator is informed that the shadow determination processing is not performed for an image without the target organ. It was not considered. Therefore, there is a possibility that an operator who does not know the contents of the discrimination process may mistakenly recognize that the shadow discrimination process has been performed on an image without the target organ.

An object of the present invention is to provide an image diagnosis support apparatus capable of notifying an operator that a shadow discrimination process is not performed when displaying an image without a target organ .

An image diagnosis support apparatus according to the present invention includes a multi-value quantization unit that performs predetermined image processing on a medical image to create a multi-value image, and at least one of the multi-value image created by the multi-value image conversion unit. In an image diagnosis support apparatus comprising an extraction unit that performs two or more determination processes and extracts a lesion candidate shadow that is a lesion candidate,
A determination unit configured to determine the presence or absence of a target organ to be subjected to the determination process by image processing; and when the determination unit determines the presence or absence of the target organ, the extraction unit extracts a lesion candidate shadow, and the determination In the case where there is no determination of the presence or absence of the target organ by the means, the display further includes display means for displaying that the focus candidate shadow is not extracted . In the present invention, since it is possible to notify the operator that the determination processing is not performed on the medical image that has not been subjected to the determination processing by displaying a mark or a pattern, the operator does not have the target organ even if the operator does not know the processing content. It was made possible to recognize that the shading determination processing was not executed on the image.

As described above, according to the diagnostic imaging support apparatus of the present invention, it is possible to provide an diagnostic imaging support apparatus capable of notifying the operator that the shadow discrimination processing is not performed when displaying an image without a target organ. effective.

  Preferred embodiments of an image diagnosis support apparatus according to the present invention will be described below with reference to the accompanying drawings.

FIG. 1 is a block diagram showing the hardware configuration of the entire diagnostic imaging support apparatus to which the present invention is applied. This image diagnosis support device displays or extracts extracted lesion candidate shadows based on a plurality of tomographic images (CT images, etc.) collected for a target region of a subject, for example, with an X-ray CT apparatus or the like. Of the candidate lesion shadows, the ones with high certainty were narrowed down and displayed. In addition, an image in the middle of these processes is displayed. This diagnostic imaging support apparatus includes a central processing unit (CPU) 40 that controls the operation of each component, a main memory 42 that stores a control program for the apparatus, and a magnetic that stores a plurality of tomographic image data, programs, and the like. A disk 44, a display memory 46 for temporarily storing image data for display, a CRT display 48 as a display device for displaying an image based on the image data from the display memory 46, and a soft switch on the screen are operated. The mouse 50 and its controller 52, a keyboard 54 having keys and switches for setting various parameters, a speaker 58, and a common bus 56 for connecting the above components. In this embodiment, a case where only the magnetic disk 44 is connected as a storage device other than the main memory 42 is shown, but in addition to this, a floppy disk drive, a hard disk drive, a CD-ROM drive, a magneto-optical disk ( MO) drive, ZIP drive, PD drive, DVD drive, etc. may be connected. Furthermore, it is possible to connect to various communication networks such as LAN (local area network), the Internet, and telephone lines via a communication interface not shown so that image data can be exchanged with other computers. Also good. Further, the exchange of image data may be performed by connecting a medical image diagnostic apparatus such as an X-ray CT apparatus or an MRI apparatus capable of collecting tomographic images of a subject to the LAN or the like.

  Hereinafter, an operation example of the image diagnosis support apparatus in FIG. 1 will be described with reference to the drawings. FIG. 2 is a diagram illustrating an example of a first flowchart executed by the image diagnosis support apparatus. The CPU 40 in FIG. 1 operates according to the first flowchart. FIG. 3 is a diagram showing an example of a display screen on the CRT display. 4 to 7 are diagrams showing how the CT image is processed according to the first flowchart. Hereinafter, details of the first flowchart will be described in the order of steps.

[Step S21]
As shown in FIG. 4, the CPU 40 reads CT images I1, I2,... In of the patient imaged by the CT apparatus from the magnetic disk 44, and stores the obtained CT images I1, I2. For example, each CT image is arranged along the XY plane and stacked in the Z direction in the three-dimensional space represented by the virtual XYZ axes on 42, and the three-dimensional image is calculated. The data is displayed on the CRT 48 via the display memory 46. The three-dimensional image here refers to an endoscopic three-dimensional image performed by a technique called central projection with a temporary viewpoint placed in the lumen of a target organ such as the large intestine, bronchus, blood vessel, and stomach. . Details of this type of three-dimensional image are described in Japanese Patent Application Laid-Open No. 7-210704 filed by the present applicant.

[Step S22]
The operator can freely advance / retreat the viewpoint along the inside of the target organ using an input means such as a mouse. In this way, the abnormal projection X that seems to be a lesion site is found while moving the viewpoint forward / backward, and the viewpoint E is set in association with the abnormal projection X.

[Step S23]
The CPU 40 projects an endoscopic three-dimensional image on the projection plane PP based on the set viewpoint E. Here, the projection plane PP is arranged on an extension with the line-of-sight vector VE from the viewpoint E as a normal vector. Of course, the line-of-sight E, the line-of-sight vector VE, and the projection plane PP are virtually set on the main memory 42.

[Step S24]
The CPU 40 stores the three-dimensional image projected on the projection plane PP in the display memory 46, and displays the stored three-dimensional image on the CRT 48 as indicated by 31 in FIG.

[Step S25]
The CPU 40 sets a cross section PE that passes through the abnormal protrusion X and the viewpoint E and is parallel to the projection plane PP, and displays the cross section PE on the CRT 48 as an image of the cross section, as indicated by 32 in FIG. This cross-sectional image is obtained by interpolating a CT image at a position where the abnormal protrusion X exists or a plurality of CT images near the position. Further, since the cross section PE passes through the curved surface P on the three-dimensional image 31, it is clearly indicated by a bold line.

[Step S26]
As shown in FIG. 5 (a), the CPU 40 calculates the center of gravity C of the lumen of the target organ (curved surface P in the previous step), draws a straight line of a predetermined length with the center of gravity C as one end, and Are read out (for example, three pixel values a1 to a3). Here, the center point for rotating the straight line is the center of gravity. However, in addition to the center of gravity, an inner center that can be the center point of the lumen region may be used. Further, the center point may be in a substantially central portion, and a slight positional deviation from the true center point is allowed.

  Next, as shown in FIG. 5 (b), the CPU 40 rotates the straight line by an angle θ until it reaches the abnormal projection X, and the points b1 to b3 3 on the concentric circles of the points a1 to a3, respectively. Read the pixel value of a point. Finally, the CPU 40 determines whether or not the read pixel point is a lesion part from the following discriminant.

Sθ＞定数 ・・・式(2)

式(1)の演算の結果、式(2)、式(3)の何れかが成立すれば病巣部と判別する。この判別に基づいて病巣部候補の陰影を抽出する(第1の抽出手段)。

Sθ> Constant ・ ・ ・ Equation (2)

As a result of the calculation of Expression (1), if either Expression (2) or Expression (3) is satisfied, it is determined as a lesion. Based on this determination, the shadow of the lesion candidate is extracted (first extraction means).

  The pixel value readout point may be selected from pixel values within a certain distance from the boundary between the outside of the target organ (mainly air) and the wall of the target organ to the wall. In this way, pixel points to be read out are set within the range of the target organ, so that the extraction of the lesion is more reliable.

[Step S27]
Here, as shown in the shaded portion in FIG. 6 (a), the stomach portion of the CT image is the processing target. FIG. 6 (b) is an enlarged view of FIG. 6 (a), and the stomach includes a protrusion that is recognized as a polyp.

  As shown in FIG. 6 (c), the CPU 40 draws a tangent line that complements so that a, b, and c of the depressed portion of the stomach become a closed figure, thereby generating a complement region. Here, attention is focused on the dent of the stomach, but if a feature is recognized in the convex portion, a complementary region may be generated for the convex portion.

  In general, it is known that if the shape of the generated complementary regions a, b, and c is slender, it is a shadow indicating a normal tissue, and if it is round, it is a shadow indicating a lesion candidate. It is disclosed in Document 1.

  Next, the CPU 40 determines that a and c, which are elongated shapes, are normal tissues and b is a lesion candidate from FIG. Finally, as shown in FIG. 6 (e), the CPU 40 extracts the shadow of the lesion candidate based on the determination (second extraction means).

[Step S28]
The CPU 40 leaves the portion determined to be an abnormal shadow in step S26 and step S27 as the lesion, and deletes the portion that is not. For example, as shown in FIG. 6 (e), a circular marker process is performed on a CT image or a cross-sectional image and displayed on the CRT 48 using the remaining lesion portion as the candidate lesion shadow.

  In the present embodiment, step S27 is executed after step S26. However, since steps 26 and 27 are independent of each other, the execution order of the processes may be reversed. Each process may be performed in parallel and then combined.

  Further, in the present embodiment, since the complementary region in step S27 is new, there is a method in which only this process is executed in a case where a lesion candidate shadow can be formed only by this process.

  As described above, the CPU 40 scans a predetermined radius from the center point set at the approximate center of the lumen of the target organ, reads pixel data of a plurality of points through which the scanned radius passes, and reads the read pixel The first extraction step (step S26) for obtaining difference information of the data and extracting the lesion candidate shadow, and calculating a complementary region to be a closed figure in the concave or convex region for the cross section, and the calculated complement A second extraction step (step S27) for extracting a lesion candidate shadow by the shape of a closed figure complemented with a region and a result extracted by each of the first and second extraction means are combined to provide high lesion confidence. A program comprising the step of extracting only the lesion candidate shadow (step S28) is executed. For this reason, it is possible to accurately determine the lesion shadow and other shadows on the wall of the target organ.

  Next, another operation example of the image diagnosis support apparatus in FIG. 1 will be described. FIG. 7 is a diagram illustrating an example of a second flowchart executed by the image diagnosis support apparatus. The CPU 40 in FIG. 1 operates according to the second flowchart. FIG. 8 is a diagram showing an example of a display screen on the CRT display. FIG. 9 is a diagram showing an example of a display form different from FIG. Details of the second flowchart will be described below in the order of steps.

[Step S71]
The CPU 40 determines whether or not the CT image has a target organ including the stomach, large intestine, and bronchus, and the CT image without the target organ executes the next step, and the CT image with the target organ moves to step S74. Perform the process of jumping.

  The above determination is made by referring to keywords and codes such as the abdomen when there is an imaging part as image supplementary information of the CT image, or when there is no image supplementary information, the anatomical shape of the target organ or the X-ray CT apparatus In the case of the CT image taken in step 1, there is a method of extracting an organ by image processing using the CT value reflected by the X-ray absorption amount. Moreover, in order to extract an organ more reliably, an imaging part as image supplementary information and image processing may be combined.

[Step S72]
In order to clearly indicate that the CT image does not execute the lesion shadow candidate extraction (CAD) process, the CPU 40 attaches an “x” mark indicating that the CT image is not executed to the background portion of the image area, as shown in FIG. However, care is taken so that this “×” mark does not overlap the image. Further, instead of the “x” mark as shown in FIG. 8, a pattern may be added to the margin portion of the image as shown in FIG. In short, the CT image branched to this step only needs to make it possible for the operator or image diagnostician to recognize that the CAD processing has not been performed.

[Step S73]
The CPU 40 stores the CT image marked in the previous step in the main memory 42 as the first diagnosis support image.

[Steps S74 to S76]
The CPU 40 multivalues the CT image, extracts the lesion shadow, associates the lesion shadow with the image, and stores the CT image in the main memory 42 as the second diagnosis support image. Since each processing here is disclosed in Patent Document 1, description thereof is omitted.

[Step S77]
The CPU 40 rearranges the first diagnosis support image and the second diagnosis support image stored in the main memory 42 in the order of the original CT images.

[Step S78]
The CPU 40 displays the rearranged diagnosis support images on the CRT 48 via the display memory 46.

  As described above, the CPU 40 determines whether to execute the determination process based on the presence or absence of a target organ to be examined for a lesion (step S71), and adds a predetermined mark to the medical image that has been determined and not executed. A program comprising a step (step S72) of displaying and the step of displaying the attached mark and the medical image not subjected to the discrimination processing (step S78) is executed. For this reason, since it is possible to notify the operator that the discrimination process has not been performed with a mark or pattern display for a medical image that has not been subjected to the discrimination process, an image without the target organ can be displayed even for an operator who does not know the processing content. It can be recognized that the shadow discrimination processing is not executed.

The block diagram which shows the hardware constitutions of the focus candidate extraction and display apparatus to which this invention is applied. FIG. 2 is a diagram showing a first flowchart executed by the apparatus of FIG. FIG. 2 is a diagram showing an example of a display screen on the CRT display of FIG. FIG. 3 is a diagram for explaining steps such as a three-dimensional image configuration and a cross-sectional image creation in the process of the first flowchart of FIG. 2; FIG. 3 is a view for explaining the first half of the lesion shadow determination in the first flowchart of FIG. 2; FIG. 3 is a view shown for explaining the latter half of the focus shadow determination in the first flowchart of FIG. 2; FIG. 4 is a diagram showing a second flowchart executed by the apparatus of FIG. FIG. 8 is a diagram showing an example of a display form in which a portion other than an image is marked for a medical image that has not been subjected to shadow determination as a result of the second flowchart of FIG. FIG. 9 is a diagram showing an example of a display form different from FIG.

Explanation of symbols

  40 Central processing unit (CPU), 42 Main memory, 44 Magnetic disk, 46 Display memory, 48 CRT display, 50 Mouse, 52 Controller, 54 Keyboard, 56 Common bus

Claims (3)

A multi-value quantization unit that performs predetermined image processing on a medical image to create a multi-value image, and at least one determination process is executed on the multi-value image created by the multi-value image conversion unit. In an image diagnosis support apparatus comprising an extraction means for extracting a lesion candidate shadow that is a candidate of
Wherein comprising a determining means for the presence or absence of a target organ to be discrimination processing determined by the image processing, if the determination of the presence or absence of the target organ by the determining means is closed, the extraction means extracts a focus candidate shadow, the determination An image diagnosis support apparatus, further comprising display means for displaying that the focus candidate shadow is not extracted because the means does not determine whether or not the target organ is present. The determining means uses the anatomical shape of the target organ, the image diagnosis support device according to claim 1, wherein the determining the presence or absence of the target organ.   2. The image diagnosis apparatus according to claim 1, wherein the display unit marks a portion that does not overlap the image region in order to display that the focus candidate shadow is not extracted.

Images