JP2002325762A - Image diagnosis support instrument - Google Patents

Abstract

PROBLEM TO BE SOLVED: To efficiently separate a part which is digitized shade where a blood vessel is branched or beard shape blood vessels are connected in a focus shade, etc., and also to perform a detection processing with a high detection rate concerning an abnormal shade candidate through the use of a feature in the shade. SOLUTION: A digitizing means makes a medical image into the digital one, the center coordinates of the shade are obtained based on the multi-valued image, various image processings are performed in the medical image or the multi-valued image with the center coordinates as reference and, then, the part which is supposed to be a focus candidate is discriminated. A density difference, i.e., the differential value of pixel values between two inner and outer places with the edge of the shade in the multi-valued image as reference is obtained in the shade of the actual medical image or the image to be discriminated so that it is discriminated whether the shade is the focus candidate shade or not. The attribute of an object isolated area is reflected on the radius of a cut circular shape. That is, the radius of the separated circular shape is decided based on the minimum radius vector of the shade and, then, a separation processing is performed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to image diagnosis support for extracting shadows and the like as lesion candidates from medical images using computer image processing, and displaying the extracted shadow and lesion candidates as identifiable images. Related to the device.

[0002]

2. Description of the Related Art Conventionally, a computer aided computer analyzes the shadows of an image captured by a CT or MRI apparatus, narrows down lesion candidates from the shadows, presents them to a doctor, and asks the doctor for a diagnosis. Has been done. Various medical images of lung fields have been reported as examples of narrowing focus candidates from shadows. As one of the methods, as a method for distinguishing elongated blood vessel shadows and cancer shadows close to a circle from medical images of lung fields, for example, "Quitofilter" (November 1999, 9th Annual Meeting of the Computer Aided Diagnostic Imaging Society of Japan) Vol.21). In the medical image of the lung field, in addition to shadows such as cancer, blood vessels, cross-sections of blood vessels, cross-sections of bronchi, etc. are mixed, so the shadows that are considered cancer candidates are extracted from these images. It is desirable to present to a doctor.

[0003]

However, the actual shadows have various sizes and shapes, and in order to enhance the ability to discriminate the shadows, much effort is required to adjust the parameters and it is difficult to use. If there is a method that can handle shadows having different sizes and shapes in a unified manner, it will be easier to create a computer program, and it will be easier to adjust parameters to increase discrimination ability. Further, if lesion candidates can be narrowed down from shadows by simple processing, the calculation time of the computer can be reduced, and accurate lesion candidates can be extracted quickly. Further, it is desirable that the extracted lesion candidate can be instantaneously displayed to a doctor. Therefore, the applicant of the present application has applied for an image diagnosis support apparatus that can uniformly handle shadows having different sizes and shapes and that requires only a short time for a computer operation (Japanese Patent Application No. 2000-131).
001-187969). This image diagnosis support device,
A CT image of a patient captured by a CT device is read from a storage device such as a magnetic disk, and a multi-value processing is performed on a diagnosis target organ from the read CT images to generate a multi-value image. In this multi-valued image, there are cases where parts or organs of a plurality of organs are connected to each other, and therefore a cutting process is performed to separate them into individual parts or organs. Then, an optimum detection process corresponding to the site or the type of the cut organ to be diagnosed is performed. In this detection processing, the type of a part or an organ is determined, image processing suitable for the type is performed, a focus candidate shadow is narrowed down, and a shadow as a focus candidate, that is, an abnormal shadow is detected. This abnormal shadow detection processing is performed without using the original image (CT image).
This is performed based on only the multi-valued image or based on both the CT image and the multi-valued image. Then, those determined as abnormal shadows are left as lesions, those that are not are deleted, and the abnormal shadows are displayed with color information, markers, and the like added so as to be easily understood in the CT image. In the above-mentioned application, various proposals have been made as a method of cutting, but for those in which a whisker-like blood vessel is connected to a focus shadow such as a branch portion of a blood vessel or a cancer,
The cutting process did not always work efficiently. Therefore,
It has been an important issue to efficiently perform a cutting process and perform an abnormal shadow process. Further, in the above-mentioned application, various proposals have been made as a method for detecting and processing an abnormal shadow, and it is determined whether or not the shadow should be a candidate for the abnormal shadow. The detection rate of abnormal shadow candidates differs depending on whether or not it is used. Therefore, how to perform a detection process using an excellent feature amount has been an important issue.

[0004] An object of the present invention is to automatically determine a lesion candidate or the like from a CT image, an MR image, an ultrasonic image, and a medical image including a difference image between a past image and a current image using a computer. It is an object of the present invention to provide an image diagnosis support apparatus capable of efficiently performing a cutting process on a multivalued shadow in which a whisker-like blood vessel is connected to a branch portion of a blood vessel or a focus shadow. .

An object of the present invention is to automatically determine a lesion candidate or the like from a CT image, an MR image, an ultrasonic image, and a medical image including a difference image between a past image and a current image by using a computer. It is an object of the present invention to provide an image diagnosis support device capable of performing a detection process with a high candidate detection rate of an abnormal shadow by using a feature of the shadow.

[0006]

According to a first aspect of the present invention, there is provided an image diagnosis support apparatus which extracts only pixels belonging to a pixel value range corresponding to a type of a shadow to be determined from a medical image or the medical image. A multi-level converting means for performing predetermined image processing on the created medical image to be determined to generate a multi-level image; detecting a center or a center of gravity of the shadow based on the multi-level image; The multi-valued image, the medical image or the medical image for discrimination is rotated on a shade in the medical image for the discrimination target, and the radius and the edge of the multi-valued image, Sampling the pixel values of the shades in the medical image or the medical image for discrimination at at least two locations separated by a predetermined distance inside and outside the multi-valued image with reference to the intersection of Based on the difference value, Those having an extraction means for determining whether the nest candidate shadow. Medical images include shadows such as cancer, blood vessels, cross-sections of blood vessels, cross-sections of bronchi, etc., so that even if image processing is performed directly on medical images, lesion candidates that are candidates for lesions It is very difficult to extract shadows. Therefore, in the present invention,
The medical image is multi-valued by the multi-value conversion means, the center of the shadow or the barycentric coordinates are obtained based on the multi-valued image, and various image processing is further performed on the multi-valued image based on the center coordinates. In addition, shadows that are not lesions, such as blood vessels, blood vessel cross-sections, and bronchial cross-sections, are efficiently deleted, and as a result, only lesion candidate shadows with high lesion confidence are extracted. As one method of extracting only lesion candidate shadows, in the present invention, a shadow in a medical image or a discrimination target image is
It is known that the density near the boundary, that is, the CT value is blurred, and shows various density values (CT values). Therefore, the shadow density difference of the actual medical image or the image for the discrimination target at two locations inside and outside the multivalued image is determined with reference to the edge of the shadow of the multi-valued image, that is, the difference value of the pixel value is obtained. Or not.

According to a second aspect of the present invention, in the first aspect, the extracting means determines whether or not the extraction means is inside or outside the multi-level image based on an intersection of the radius and an edge of the multi-level image. A pixel value of a shadow in the medical image or the medical image for discrimination at at least two positions on the radius separated by a distance is sampled. This method uses a point on the radial radius (the radius of rotation around the center of the shadow or the center of gravity of the multi-level image) as two positions inside and outside of the multi-valued image with reference to the edge of the shadow. It was done.

According to a third aspect of the present invention, in the first aspect of the present invention, the extraction means forms the edge of the multi-level image so as to include an intersection of the radius and the edge of the multi-level image. Of the shadow in the medical image or the medical image for discrimination at at least two places separated by a predetermined distance on the line perpendicular to the tangent line and at a predetermined distance inside and outside the multilevel image based on the intersection point The pixel value is sampled. In this method, a point on a line perpendicular to a tangent line at which the shadow edge of the multi-valued image is formed is used as two positions inside and outside of the multi-valued image with reference to the shadow edge. It is.

According to a fourth aspect of the present invention, in the first, second or third aspect, the extraction means is configured to determine the rotation angle of the radius as a horizontal axis based on a difference value between the two pixel values. Then, it is determined whether or not the shadow is a lesion candidate shadow based on the difference value waveform. In this method, lesion candidate shadows are determined based on a difference value waveform created using the difference value waveforms at two locations inside and outside the multi-valued image with the angle of the radial axis as the horizontal axis. This is because a remarkable difference appears in the difference value waveform between the case where the shadow is a lesion candidate and the case where the shadow is a shadow of a blood vessel cross section. In general, it is known that the density of a shadow that is not a focus candidate shadow is as simple as having one peak, and the density of a focus candidate shadow is as complicated as having multiple peaks. Have been. Therefore, when a difference value waveform is obtained for such a shadow, a waveform with relatively little change is shown when the shadow is not a lesion candidate shadow, and a waveform with many changes is shown for a lesion candidate shadow. Therefore,
Here, it is determined whether or not the shadow is a lesion candidate shadow based on these difference value waveforms.

According to a fifth aspect of the present invention, in the image diagnosis support apparatus according to the fourth aspect, the extracting means performs a Fourier transform on the difference value waveform, and based on a result of the Fourier transform, determines whether the shadow is a focus shadow. It is to determine whether or not. In this method, Fourier transform is performed to extract a feature of a difference value waveform, and a lesion candidate shadow is determined using the result. As described above, the difference value waveform shows a waveform with relatively little change when it is not a focus candidate shadow,
In the case of a lesion candidate shadow, a waveform with a large change is shown. Similarly, the result of the Fourier transform tends to shift to a higher frequency side in the case of the lesion candidate shadow as compared with the shadow of a blood vessel terminal or the like. Therefore, the shift amount toward the high frequency side, for example, the shift of the peak position becomes the feature amount of the shadow. If it is determined that the peak position is on the higher frequency side than the predetermined value, the shadow is determined to be a lesion candidate shadow, otherwise, it is determined that the shadow is not a lesion candidate shadow.

According to a sixth aspect of the present invention, in the fifth aspect, the extraction means determines whether or not the shadow is a focus shadow based on a magnitude relationship between frequency components as a result of the Fourier transform. Is determined. this is,
The determination is made based on the magnitude relationship between the frequency components as a result of the Fourier transform.

According to a seventh aspect of the present invention, in the image diagnosis support apparatus according to the fifth aspect, the extraction means is a focus shadow based on a magnitude of a frequency indicating a peak of the frequency component as a result of the Fourier transform. It is to determine whether or not there is. This is determined based on the magnitude of the frequency indicating the peak of the frequency component as a result of the Fourier transform.

According to an eighth aspect of the present invention, in the image diagnosis support apparatus according to the fourth aspect, the extracting means is arranged such that, in the difference value waveform, a pixel value located inside the multi-valued image is outside the multi-valued image. The ratio when the pixel value is smaller than the pixel value located at the position is determined, and based on whether the ratio is larger than a certain value, the shadow is regarded as the end of the blood vessel and excluded from the focus candidate shadow. is there. This is to make a direct determination based on the difference value waveform obtained in claim 4. In the case of a shadow at the end of a blood vessel, a shadow of a blood vessel portion connected to the shadow exists, so that the difference value waveform In some cases, the pixel value located inside the multi-valued image may be smaller than the pixel value located outside the multi-valued image, and the ratio may be larger than a certain value. It was considered to be the end of a blood vessel.

According to a ninth aspect of the present invention, in the first aspect, the extracting means detects a center or a center of gravity of the shadow based on the multi-valued image, and determines a center or a vicinity of the center of the shadow as a reference. By rotating a straight line of a predetermined length as a point on the shade in the multi-valued image, the medical image or the medical image for discrimination, the straight line and the multi-valued image, the medical image or the medical image for discrimination. Obtain the minimum value of the length of the straight line portion that intersects the shadow in the image, obtain the cutting radius based on the minimum value, leave the shadow included in the cutting circle formed by the cutting radius, other shadows And remove
It is determined whether or not the remaining shadow is a lesion candidate shadow. The simplest case of cutting an isolated area is cutting with a circle. This is to reflect the attribute of the target isolated region in the radius of the cutting circle, determine the radius of the cutting circle based on the minimum radius of the shadow, and perform the cutting process. .

According to a tenth aspect of the present invention, in the first aspect, the extracting means detects a center or a center of gravity of the shadow based on the multi-valued image, and determines a center or a vicinity of the center of the shadow as a reference. By rotating a straight line of a predetermined length as a point on the shade in the multi-valued image, the medical image or the medical image for discrimination, the straight line and the multi-valued image, the medical image or the medical image for discrimination. Determine the minimum and maximum values of the length of the straight line portion that intersects the shadow in the image, determine the cutting radius based on the ratio of the minimum value and the maximum value, to the cutting circle formed by the cutting radius This is to leave the included shadows and remove other shadows, and determine whether or not the remaining shadows are lesion candidate shadows. In this method, the radius of the cutting circle is determined based on the ratio between the minimum value and the maximum value of the moving radius, and the cutting process is performed.

According to an eleventh aspect of the present invention, in the ninth aspect, the extracting means detects a center or a center of gravity of the shadow based on the multi-valued image, and detects an edge of the shadow from the center or the center of gravity. The distance to the part is determined over the entire circumference of the shadow, the variance or standard deviation of the distance determined over the entire circumference is determined, and the variance or standard deviation is calculated based on the minimum value and the shadow. Is to determine whether or not is a focus candidate shadow. In this method, the variance value or the standard deviation of the radius and the distance to the edge of the shadow is obtained, and the discrimination processing is performed in association with the minimum value obtained in claim 9.

According to a twelfth aspect of the present invention, in the image diagnosis support apparatus according to the tenth aspect, the extracting means obtains an area of the shadow area and an area of a concave area formed at an edge of the shadow area. Determining the ratio between the area of the shadow region and the area of the recessed region, and determining whether the shadow is a lesion candidate shadow based on the calculated ratio and the ratio between the minimum value and the maximum value. It is. In this method, the area of the shadow and the area of the concave area at the edge are obtained, and the ratio of the area is associated with the ratio between the minimum value and the maximum value obtained in claim 10, and the discrimination processing is performed. It is.

According to a thirteenth aspect of the present invention, in the tenth aspect, the extracting means obtains an area of the shadow area and an area of a concave area formed at an edge of the shadow area. The vicinity of the position of the center of gravity of the two largest areas among the areas of the recessed areas is connected by a straight line or a curve, and the shadow area is cut using the straight line or the curve. In this method, the area of the concave region at the edge of the shadow is obtained, and the shading is cut using two of the larger areas. When the lesion candidate shadow overlaps the vascular shadow, this process can cut the vascular shadow and the lesion candidate shadow.

According to a fourteenth aspect of the present invention, in the thirteenth aspect, the shadow cut using the straight line or the curve is used to determine whether the shadow is a lesion candidate shadow. In this method, a lesion candidate shadow discriminating process is performed for both shadows after the cutting process.

According to a fifteenth aspect of the present invention, in the image diagnosis support apparatus according to the fourteenth aspect, the shadow after cutting using the straight line or the curve, which does not include the center of the shadow before cutting or the center of gravity. Is deleted, and it is determined whether or not the shadow after deletion is a lesion candidate shadow. In general, since the center or the center of gravity of the shadow candidate is often present in the portion including the shadow candidate lesion, the shadow which does not include the center or the center of gravity is deleted after the cutting processing according to claim 13. is there.

According to a sixteenth aspect of the present invention, in the first aspect, the extracting means divides a predetermined area of the multi-valued image into a plurality of areas, and assigns different parameters to each of the divided areas. Is used to determine whether or not the shadow is cut or a focus candidate shadow using a parameter assigned to the divided region where the center or the center of gravity of the shadow detected based on the multi-valued image is located. It performs processing. In this method, a parameter that is a criterion value is made dependent on the position of an organ in an image. A region corresponding to an organ is divided based on the multi-valued image, and a different parameter is assigned to each divided region. It is like that. This makes it possible to set appropriate parameters according to the position of the organ.

According to a seventeenth aspect of the present invention, in the first aspect, the extracting means generates an average image based on the multi-valued image, and generates an average image based on the multi-valued image or the averaged image. Detecting the center or center of gravity of the shadow based on, and rotating a straight line of a predetermined length on the shadow in the multi-valued image and the average image as a reference point near the center of the shadow or the center of gravity, the straight line and the Obtain a difference value of the length of a straight line portion that intersects the shadow in the multi-valued image and the average image, and creates a difference value waveform based on the difference value with the angle at the time of rotation of the straight line as the horizontal axis, It is to determine whether or not the shadow is a lesion candidate shadow based on the difference value waveform.
That is, the multi-valued image is further averaged in a predetermined sectioned area to create an average value image, and a difference value of the length of a straight line portion where the multi-valued image and the average value image intersect with the radial is calculated. , Based on the waveform of the difference value. In the case of a lesion candidate shadow, the average value image is almost the same as the multi-valued image, but in the case of a cross-sectional shadow of a blood vessel, the shape of the shadow in the multi-valued image and the average value image is different, and the change in the difference value waveform is different. Because it gets bigger
Based on this, it is possible to easily determine a lesion candidate shadow.

In the image diagnosis support apparatus according to the present invention, an unnecessary image is removed from the medical image before it is processed by the multi-level converting means, and is subjected to abnormal shadow detection processing. It is provided with processing image selecting means for selecting an object. The number of medical images and the like tends to increase, and a large number of medical images are generated by one examination.
Therefore, in the present invention, unnecessary images are removed from a large number of medical images, and an image to be subjected to abnormal shadow detection processing is selected. Thereby, the efficiency of the abnormal shadow detection process can be increased.

According to a nineteenth aspect of the present invention, in the image diagnostic assistance apparatus according to the eighteenth aspect, the processing image selecting means extracts a first medical image having a largest area of a predetermined region from the medical images, Correlating the predetermined area between the first medical image and the medical images before and after the first medical image, and determining whether the preceding and following medical images are to be subjected to the abnormal shadow detection processing based on the magnitude of the correlation value. Then, the medical image to be compared with the correlation is sequentially shifted back and forth to determine whether or not the medical image is to be subjected to the abnormal shadow detection processing.
This relates to a process for selecting an unnecessary image from a large number of medical images, and focuses on a predetermined region in the medical image, for example, a lung field region, and the first medical Extract the image. Normally, the medical image having the largest predetermined area is located in the middle of a large number of images, so that a large number of medical images exist before and after the first medical image. Therefore, a predetermined area is correlated between the preceding and succeeding medical images, and it is determined whether or not the predetermined area exists based on the correlation. If the predetermined area exists, it is set as a processing target. By shifting this process back and forth and sequentially, there is an image that cannot be correlated. In this case, the image before and after the image is excluded from the target of the process as not including the predetermined region. I did it. As a result, an image to be processed can be extracted from a large number of images.

According to a twentieth aspect of the present invention, there is provided an image diagnosis support apparatus for extracting only pixels belonging to a pixel value range corresponding to a type of a shadow to be determined from a medical image or the medical image. A multi-level converting means for performing predetermined image processing on the image to generate a multi-level image; and at least one of the multi-level image, the medical image and the medical image to be determined based on one of the multi-level image Extraction means for executing the above-described discrimination processing to extract a lesion candidate shadow that is a lesion candidate, and a medical image before abnormal shadow detection processing and a medical image including the lesion candidate shadow extracted by the extraction means are similar. Display means for displaying a medical image including non-abnormal shadows side by side. This is to simultaneously display a medical image containing a cancer shadow and a medical image containing a similar non-cancer shadow which is not a cancer shadow but is close to the medical image at the same time as the medical image before the abnormal shadow detection process. By
The doctor can make an appropriate decision on the shadow of the unknown medical image while referring to the shadow of the medical image.

[0026]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS 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 a hardware configuration of the entire image diagnosis support apparatus to which the present invention is applied. This image diagnosis support apparatus includes, for example, a plurality of tomographic images (such as CT images) collected for a target portion of a subject by an X-ray CT apparatus or the like.
Based on the displayed focus candidate shadows extracted, etc.,
Among the extracted lesion candidate shadows and the like, those with high certainty were narrowed down and displayed. Also, an image is displayed during these processes. This image diagnosis support apparatus includes a central processing unit (C) that controls the operation of each component.
PU) 40, a main memory 42 storing a control program of the entire apparatus, a magnetic disk 44 storing a plurality of tomographic image data and programs, a display memory 46 for temporarily storing image data for display, This display memory 46
CRT display 48 as a display device for displaying an image based on image data from the camera, mouse 50 for operating a soft switch on the screen, and controller 52 thereof
, A keyboard 54 having keys and switches for setting various parameters, a speaker 57, and a common bus 56 for connecting the above components. In this embodiment, the case where only the magnetic disk 44 is connected as a storage device other than the main memory 42 is shown. MO)
A drive, a ZIP drive, a PD drive, a DVD drive, or the like may be connected. Furthermore, it can be connected to various communication networks 58 such as a LAN (local area network), the Internet, and a telephone line via a communication interface so that image data can be exchanged with another computer, a CT device 59, or the like. It may be. The exchange of image data is
A medical image diagnostic apparatus capable of collecting a tomographic image of a subject, such as an X-ray CT apparatus or an MRI apparatus, may be connected to the communication network 58 such as the LAN.

Hereinafter, an example of the operation of the image diagnosis support apparatus of FIG. 1 will be described with reference to the drawings. FIG. 2 is a diagram illustrating an example of a main flow executed by the image diagnosis support device. FIG.
CPU 40 operates according to this main flow.
FIG. 3 is a diagram showing how a CT image is processed by this main flow. FIG. 4 is a diagram showing an example of a display screen on the CRT display 48. This main flow is started when the operator inputs a patient name to be subjected to lesion candidate extraction and display processing in the column of patient name on the display screen of FIG. 4 and clicks a calculation button. Hereinafter, the details of the main flow will be described in the order of steps.

[Step S80] The CPU 40 reads from the magnetic disk 44 the CT image 20 (FIG. 3 (a1)) of the patient corresponding to the patient name shown in FIG. [Step S81] The CPU 40 performs multi-level processing on the organ to be diagnosed from the read CT images, and generates a multi-level image as shown in FIG. 3 (b1). Details of the multi-value processing will be described later. [Step S82] The CPU 40 executes an optimum detection process corresponding to the region of the organ to be diagnosed or the type of the organ.
Determines the type of the part or the organ, etc., and proceeds to step S
It is determined whether to proceed to step S83 or step S83. [Step S83] The CPU 40 performs various kinds of image processing on the multi-valued image of FIG. 3 (b1), narrows down lesion candidate shadows, and selects shadows that are lesion candidates, that is, abnormal shadows 2.
2 is detected. This abnormal shadow detection process detects the abnormal shadow 22 based only on the multi-valued image generated in step S81 without using the original CT image. The details will be described later. By performing the abnormal shadow detection processing based on the multi-valued image as in this embodiment, it is possible to reduce the time required for a computer operation or the like and reduce the load of the operation processing. [Step S84] The CPU 40 determines whether the CT in FIG.
Various image processing is performed on the image and the multi-valued image of FIG. 3B1 to narrow down lesion candidate shadows and detect shadows that are lesion candidates, that is, abnormal shadows 22.

Steps S83 and S84
Is displayed in parallel on the CRT display 48 next to the CT image 20 of FIG. 3 (a1) on the CRT display 48 as shown in FIG. When the synthesizing button shown in FIG. 4 is clicked, the in-determination image 24 is superimposed on the CT image 20 and displayed accordingly. The display content of the image 24 during determination is sequentially changed according to the process of processing the data of the multi-valued image (that is, according to the extraction stage of the lesion candidate shadow). When the number of extracted abnormal shadows detected by the abnormal shadow detection processing is larger than a predetermined number, the display may indicate "not possible to determine" and may end the processing. The result is recorded on the magnetic disk as needed. Details of the abnormal shadow detection processing will be described later.

[Step S85] The CPU 40 leaves those determined as abnormal shadows in step S83 or S84 as lesions and deletes those that are not. [Step S86] The CPU 40 determines whether or not the three-dimensional image formation button 3D in FIG. 4 is clicked, that is, whether the three-dimensional image formation flag is “1” or “0”.
If “1” (yes), the process proceeds to step S87,
If “0” (no), the process proceeds to step S88. It should be noted that the three-dimensional image configuration flag can be set to “1” or “0” by the operator arbitrarily clicking the three-dimensional image configuration button in FIG. [Step S87] The process in step S87 is executed when the determination in step S86 is yes. The CPU 40 starts a process of forming a three-dimensional image from a plurality of CT images near the abnormal shadow. The three-dimensional image forming process is performed in parallel with the process of step S88. However, after the three-dimensional image forming process is completed, the process proceeds to step S88 and proceeds to step S8.
8 may be executed.

[Step S88] The CPU 40 adds and displays color information in the CT image of FIG. A composite process is performed to display the extracted lesions or markers colored in the original image (CT image). FIG. 3 (a2) shows an example of a composite image in the case where an abnormal shadow is surrounded by the marker M. [Step S89] The CPU 40 determines whether or not the multi-function image display button has been turned on.
If es), the process proceeds to step S8A, and if not on (no), the process proceeds to step S8B. [Step S8A] Since the multi-function image display button is on, the CPU 40 displays the three-dimensional image formed in step S87 and the like. [Step S8B] The CPU 40 determines whether or not an instruction has been given by the operator to perform similar lesion candidate extraction and display processing on the image of another patient, and displays the image of another patient (yes). If determined to be, step S8
0, the same process is repeatedly executed, and when it is determined that the image of another patient is not displayed (no), the process proceeds to step S
Proceed to 8C. [Step S8C] The CPU 40 determines whether or not the operation of turning on the end button in FIG. 4 has been performed by the operator. The display or the multi-function image display is continued, and when it is determined that the display is turned on (yes), the process ends.

The multi-valued image processing in step S81 in FIG. 2 is performed based on the CT image 20 as shown in FIG. As shown in FIG. 3, this multi-valued image processing is performed as shown in FIG.
The result of calculating the standard deviation and the like of the original CT image 20 as shown in (a1) is subjected to a predetermined threshold value processing to be multi-valued. FIGS. 5 and 6 show steps S8 of FIG.
It is a flowchart figure which shows the detail of the multivalued image processing of one organ for diagnosis. Here, the most basic binarized image processing in the multi-level image processing will be described. Conventionally, one of the image processing methods for highlighting a shadow is to use each CT.
There is a method for taking the difference between images. For example, image size 51
The difference between the CT values of the pixels at the same address (x, y) is determined between two adjacent 2 × 512 CT images, and the difference between the CT values is stored in the address (x, y) of the memory. Thus, an emphasized image in which the shadow is emphasized is obtained. Also,
There is also a method using a standard deviation value (including a variance value).
These methods do not particularly highlight the vicinity of the boundary of the shadow, nor do they extract the boundary (edge) of the shadow or extract only the shadow. Therefore, in this embodiment, multi-valued image processing capable of extracting a shadow (especially near the boundary of the shadow) in the CT image and highlighting the shadow is adopted. FIG. 7 is a diagram for explaining this multi-valued image processing in principle. Hereinafter, the details of the main flow will be described in the order of steps.

[Step S11] The CPU 40 sets a specific area having a predetermined shape as an initial position on the CT image. That is, for example, as shown in FIG.
The square specific regions (small regions) 12A and 12B of about 10 pixels are set in the CT image 20 (tomographic image of the subject), and are set at the initial position of the upper left corner. This small area 1
When the coordinates of the center position of 2A and 12B are (X, Y), the coordinates (X, Y) are set to (0, 0), respectively. Note that, in FIG. 7, the small area 12A is
And the small area 12B is set so as to straddle the boundary (edge) of the shadow 16. The size of the small region is not limited to 10 × 10 pixels, and may be, for example, a rectangle other than a square, a rhombus, or a circle. When the center position and the center of gravity are different from each other, the center of gravity is given priority. [Step S12] The CPU 40 calculates an average value AV of the density values (CT values) in the small area. Average value A obtained
V indicates a high value when present in the shadow 15 as in the small area 12A in FIG. 7, a low value when the small area does not straddle the shadow, and indicates a low value in the shadow 16 as in the small area 12B. In the case of straddling, the values are almost in between. [Step S13] The CPU 40 determines whether or not the
The average value p (xp, yp) of the coordinates of the pixel whose density value is equal to or greater than the average value AV and the average value m (xm, ym) of the coordinates of the pixel whose density value is smaller than the average value AV are obtained. In the case of the small area 12A in FIG. 7, the average value pA,
The mA is substantially near the center of the small area 12A, and the coordinate positions of the two substantially coincide. On the other hand, in the case of the small region 12B, the average value pB is approximately near the center of the overlapping portion between the shadow 16 and the small region 12B, and the average value mB is approximately near the center of the non-overlapping portion between the shadow 16 and the small region 12B. The coordinates of both are separated.

[Step S14] The CPU 40 determines the coordinates (xp, yp) of the average value p and the coordinates (xm, y) of the average value m.
m) is obtained. In the case of the small area 12A in FIG. 7, since the average values pA and mA are the same value, the distance D is “0”. In the case of the small area 12B, the average value pB and the average value mB are apart from each other.
Becomes That is, the distance D tends to increase when the small area is located near the edge of the shadow, and to have a small value when the small area does not straddle the shadow. [Step S15] In order to make the above-mentioned tendency more conspicuous, in this step S15, the CPU 40 sets M as a moment M at the center coordinates (X, Y) of the small area based on the distance D obtained in step S14. = G · f
(D) is obtained. This moment M is a value related to (X, Y). For example, in a small area, when the number of pixels whose density value is equal to or more than the average value AV is Np, and the number of pixels whose density value is smaller than the average value AV is Nm, each moment obtained based on the following equation is obtained. M1 to M3 are defined as moments M in step S15. Moment M
1 is M1 = Np × Nm × D. The moment M2 is
It is assumed that M2 = Nor × D. (Nor is the larger of Np and Nm.) The moment M3 is M3 = conventional variance × D.
(However, D may be a value of about 1 to 3 raised to the power of δ.) In general, an operation including D is effective. Also, in a determination process described later, a calculation result including D for a lesion area can be used for the determination.

[Step S16] The CPU 40 sets the center coordinate X of the small area to move the small area in the X direction of the image.
Add 1 to. [Step S17] The CPU 40 determines the center coordinate X of the small area.
Is determined to be the maximum value (the position where the small area exceeds the right end of the image). If the value is determined to be the maximum value (yes), the process proceeds to step S17. Returns to step S12, and repeats the processing of steps S12 to S17 until the value of the center coordinate X becomes maximum. [Step S18] The CPU 40 proceeds to step S17.
Since it was determined that the center coordinate X of the small area was the maximum,
To return the small area to the left end of the image, the center coordinate X is returned to an initial value (usually 0). [Step S19] The CPU 40 adds “1” to the center coordinate Y of the small area in order to move the small area in the Y direction of the image. [Step S20] The CPU 40 determines the center coordinate Y of the small area.
Is determined to be the maximum value (the position where the small area exceeds the lower end of the image). Proceeding to S21, if it is determined that the value is not the maximum (no), the process returns to step S12, and the processes of steps S12 to S20 are repeated until Y reaches the maximum. Thus, the CP
U40 scans the small region from the upper left to the lower right of the CT image 20, and sequentially calculates the moment M at the center coordinate position.

A method of extracting pixels located near a shadow or a boundary of the shadow from each CT image 20 using the moment M obtained in this manner will be described with reference to the flowchart shown in FIG. [Step S21] The CPU 40 stores in advance a constant from a keyboard input by an operator or a magnetic disk 44 or the like as a threshold value for determining whether each pixel of the CT image 20 is a shadow or a boundary of the shadow. Read the constants and specify one of them as a constant. [Step S22] The CPU 40 sets the coordinates (X, Y) to (0, 0) in order to set the pixel to be determined (determined pixel) to the initial position of the upper left corner of the CT image 20.
Set to 0). [Step S23] The CPU 40 reads the moment M obtained in step S15 of FIG. 5 for a small area centered on the coordinates (X, Y) of the pixel to be determined.

[Step S24] The CPU 40 determines whether or not the read moment M is larger than the constant specified in step S21. If it is larger (yes), the process proceeds to step S25, where the smaller (no) ), The process jumps to step S26. [Step S25] When the CPU 40 determines that the moment M is larger than the constant in step S24, it means that the pixel to be determined corresponding to the coordinates (X, Y) corresponds to a shadow or a boundary of the shadow. Therefore, in this step, the coordinates (X, Y) are extracted and stored in the memory (the main memory 42 or the magnetic disk 44). That is, when the CPU 40 determines that the moment M is larger than the constant (yes) in step S24, the CPU 40 sets the coordinate (X, Y) to the high level “1” of the binarization, and conversely, the step S24. When it is determined in S24 that the moment M is smaller than the constant (no), the low level “0” of the binarization is set to the coordinates (X, Y). In this way, each coordinate is set to one of the low level “0” and the high level “1” and is binarized. By binarizing each coordinate in this manner, each coordinate can be represented by one bit, so that subsequent processing can be simplified. [Step S26] The CPU 40 adds 1 to the coordinate X to move the coordinate of the pixel to be determined in the X direction.

[Step S27] The CPU 40 determines whether or not the value of the coordinate X of the pixel to be determined is the maximum (the position beyond the right end of the image). If it is determined that the value is the maximum (yes), the CPU 40 proceeds to step S28. When it is determined that the maximum value is not reached (no), the process returns to step S23, and the processes of steps S23 to S26 are repeated until X reaches the maximum value. [Step S28] The CPU 40 proceeds to step S27.
Since the coordinate X of the pixel to be determined is determined to be the maximum, the coordinate X is set to “0” in order to return the pixel to be determined to the left end, and the coordinate Y is set to 1 in order to move the image to be determined in the Y direction. Add. [Step S29] The CPU 40 determines the coordinates Y of the pixel to be determined.
Is determined to be the maximum (position beyond the lower end of the image), and if it is determined to be the maximum (yes), the process is terminated
If it is determined that it is not the maximum (no), step S23
And the processing of steps S23 to S28 is repeated until Y reaches the maximum.

In this manner, the CPU 40 scans the pixel to be determined from the upper left to the lower right of the CT image 20 and determines whether or not the pixel is a shadow or a boundary of the shadow. By the above processing, the center point (X, Y) of the small area having the moment M larger than the constant, that is, the coordinate point of the pixel which is the shadow or the boundary of the shadow is sequentially stored in the memory (the main memory 42 or the magnetic disk 44). You. 5 and 6
In the above, the binarization of the low level “0” and the high level “1” has been described. For example, three constants C1, C2, and C3 are designated, and when the moment M is smaller than the constant C1, when the moment M is greater than the constant C1 and smaller than the constant C2,
When the CT image is smaller than the constant C3, the CT image can be quaternized by determining which one of the cases corresponds to the constant C3 or more. In the case of quaternary conversion, one pixel is represented by 2 bits. It should be noted that a plurality of constants can be designated in the same manner in the case of multi-value conversion to a number other than this, and multi-value conversion can be performed based on the constants.

FIG. 8 is a diagram showing the concept of how a shadow is extracted by the above-described method for extracting a pixel located at the boundary of the shadow or the shadow. As shown in FIG. 8A, a CT image 21 having a circular shadow in which the CT value is highest near the center of the shadow and gradually decreases in the radial direction. By executing the processing, the memory stores the shadow 22 of the multi-valued image in which the boundaries of the shadow are clear as shown in FIG. . Further, by increasing the constant specified in step S21, a ring-shaped shadow in which the shadow boundary 23 is enhanced as shown in FIG. 8C is extracted.
Therefore, by variously changing the constant specified in step S21, it is possible to extract only the boundary of the shadow or to extract the entire shadow. In addition, it is also possible to highlight the boundary of the shadow extracted in this way.

Using the multi-valued image generated by the above-described multi-valued image processing, the abnormal shadow detection processing in step S83 in FIG. 2 is performed. Further, using the multi-valued image and the CT image 20 as the original image, an abnormal shadow detection process in step S84 in FIG. 2 is performed. When the abnormal shadow detection processing is performed using only the multi-valued image as in step S83, a binarized image and a multi-valued image larger than this (for example, an octalized image or a 16-valued image) are used. It is desirable to carry out. In the following description, the binarized artigraphy and the CT image 20
A description will be given of a case in which the abnormality detection processing is performed using. When the abnormal shadow detection process is performed using only the multi-valued image as in step S83 in FIG. 2, it is needless to say that the CT image 20 can be similarly handled by replacing the CT image 20 with the multi-valued image. .

Hereinafter, the details of the abnormal shadow detection processing according to the present invention will be described. In this abnormal shadow detection processing, the shadow is binarized and extracted, and a radius is set to rotate around the center of gravity of the extracted binarization shadow. Determine the density difference on the tomographic image at a predetermined position as a function of the angle of the radial, further Fourier transform this function based on the magnitude of the frequency component or based on the magnitude of the frequency showing the peak of the frequency component, The shadow is distinguished into a focus candidate shadow and a normal shadow.

FIGS. 9 and 10 are diagrams conceptually showing the appearance of the abnormal shadow detection processing. FIG. 9A is an enlarged view showing a lung field portion of a CT image, and a shadow 1 is shown in the lung field.
The state where a exists is shown. FIG. 9 (b)
9A is extracted after the binarization processing, and shows a state where the binarized shadow 1b exists. FIG. 10A shows the original CT image of FIG.
It is the image which superimposed and shown the binarized extraction image of (b). Although stored in separate memory areas on the computer, the same coordinates are taken at the time of processing as shown in FIG. FIG. 10B is an enlarged view of a part of FIG. 10A, showing how the abnormal shadow detection process is performed.

The center (center of gravity) is determined based on the binarized shadow 1b in FIG. 10B, and the center O of the shadow 1b is set to C
The method is applied to the shadow 1a in the T image, and a moving radius 100 that rotates around the center O is set. Points b and c are set on the moving radius 100 by minute distances dR1 and dR2, respectively, from the intersection a of the outer circumference of the moving radius 100 and the binarized shadow 1b. The difference between the CT values of points b and c in the CT image (point b
CT value−CT value of point c) is obtained for each angle Θ, and the angle Θ
Is plotted on the horizontal axis and the difference value is plotted on the vertical axis.
A curve as shown in FIG. When a curve as shown in FIG. 11A is subjected to a Fourier transform, and a horizontal line represents a frequency f and a vertical axis represents a Fourier coefficient C, a line graph is created.
A graph as shown in (b) is obtained. A curve 110 in FIG. 11B showing a case of a cancer shadow is obtained by performing a Fourier transform on the curve in FIG. 11A. On the other hand, the curve 111 showing the case of the blood vessel shadow in FIG. 11B is obtained by performing the above-described processing on the shadow corresponding to the end of the blood vessel as shown in FIG. As is clear from FIG. 11 (b), the cancer shadow tends to be shifted to the high frequency side as compared with the shadow of the blood vessel terminal and the like. Therefore, the shift amount toward the high frequency side, for example, the shift of the peak position becomes the feature amount of the shadow. When it is determined that the peak position is on the higher frequency side than the predetermined value, the binarized shadow 1b is left as a lesion candidate shadow, and otherwise, the binarized shadow 1b is deleted from the lesion candidate shadow. The minute distances dR1 and dR2 described above
May be a predetermined constant or may be determined based on the size (major axis or minor axis) of the binarized shadow 1b. Further, the minute distance dR1 and the minute distance dR2 may have the same value or different values. In the drawing, an example is shown in which the minute distance dR1 is smaller than the minute distance dR2 by about 1/2.

In the line graph of FIG. 11B, the Fourier coefficient at the frequency f0 is C0 and the frequency f1
, The Fourier coefficient at the frequency f2 is represented by C2, the Fourier coefficients are represented by Ci, each frequency is represented by fi, and the absolute value | fi of the product thereof is represented by fi.
× Ci | and the sum Σ | fi × Ci |
The divided value ff is calculated by dividing by | Ci |. The discriminated value ff is expressed by the following equation. ff = Σ | fi × Ci | / Σ | Ci | Whether or not the shadow is a focus candidate shadow may be determined according to whether or not the discriminated value ff is smaller than a predetermined value. When the shadow is a shadow corresponding to the end of a blood vessel as shown in FIG. 13, the Fourier transform results in a large Fourier coefficient of a low-order frequency component. Conversely, the shading is shown in FIG.
In the case of the focus candidate shadow 1a as in (a), as a result of the Fourier transform, many high-order frequency components are included, and low-order frequency components are reduced. Therefore, it is possible to determine whether the shadow is a lesion candidate shadow or a blood vessel cross-sectional shadow based on the value of the determination target value ff. Note that a specific | fi × Ci | may be used as the determined value instead of the determined value ff.

In FIG. 10B, the points b and c are set to the
Although the case of setting above has been described, in FIG.
1 and c1 are set on a straight line 121 in a direction perpendicular to a tangent 120 to the outer peripheral line of the binarized shadow 1b on the intersection point a of the radial radius 100 and the outer peripheral portion of the binarized shadow 1b. That is, a radius 100 that rotates around the center O of the binarized shadow 1b and a tangent 120 on the intersection a of the outer periphery of the binarized shadow 1b are created, and a line 121 perpendicular to the tangent is created. On the vertical line 121, points b1 and c1 separated by minute distances dR1 and dR2 are set. Points b1, c1 in CT image
(CT value at point b1−CT value at point c1)
Is obtained for each angle 、, and the angle Θ is plotted on the horizontal axis, and the difference value is plotted on the vertical axis, a curve similar to that shown in FIG. 11A can be obtained. A line graph in which the frequency f is set and the vertical axis is the Fourier coefficient C is created, and a shadow is determined.

FIG. 13 is a diagram showing a CT image of a blood vessel portion and a binarized extraction region obtained by binarizing and extracting the CT image. is there. In the figure, the end of the blood vessel is extracted as a binarized extraction area 131 by the binarization processing. FIG. 13 shows the original C
The blood vessel of the T image and the binarized extraction region 131 are shown superimposed. Generally, a CT image is a sliced CT
When a blood vessel exists between images, that is, when a blood vessel exists continuously in an adjacent slice image, the density of the terminal portion of the blood vessel tends to be low. Therefore, FIG.
As shown in (b), the CT value on points b and c or the CT value on points b1 and c1 is obtained as shown in FIG. 12, and the density of point b (or point b1) is point c (or point c1). The ratio in the case where the concentration is smaller than the density is calculated. If the ratio is larger than a certain value, it is regarded as the end of the blood vessel and is excluded from the focus candidate shadow. That is, since the binarized extraction region 131 at the end of the blood vessel as shown in FIG.
The T value is smaller than the CT value of the blood vessel portion. At the location where the blood vessel is present, the density at point b (or point b1) is lower than the density at point c (or point c1), and the ratio is relatively large, corresponding to the size of the location where the blood vessel is present. As shown. Therefore, in the case of FIG. 13, the binarized extraction area 131 is regarded as the end of the blood vessel and is deleted. In this case, since it is not necessary to perform the above-described Fourier transform processing, the calculation time can be reduced.

In the above-described embodiment, the CT image is subjected to the binarized image processing shown in FIGS. 5 and 6 to remove the shadow of a blood vessel portion smaller than a predetermined value. This processing is to extract a shadow having a predetermined number of pixels or less in the horizontal (X-axis) or vertical (Y-axis) direction and delete it. Relatively large vascular shadows that do not meet this condition were not removed. Even in the case of such a shadow of a blood vessel part, a shadow of a blood vessel 141 as shown in FIG. 14A or a relatively large lesion candidate shadow 151 and blood vessel shadows 152 to 154 as shown in FIG. Are overlapped, the blood vessel shadows 141 and 152
154 could not be removed. Therefore, in this embodiment, the blood vessel shadow 14 shown in FIG. 14A or FIG.
1, 152 to 154 were cut and removed.

First, the CT image is subjected to the binarized image processing shown in FIGS. 5 and 6 to remove the shadow of a blood vessel portion smaller than a predetermined value. After this processing, a temporary center of gravity O is determined for the shadow in FIG. 14A or 15A. The method of obtaining the position of the center of gravity is performed by using various methods described in the previously proposed application. While rotating the moving diameters 145 and 155 in the directions of arrows 146 and 156, the center of gravity O of the moving diameters 145 and 155
, And the minimum value (radial minimum length) Rmin and the maximum value (radial maximum length) Rmax thereof are determined. Here, the minimum radius Rmin and the maximum radius Rm
ax denotes a pixel as a unit of its size. Radial minimum length R
When min is determined, a value obtained by multiplying the minimum length Rmin by a constant a and adding a constant b thereto is set as a cutting radius Rc.
That is, the cutting radius is set to Rc = a × Rmin + b. FIG. 16A shows a curve corresponding to this equation. As is clear from FIG. 16 (a), the minimum radial length Rmi
The cutting radius Rc is determined by n. FIG. 16A shows the case where the cutting radius Rc changes linearly, but it may be changed non-linearly. FIG.
(B) is a value obtained by dividing the maximum radius Rmax by the minimum radius Rmin, that is, the maximum radius Rmax and the minimum radius R.
The cutting radius Rc is determined by converting the ratio with min to a non-linear curve. After the cutting major radius Rc is determined, as shown in FIG. 14B or FIG. 15B, cutting circles 147 and 157 are formed based on the cutting radius Rc and included in the cutting circles 147 and 157. The shading of the blood vessel portion is cut by removing the unshaded portion. FIG. 17A shows the result of the cutting process in FIG. 14B, and FIG. 17B shows the result of the cutting process in FIG. It is preferable that the constants a and b for determining the cutting radius Rc and the shape of the nonlinear curve in FIG. In the above embodiment, the cutting radius is set to the maximum radius Rmax or the minimum radius Rmi.
Although the case of determining based on n has been described, the cutting radius may be determined based on the area of the shadow.

FIG. 18 shows a process for detecting an abnormal shadow using the area of the shadow and another area relating to the shadow. In FIG. 18, the ratio of the total area of the entire lesion candidate shadow to the area of the concave portion formed at the edge of the shadow is determined, and the abnormal shadow detection processing is performed based on the ratio. FIG. 18 (a)
FIG. 17A and FIG. 18B are diagrams showing specific examples in the case where abnormal shadow detection processing based on the area ratio is performed on the shadow in FIG. 17B. The difference between the shadow shown in FIG. 18B and the shadow of the blood vessel region shown in FIG. 18A can be easily understood, and it can be understood that the shadow has a shape close to a circle. In the case of the blood vessel shadow shown in FIG. 18A, the lesion candidate shadow is obtained by using the area ratio RSA of the total area S1A of the areas s1a, s2a, s3a, and s4a of the concave portions 181 to 184 and the total area S2A of the shadow. Is determined. In the case of the shadow shown in FIG. 18B, the areas s1b, s2b, and s of the concave portions 185 to 189, respectively.
The lesion candidate shadow is determined using the area ratio RSB of the total area S1B of 3b, s4b, and s5b and the total area S2B of the shadow. The area ratio RSA, RSB is the total area S of the concave portions.
A ratio expression that simply shows the ratio between 1A, S1B and the total area S2A, S2B of the shadow: RSA = SLA / S2A, RSB =
SlB / S2B may be used, or the areas S1A,
The sum of S1B and areas S2A and S2B and areas S2A and S2
2A: RSA = S2A / (SLA +
S2A) and RSB = S2B / (S1B + S2B).

FIG. 18C shows a method for obtaining the area of the concave portion. Hereinafter, the details of the process of obtaining the area of the concave portion will be described with reference to the flowchart of FIG. [Step S821] As shown in FIG. 18C, the CPU 40 selects two points pl and p2 on the contour line of the shadow as a pair, and selects both points with a straight line. Here, only the first one is selected as a pair. [Step S822] A point p is assumed that moves on the straight line connecting the two points pl and p2 from one point p1 toward the other point p2 by a fixed length. The CPU 40 determines whether or not the point p exists in the extraction area (s2) every time the point p moves by a certain length. If the determination result is yes (when the point p exists in the extraction area (s2)), step S
Proceeding to 824, if the determination is no, step S82
Proceed to 3. [Step S823] Since the point p does not exist on the extraction area (s2), the CPU 40 records a specific value (for example, “5”) in that part. [Step S824] The CPU 40 determines whether or not the point p has moved on a straight line connecting the points pl and p2. If the result is yes (the movement has been completed), the process proceeds to step S825. By the processing of steps S822 to S824, a specific value (for example, 5) is recorded in an area other than the extraction area (s2) while the point p moves from the point p1 to the point p2. [Step S825] The CPU 40 sets the point pl to a fixed point,
When the point p2 is set as the movement point, it is determined whether or not the movement point p2 has moved on all the contour lines of the extraction area. CP
When the point p2 is a fixed point and the point p1 is a moving point, the U40 determines whether the moving point p1 has moved on all the contour lines of the extraction area. If the determination result is no (if the movement of the movement point has not been completed), the process returns to step S821, and the same processing is performed for the next two points. If the determination is yes, the process proceeds to step S826.

[Step S826] The CPU 40 determines the area (SLA, S1) of the area where the specific value (for example, 5) is recorded.
B) is obtained. These areas S1A and S1B are the areas of the concave portions. [Step S827] The CPU 40 calculates the areas SlA, S1
B and area ratio RSA, R of area S2A, S2B of extraction region
Obtain SB. [Step S828] The CPU 40 determines that the area ratios RSA, R
It is determined whether SB is larger than a predetermined value. If the determination is yes (large), the process proceeds to step S829; if the determination is no (small or equal), the process proceeds to step S82A. [Step S829] The area ratio RSA in step S828
Since it is determined that is larger than a certain value, the extracted shadow is highly likely to be a blood vessel shadow. Therefore, the CPU 40
The shadow in FIG. 18A is deleted from the lesion candidate shadow. [Step S82A] The area ratio RSB in step S828
Is determined to be equal to or less than a certain value, the extracted shadow is likely to be a lesion candidate shadow. Therefore, the CPU 40 stores information such as the coordinate position using the shadow in FIG. 18B as a lesion candidate shadow.

FIG. 20 shows another embodiment in which an abnormal shadow is detected using the radial length of the shadow, the area of the shadow, and other areas related to the shadow. FIG.
In (a), the ratio between the maximum length and the minimum length of the radial length of the binarized extracted shadow is plotted on the horizontal axis, and the ratio of the total area of the entire lesion candidate shadow to the ratio of the area of the concave portion formed at the edge portion of the shadow is plotted vertically. An abnormal shadow detection process is performed as an axis. That is, a value (percent value) obtained by dividing the total area of the concave portions by the total area of the shadow is defined so as to be a function of the ratio of the length of the maximum radius to the minimum radius of the binarized extracted shadow. However, the lesion candidate shadow is determined based on whether the region corresponds to a non-cancer side or a cancer side with respect to a curve serving as a determination reference. In FIG. 20B,
The abscissa represents the minimum radius and the ordinate represents the variance (including the standard deviation) of the radius, and the abnormal shadow detection process is performed. That is, the variance value (including the standard deviation value) of the radial length is defined so as to be a function of the minimum length of the radial length of the binarized extracted shadow, and this variance value is determined with respect to a curve serving as a criterion. Thus, the focus candidate shadow is determined based on whether the region corresponds to the non-cancer side or the cancer side.

FIG. 21 is a diagram showing an example of a case where the above-mentioned parameter serving as a criterion depends on the position of the organ in the image. FIG. 21A is a diagram illustrating an example of a case where a lung field region is divided. In this division process, after the entire lung field 211 is extracted by binarization extraction processing or the like, the lung field 211 is cut inward by one pixel from the edge of the lung field 211. Lung field 21
The determination as to whether or not it is the edge of 1 is made, for example, considering a small area of 3 × 3 pixels, and any of the pixels of this small area is determined to be a lung field
In the case where the area deviates from the eleven extraction areas, the center of the small area of 3 × 3 pixels may be regarded as an edge and deleted. By this division processing, the lung field 211 becomes the lung field area 212 after the first deletion processing. Hereinafter, by similarly deleting, lung field regions 213 to 215 after the second to fourth deletion processes are formed. When the lung field regions are formed in this way, parameters are assigned to the respective lung field regions 211 to 215. In the present embodiment, as an example, the cutting radius r1 is set in the lung
2, a cutting radius r2, and a lung field area 213 a cutting radius r3,.
・ Allocate sequentially. As a result, a parameter (cutting radius) corresponding to the position is assigned to the entire lung field area as shown in FIG. When a cutting radius is assigned as a parameter to a lung field region, a case of a branched blood vessel 141 as shown in FIG. 14A or a relatively large lesion candidate shadow 151 and a blood vessel shadow 152 as shown in FIG. When 154 overlaps, the cutting radius is determined according to where the center of gravity position O is located in the lung field regions 211 to 215. That is, the position of the center of gravity of the blood vessel shadow 141 in FIG.
When it is located at 2, the cutting radius is r2, and the blood vessel shadow 141 is cut by this cutting radius r2.
Further, the lesion candidate shadow 151 and the blood vessel shadow 1 in FIG.
When the centroid position O of 52 to 154 is located in the lung field region 213, the cutting radius is r3, and the lesion candidate shadow 151 and the blood vessel shadows 152 to 154 are cut by the cutting radius r3. Although the case where the parameter is determined depending on where the center of gravity position O is located in the lung field region has been described, the parameter is determined depending on which lung field region the shadow exists in, or the ratio of the presence of the shadow. The parameter may be calculated according to.

FIG. 22 is a diagram showing another embodiment of the cutting process. Here, the cutting process is performed based on the area of the concave portion formed at the edge of the shadow. FIG.
FIG. 22A shows a case where cutting processing is performed on a branched blood vessel shadow 220, and FIG. 22B shows a case where cutting processing is performed on a shadow 225 in which a lesion candidate shadow and a blood vessel shadow are combined. Is shown. First, for each of the shadows 220 and 225, the area s1a and s2 of the concave portion by the above-described processing.
a, s3a, s1b, s2b, s3b, and s4b are obtained. Compare the size of each area and arrange them in descending order. In the case of FIG. 22A, s1a>s2a> s3a, and
In the case of FIG. 22B, s1b>s2b>s3b> s4
b. The top two centroids (which may be near the centroid) with the larger area are obtained. In the case of FIG. 22 (a), the centers of gravity O1a, O1a of the upper two areas s1a and s2a having the larger areas are shown.
2a is obtained, and a shadow 22 is formed by a line segment O1a-O2a connecting the two.
Cut 0. Similarly, in the case of FIG. 22B, the center of gravity O1b of the upper two areas s1b and s2b,
O2b is obtained, and a shadow 2 is formed by a line segment O1b-O2b connecting the O2b.
Cut 25. Specifically, the line segments O1a-O2a, O
The cutting process is performed by writing the same value as the background value of the binarized shade into the memory in 1b-O2b. As a result, the shadow of the blood vessel as shown in FIG.
Since the combination of the focus shadow and the blood vessel as in 2 (b) is appropriately cut, the abnormal shadow detection processing to be performed later is appropriately performed. Note that, of the cut shadows, only the one to which the center (center of gravity) O of the original shadows 220 and 225 belongs may be left, and the other shadows may be deleted. Further, in the above-described embodiment, a straight line is used as a line segment. . Alternatively, the shadows 220 and 225 may be cut at an arc centered on the center (center of gravity) O.

FIG. 23 shows an embodiment in which an average value image of a shadow is created, and an abnormal shadow is detected using the distance from the edge of the average value image to the edge of the shadow using the radius. It is. A predetermined area (eg, 3 ×
(3,5 × 5, 7 × 7, 9 × 9,..., Etc.)
Is applied to calculate the average value of the binarized shadow 231, and the calculated average value is used as the average value of the center of the predetermined area to create the average value image 232. The relationship between the binarized shadow 231 and the average image differs depending on the size of the predetermined area. Then, the moving radius 233 having a predetermined length is rotated by about 1 degree from 0 degree to 360 degree with the angle θ around the center of the shadow 231 as the rotation center. At this time, a distance dL between the edges when the moving radius 233 intersects the edge of the shadow 231 and the edge of the average image 232 at each angle is obtained. This distance dL is the shade 231
Is positive when the edge of the image exists outside the average image 232, and is negative when the edge exists inside the average image 232. Based on this length dL, as shown in FIG. 23 (b), a curve is drawn with the angle θ as the horizontal axis and the distance dL between both edges as the vertical axis, and based on this curve, the shadow 231 is a focus shadow. Determine if there is. In the case of a shadow representing a branch of a blood vessel as shown in FIG. 23 (a), there are many negative portions in the distance dL as shown in FIG. 23 (b). However, in the case of a lesion shadow, this negative portion is present. Since it is small or does not exist, it can be determined by this. The curve is Fourier-expanded, and based on the Fourier-expanded result, a line graph is created with the horizontal axis as the frequency and the vertical axis as the Fourier coefficient, and the shadow is a focus shadow based on the line graph. It may be determined whether or not. In the above-described embodiment, the case has been described where the center of rotation of the binarized shadow 231 is set as the center of rotation. However, the center of rotation of the average image 232 may be set as the center of rotation.

FIGS. 24 to 26 are views showing an example of an image selection process for selecting a CT image to be subjected to the above-mentioned abnormal shadow detection process. In CT images, MR images, ultrasonic images, etc., the number of measurement images tends to increase more and more,
A large number of medical images are generated. Therefore, in this embodiment, unnecessary ones are removed from the medical images, and an image to be subjected to the abnormal shadow detection processing is selected. FIG. 24 is a diagram schematically showing how to select from a large number of CT images. FIGS. 25 and 26 are flowcharts showing an example of this image selection processing. Hereinafter, the details of the image selection processing will be described with reference to these flowcharts. [Step S251] For a plurality of (Nmax) CT images as shown in FIG. 24, the CPU 40 sets the flag flg [N] to the high level “1” for the image number N where the lung field region exists. That is, in this process, for example, an image N = 1 and N = Nmax in which the lung field regions 243 to 246 do not exist among a plurality of images as shown in FIG.
Is extracted. The details of the flag setting process will be described later. [Step S252] Since the flag setting process in step S251 is completed, “1” is set in the image number register N. [Step S253] It is determined whether or not the flag flg [N] of the image number register N is at a high level “1”. In the case of ()), the process proceeds to step S255. [Step S254] Perform candidate point detection processing, that is, the above-described abnormal shadow detection processing. [Step S255] Based on the extracted result, a display / storing process of displaying a CT image with a marker as described above and storing the extracted result in a memory or a magnetic disk is performed. In addition, C which did not perform the candidate point detection process
Processing of T images may also be performed. [Step S256] The image number register N is incremented by "1", and the process proceeds to the next step. [Step S257] It is determined whether or not the value of the image number register N has reached the maximum value.
Returning to step S253, steps S253 to S2 are performed until the image number register N reaches the maximum value.
Step 57 is repeated.

FIG. 26 is a diagram showing details of the flag setting process in step S251 of FIG. Hereinafter, the details of the flag setting process will be described with reference to a flowchart. [Step S261] The CPU 40 obtains an image number Npeak that gives the maximum lung field area for a plurality of (Nmax) CT images as shown in FIG. In the figure, the lung field area 245 gives the maximum. [Step S262] Reset the variable k to “0” and set the flag flg [Np corresponding to the image number Npeak.
eak] is set to a high level “1”. [Step S263] A variable k is assigned to this image number Npeak.
Is added to the lung field area of the image number Npeak + k, and the image number Npe is obtained by adding the variable k + 1 to the image number Npeak.
Correlate with the ak + k + 1 lung field area. [Step S264] It is determined whether or not the correlation value between the lung field area of the image number Npeak + k and the lung field area of the image number Npeak + k + 1 is greater than a certain value. Proceed to
If it is smaller (no), the process proceeds to step S266. [Step S265] The flag flg [Npeak + k + 1] corresponding to the image number Npeak + k + 1 is set to the high level “1”. [Step S266] A low level “0” is set to the flag flg [Npeak + k + 1] corresponding to the image number Npeak + k + 1. [Step S267] The variable k is incremented by “1” and the process proceeds to the next step. [Step S268] It is determined whether or not the variable k has reached the maximum value. If the variable k has reached (yes), the next step S26
9 if not reached (no), step S26
3 and the above processing is repeated until the variable k reaches the maximum value. [Step S268] It is determined whether or not the variable k has reached the maximum value. If the variable k has reached (yes), the next step S26 is performed.
9 if not reached (no), step S26
3 and the above processing is repeated until the variable k reaches the maximum value.

[Step S269] The above-described step S2
In the processing from step 63 to step S268, the image number is moved in the plus direction with respect to the image number Npeak, and the correlation between the two images is obtained. Next, the image number is moved in the minus direction with respect to the image number Npeak. Both images were correlated. Therefore, the variable k is reset to “0”. [Step S26A] Lung field area of image number Npeak + k obtained by adding variable k to image number Npeak, and image number Npeak obtained by adding variable k−1 to image number Npeak
+ K-1 is correlated with the lung field area. That is, the image number is moved in the minus direction with respect to the image number Npeak, and the correlation is obtained. Here, since the variable k is a minus value, + k of the image number means minus, and as a result, the image number moves from Npeak in the minus direction. [Step S26B] It is determined whether or not the correlation value between the lung field area of the image number Npeak + k and the lung field area of the image number Npeak + k-1 is larger than a certain value. Proceed to step S26C,
If smaller (no), the process proceeds to step S26B. [Step S26C] The high-level "1" is set to the flag flg [Npeak + k-1] corresponding to the image number Npeak + k-1. [Step S26D] A low level “0” is set to the flag flg [Npeak + k-1] corresponding to the image number Npeak + k-1. [Step S26E] The variable k is decremented by "1", and the process proceeds to the next step. [Step S26F] It is determined whether or not the variable k has reached the maximum value in the negative direction. If the variable k has reached (yes), the process is terminated, the process proceeds to step S252 in FIG. 25, and if not (no), Returning to step S26A, the above-described processing is repeated until the variable k reaches the maximum value. Here, the case of processing an axial image has been described, but it goes without saying that the same can be applied to other sagittal images and coronal images.

FIG. 27 shows a tomographic image on the CRT display 4.
FIG. 8 is a diagram showing an example of a case of displaying on a display 8; Normally, when displaying a tomographic image, one tomographic image is displayed on the CRT display 48, and the display is sequentially switched, or a plurality of tomographic images are simultaneously displayed on the CRT display 48, and the display is performed for each of a plurality of images. Switching. When a plurality of tomographic images are displayed continuously as described above, instead of switching and displaying the images, shift display is performed as shown in FIG. That is, in FIG. 27A, the first tomographic image 271 and the second tomographic image 272 are displayed. When the next button is clicked with the mouse pointer in this state, the first tomographic image 271 and the second tomographic image 272 shift to the right and a new third tomographic image to the left as shown in FIG. The tomographic image 273 is displayed. Hereinafter, as shown in FIGS. 27C and 27D, the tomographic images 272 to 275 are sequentially shifted and displayed on the right side in response to the click of the next button. in this way,
By performing shift display, when diagnosing with the center image, the images before and after it are always displayed on both sides, so the doctor refers to the images on both sides and makes an appropriate judgment on the center image. Will be able to do it.

FIG. 28 shows a tomographic image on the CRT display 4.
FIG. 9 is a diagram showing a modified example in a case where the image is displayed on a screen 8; Magnetic disk 4 storing a plurality of tomographic image data and programs
4 calculates in advance image data 281 including a cancer shadow and a feature amount of an image which is not cancer but includes a shadow similar to cancer (similar non-cancer shadow), and calculates the feature amount and the image. The associated image data 282 is stored. When an unknown CT image in which it is not determined whether the shadow is a focus candidate shadow or not is displayed on the left side of the CRT display 48, if the lower right reference image button is clicked, the unknown CT image is displayed. On the right side of the image, from among the image data 281 including the cancer shadow, a cancer shadow having the closest calculated value using the feature value is selected and displayed, and the image data 282 is also displayed.
, A similar non-cancer shadow having the closest calculated value using the feature value is selected and displayed. This allows the doctor to make an appropriate decision on the shadow of the unknown CT image while referring to the cancer shadow or similar non-cancer shadow. Note that the calculated value using the feature amount includes, for example, a Mahalanobis distance, a Euclidean distance, a neural network, and the like.

When judging whether or not to be a lesion candidate using the above various feature amounts, even if processing such as statistical processing or neural network is employed in the middle, finally, threshold processing or the like is performed. It may be necessary to determine the exact parameters of It goes without saying that the conventional means in such a case is based on so-called "learning" in which the parameters are made more accurate from the images obtained every day. The CPU 40 may use the extraordinary amount, the variance value, or the standard deviation value as an input value of a Mahalanobis distance, a Euclidean distance, a neural network, or the like, and make a determination using the result. As a result, it is possible to provide a feature amount for discriminating a lesion shadow and a processing procedure using the feature amount, which are not available in the related art. In the above-described embodiment, the case where cutting is performed using a cutting circle has been described, but cutting may be performed using an ellipse or another shape. In this case, it is preferable to make the major axis direction of the shadow coincide with the major axis direction of the ellipse. In the above-described embodiment, a two-dimensional image has been described as an example.
It goes without saying that the same effect is obtained even when the same processing is performed on the three-dimensional image space.

[0063]

As described above, according to the image diagnosis support apparatus of the present invention, when automatically determining a lesion candidate or the like from a medical image by a computer, shadows having different sizes and shapes can be uniformly handled. In addition, there is an effect that a computer operation can be completed in a short time. Further, according to the image diagnosis support apparatus of the present invention, there is an effect that a shadow which is considered to be an extracted lesion candidate can be displayed easily and instantly so as to be identifiable.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a hardware configuration of an entire lesion candidate extraction and display device to which the present invention is applied.

FIG. 2 is a diagram showing a main flow executed by a lesion candidate extraction display device;

FIG. 3 is a diagram showing how a CT image is processed according to the main flow of FIG. 2;

FIG. 4 is a view showing an example of a display screen on the CRT display of FIG. 1;

FIG. 5 is a detailed flowchart showing the first half of the multilevel image processing in step S81 in FIG. 2;

FIG. 6 is a detailed flowchart showing the latter half of the multilevel image processing in step S80 of FIG. 2;

FIG. 7 is a diagram for explaining in principle the multilevel image processing of FIGS. 5 and 6;

FIG. 8 is a view showing a concept of how a shadow is extracted by a method of extracting a pixel located at a shadow or a boundary of the shadow.

FIGS. 9A and 9B are diagrams conceptually showing a state of abnormal shadow detection processing, wherein FIG. 9A is a partially enlarged view of a CT image, and FIG. 9B is a view showing an extracted shadow after binarization;

10A and 10B are diagrams conceptually showing a state of abnormal shadow detection processing, in which FIG. 10A is a diagram in which a CT image and an extracted image are superimposed, and FIG. 10B is an enlarged view in which a part of FIG.

11A shows a curve plotted with the angle Θ plotted on the horizontal axis and the difference between the CT values of points b and c in the CT image plotted on the vertical axis, and FIG. 11B shows a curve as shown in FIG. Diagram showing a line graph in which Fourier transform is performed and the horizontal axis is frequency f and the vertical axis is Fourier coefficient C

FIG. 12 is a diagram showing a modification of FIG.

FIG. 13 is a diagram showing a CT image of a blood vessel portion and a binarized extraction area obtained by binarizing and extracting the CT image.

FIG. 14 is a diagram conceptually showing a state of a cutting process for a branched shadow of a blood vessel.

FIG. 15 is a diagram conceptually showing a state of a cutting process when a relatively large lesion candidate shadow and a blood vessel shadow overlap each other.

FIG. 16 is a diagram showing a method of determining a cutting radius when performing a cutting process.

17A illustrates a result of the cutting process in FIG. 14B, and FIG. 17B illustrates a result of the cutting process in FIG.

FIG. 18 is a diagram showing a state in which an abnormal shadow is detected using the area of the shadow and another area related to the shadow.

FIG. 19 is a flowchart showing details of a process for obtaining the area of a concave portion;

FIG. 20 is a diagram showing another embodiment in which an abnormal shadow is detected and processed using the radial length of the shadow, the area of the shadow, and other areas relating to the shadow.

FIG. 21 is a diagram showing an example of a case where a parameter serving as a criterion depends on the position of an organ in an image.

FIG. 22 is a diagram showing another embodiment of the cutting process.

FIG. 23 is a diagram showing an example in which an average value image of a shadow is created, and an abnormal shadow is detected and detected using the distance from the edge of the average value image to the edge of the shadow using a radial radius.

FIG. 24 is a diagram showing an example of image selection processing for selecting a CT image to be subjected to abnormal shadow detection processing.

FIG. 25 is a flowchart showing an example of this image selection processing.

FIG. 26 is a diagram showing details of the flag setting process in step S251 in FIG. 25;

FIG. 27 is a view showing an example of displaying a tomographic image on a CRT display 48;

FIG. 28 is a view showing a modification in the case where a tomographic image is displayed on the CRT display 48;

[Explanation of symbols]

 1a, 15, 16: shadow 1b: binarized shadow 12A, 12B: small area 20, 21, original CT image 22: lesion candidate shadow 24: image in process 30: medical image 31, 32, 33: circle ( Marker) 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 57 Speaker 58 Local area network 59 Other Computer or CT device 100, 130, 145, 155 ... Radius 131 ... Binarized extraction area 147, 157 ... Cutting circle

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI theme coat ゛ (reference) G06T 7/60 150 A61B 5/05 380 F term (reference) 4C093 AA22 AA26 CA21 CA31 FF08 FF09 FF12 FF16 FF17 FF20 FF23 FF28 FG04 4C096 AB50 AD12 AD14 DA01 DC12 DC15 DC16 DC20 DC21 DC24 DC28 DD08 5B057 AA09 BA03 BA07 CA08 CA12 CA16 CB08 CB12 CB16 CC02 CD03 CE09 CG05 DA08 DC02 DC06 DC30 DC32 5L096 AA06 BA06 BA13 EA16 FA63 FA63 FA63 FA23

Claims (20)

[Claims] 1. A predetermined image process is performed on a medical image or a medical image to be determined created by extracting only pixels belonging to a pixel value range corresponding to the type of a shadow to be determined from the medical image. Multi-valued means for creating a multi-valued image by performing the processing, detecting a center or a center of gravity of the shadow based on the multi-valued image, and defining a radius of a predetermined length with the vicinity of the center or the center of gravity of the shadow as a reference point. The multi-valued image, the medical image or the medical image for discrimination is rotated on a shadow, and a predetermined value is set inside and outside the multi-valued image with reference to the intersection of the radius and the edge of the multi-valued image. A pixel value of a shadow in the medical image or the medical image for discrimination at at least two locations separated by a distance is sampled, and whether or not the shadow is a lesion candidate shadow is determined based on a difference value between the pixel values of the two locations. Extraction means for determining whether Image diagnosis support apparatus characterized by a. 2. The multi-valued image according to claim 1, wherein the extracting means is configured to include at least a predetermined distance between the inside of the multi-valued image and the outside of the multi-valued image with reference to an intersection of the radius and an edge of the multi-valued image. An image diagnosis support apparatus, wherein pixel values of a shadow in the medical image or the medical image for a discrimination target at two locations are sampled. 3. The multi-valued image according to claim 1, wherein the extracting means includes a line perpendicular to a tangent line formed at an edge of the multi-valued image so as to include an intersection of the radius and the edge of the multi-valued image. And sampling the pixel values of the shadows in the medical image or the medical image for discrimination at at least two places separated by a predetermined distance inside and outside the multi-valued image with reference to the intersection. Image diagnosis support device. 4. The method according to claim 1, wherein the extraction means creates a difference value waveform based on a difference value between the two pixel values with the rotation angle of the radius as a horizontal axis. An image diagnosis support apparatus characterized in that it is determined whether or not the shadow is a lesion candidate shadow based on a difference value waveform. 5. The method according to claim 4, wherein the extracting means performs a Fourier transform on the difference value waveform, and determines whether or not the shadow is a focus shadow based on a result of the Fourier transform. Image diagnosis support device. 6. The image according to claim 5, wherein the extraction means determines whether or not the shadow is a lesion shadow based on a magnitude relationship between frequency components as a result of the Fourier transform. Diagnosis support device. 7. The method according to claim 5, wherein the extraction unit determines whether or not the shadow is a lesion shadow based on a magnitude of a frequency indicating a peak of the frequency component as a result of the Fourier transform. Characteristic image diagnosis support device. 8. The multi-valued image according to claim 4, wherein the extraction means is configured such that, in the difference value waveform, a pixel value located inside the multi-valued image is smaller than a pixel value located outside the multi-valued image. An image diagnosis support apparatus, wherein a ratio of a case is obtained, and based on whether the ratio is greater than a predetermined value, the shadow is regarded as a terminal of a blood vessel and is excluded from a lesion candidate shadow. 9. The method according to claim 1, wherein the extracting unit detects a center or a center of gravity of the shadow based on the multi-valued image,
The multi-valued image, the medical image or the medical image for determination or by rotating on the shadow in the medical image for determination as a reference point near the center or the center of gravity of the shadow, the straight line and the multi-valued image, The minimum value of the length of a straight line portion that intersects the shadow in the medical image or the medical image for discrimination is obtained, and a cutting radius is obtained based on the minimum value, which is included in a cutting circle formed by the cutting radius. An image diagnosis support apparatus, wherein the shadow is left, other shadows are removed, and it is determined whether or not the remaining shadow is a lesion candidate shadow. 10. The method according to claim 1, wherein:
The center or the center of gravity of the shadow is detected based on the multi-valued image, and a straight line having a predetermined length with the vicinity of the center or the center of gravity of the shadow as a reference point is included in the multi-valued image, the medical image or the medical image for discrimination target. Rotate on the shade of, the straight line and the multi-valued image, the medical image or the medical image for the determination of the minimum length and the maximum value of the length of the straight line portion that intersects the shadow in the medical image for determination, the minimum value Determine the cutting radius based on the ratio of the value and the maximum value, leave the shadow included in the cutting circle formed by the cutting radius, remove other shading,
An image diagnosis support apparatus for determining whether a remaining shadow is a lesion candidate shadow. 11. The method according to claim 9, wherein:
Detecting the center or center of gravity of the shadow based on the multi-valued image, determining the distance from the center or the center of gravity to the edge of the shadow over the entire periphery of the shadow, and determining the distance over the entire periphery. An image diagnosis support apparatus for determining a variance value or a standard deviation value of the image and determining whether the shadow is a lesion candidate shadow based on the variance value or the standard deviation value and the minimum value. 12. The method according to claim 10, wherein the extraction means obtains an area of the shadow area and an area of a concave area formed at an edge of the shadow area, and calculates an area of the shadow area and an area of the concave area. An image diagnosis support apparatus, wherein a ratio to an area is determined, and it is determined whether or not the shadow is a lesion candidate shadow based on the determined ratio and a ratio between the minimum value and the maximum value. 13. The method according to claim 10, wherein said extraction means obtains an area of said shadow area and an area of a concave area formed at an edge of said shadow area, and An image diagnosis support apparatus, wherein the vicinity of the center of gravity of two large areas is connected by a straight line or a curve, and the shadow area is cut using the straight line or the curve. 14. The image diagnosis support apparatus according to claim 13, wherein it is determined whether or not the shadow cut using the straight line or the curve is a lesion candidate shadow. 15. The shadow according to claim 14, wherein, for the shadow cut using the straight line or the curve, a post-cut shadow that does not include the center of the pre-cut shadow or the center of gravity that does not include the center of gravity is deleted. An image diagnosis support apparatus for determining whether or not a shadow is a focus candidate shadow. 16. The method according to claim 1, wherein the extracting unit includes:
Dividing a predetermined area of the multi-valued image into a plurality of areas, assigning different parameters to each of the divided areas, and performing the division where a center or a center of gravity of a shadow detected based on the multi-valued image is located An image diagnosis support apparatus, comprising: performing a cutting process of the shadow or a process of determining whether or not the shadow is a lesion candidate shadow using a parameter assigned to an area. 17. The method according to claim 1, wherein:
An average value image is created based on the multi-valued image, a center or a center of gravity of the shadow is detected based on the multi-valued image or the average value image, and a predetermined length is set using the center of the shadow or the vicinity of the center of gravity as a reference point. Is rotated on the shade in the multi-valued image and the average value image, the difference value of the length of a straight line portion that intersects the straight line and the shade in the multi-valued image and the average value image is calculated. Calculating a difference value waveform based on the difference value with the rotation angle of the straight line as the horizontal axis, and determining whether the shadow is a lesion candidate shadow based on the difference value waveform. Image diagnosis support device. 18. A processing image selecting means according to claim 1, wherein unnecessary processing is removed from said medical images before being processed by said multi-value processing means, and an image to be subjected to abnormal shadow detection processing is selected. An image diagnosis support device, comprising: 19. The medical image processing apparatus according to claim 19, wherein the processing image selection unit extracts a first medical image having a largest area of a predetermined region from the medical image, Taking the correlation of the predetermined area between the medical image, determine whether the medical image before and after is to be subjected to the abnormal shadow detection process based on the magnitude of the correlation value,
An image diagnosis support apparatus, wherein a medical image to be compared with the correlation is sequentially shifted back and forth to determine whether or not the medical image is to be subjected to the abnormal shadow detection processing. 20. A predetermined image process is performed on a medical image or a medical image to be determined created by extracting only pixels belonging to a pixel value range corresponding to the type of a shadow to be determined from the medical image. Performing multi-valued image processing to generate a multi-valued image by performing at least one or more discriminating processes based on any one of the multi-valued image, the medical image, and the medical image to be discriminated. Extracting means for extracting a lesion candidate shadow which is a candidate of the medical image, and a medical image including the lesion candidate shadow extracted by the extracting means and a medical image including the similar non-abnormal shadow extracted by the extracting means. An image diagnosis support device comprising: display means for displaying the images side by side.