JP4934786B2 - Knee joint diagnosis support method, apparatus and program - Google Patents

Description

  The present invention relates to a method for diagnosing a knee joint using an X-ray imaging apparatus. In particular, the present invention relates to a diagnosis support method, apparatus, and program that are effective for the diagnosis of knee osteoarthritis.

  Osteoarthritis of the knee (knee OA) is a case in which the cartilage of the knee joint is worn away, and is a disease that is common in middle-aged and older people. With the arrival of an aging society, the diagnosis, treatment, and eventually prevention of this knee osteoarthritis are drawing attention as important issues.

  Now, the state of the cartilage of the knee joint can be examined very clearly by a technique using MRI or the like. However, this MRI has a problem that the price of the apparatus is expensive and the measurement also takes time.

  Therefore, at present, the diagnosis of knee osteoarthritis (knee OA) is mainly based on X-ray images, but the diagnostic criteria are based on the experience of doctors and the like, and clear diagnostic criteria have been established. Absent.

  Therefore, if it is possible to provide objective diagnostic information for diagnosing knee osteoarthritis using an X-ray image, which is a general diagnostic technique, a doctor can perform a more accurate diagnosis using this X-ray image. Conceivable.

Prior Art Documents A description will be given of what is considered to be representative among prior arts relating to diagnosis of knee joint disease, imaging of X-ray images, and the like.

  Patent Document 1 below discloses a method and apparatus for determining knee osteoarthritis. In this method, acceleration is detected by an acceleration sensor attached to the tibia and the heel during walking, the peak time of the acceleration signal is measured, and the impact transmission time is obtained. A technique for determining knee osteoarthritis from this impact transmission time is disclosed.

  Patent Document 2 below discloses a joint surgery support information calculation method that supports diagnosis of arthropathy, selection and positioning of an artificial joint, and the like. In this method, it is described that approximate bone three-dimensional data of a patient is created based on an X-ray image, and information on the positional relationship of the bone is obtained based on this data.

  Patent Document 3 below discloses an imaging table that can obtain an X-ray imaging image having a diagnosis of a knee joint or a foot.

Japanese Patent Application No. 2004-261525 JP 2003-144454 A Japanese Patent Application Laid-Open No. 11-089831

  The present invention has been made in view of such a background, and provides a diagnosis support method and apparatus that provides effective quantitative diagnosis information to a doctor in order to objectively diagnose a knee joint. The purpose is to do.

  In view of the above-described problems, the present inventor has developed a computer-aided interpretation system and has developed a technique for quantifying the degree of narrowing of the joint space on an X-ray image. By performing such quantification, it is possible to provide objective diagnosis information to the doctor and make the doctor's diagnosis and judgment more accurate.

  That is, the present invention employs the following means.

  In the text, X-ray image data obtained by imaging the knee joint is used, and the femur direction is referred to as “upper” and “upper”, and the tibial direction is referred to as “lower” and “lower”.

  (1) In order to solve the above-mentioned problem, the present invention provides a method for assisting diagnosis of a knee based on X-ray image data of a knee. And an interval detection step for obtaining a lower side, a lateral range where the cartilage of the knee exists is obtained from the X-ray image data, and an area surrounded by the obtained upper side and lower side in the lateral direction range And an area acquisition step of acquiring the area of this region, and the interval detection step performs an edge extraction process on the X-ray image data to obtain a joint surface of the femur of the knee joint Detecting the femoral edge indicating the upper side, performing edge extraction processing on the X-ray image data, detecting a trailing edge indicating the rear edge of the joint surface of the tibia, and the X-ray image Performing a maximum luminance extraction process on the data, detecting a leading edge indicating a leading edge of the joint surface of the tibia, and obtaining a midline between the leading edge and the trailing edge as the lower side; The area obtaining step obtains a regression line of the lower side, obtains a range in which the lower side is located below the regression line as a range in the left-right direction where the cartilage of the knee exists, and the upper side in the range This is a knee joint diagnosis support method characterized in that the area of a region surrounded by the lower side is obtained.

(2) In the knee joint diagnosis support method according to (1), the present invention performs vertical edge extraction processing on the knee X-ray image data, obtains the center of gravity of the obtained edge, And a reference region determining step of determining a predetermined range centered on the center of gravity as a reference region, wherein the processing described in (1) is performed in the reference region.

  (3) In order to solve the above-mentioned problem, the present invention provides a method for assisting diagnosis of a knee based on the X-ray image data of the knee, and the side surface of the femur and the side surface of the tibia in the X-ray image data. And a side detection step for obtaining a femoral center curve from the obtained femoral side surface, obtaining a tibial center curve from the obtained tibial side surface, and further, a regression line of the femoral center curve And a FTA acquisition step of obtaining an FTA which is an intersection angle of the two regression lines, and obtaining a regression line of the central curve of the tibia.

  (4) In the knee joint diagnosis support method according to (3), the present invention performs vertical edge extraction processing on the knee X-ray image data, obtains the center of gravity of the obtained edge, A horizontal line segment intersects the femur at a predetermined distance above the center of gravity, and the arranged line segment is used as a reference area on the femur side, and a horizontal line segment is positioned at a predetermined distance below the center of gravity. A reference region determining step that is arranged so as to intersect with the tibia and uses the arranged line segment as a reference region on the tibia side, and the side surface detection step includes the intersection of the reference region on the femur side and the femur, The edge which is a pixel having a large luminance change in the X-ray image data is detected to detect the side surface of the femur, and the luminance change in the X-ray image data is determined from the intersection of the reference region on the tibia side and the tibia. Track edges that are large pixels A knee diagnosis support method characterized by detecting the side surface of the tibia.

  (5) Further, in the knee joint diagnosis support method according to any one of (1), (2), (3), and (4), the present invention performs left and right determination of the X-ray image data of the knee. A knee joint diagnosis support method characterized by including a left / right determination step.

  (6) In the knee joint diagnosis support method according to any one of (1), (2), (3), and (4), the present invention provides noise that removes noise from knee X-ray image data. A knee joint diagnosis support method comprising a removal step.

  (7) In order to solve the above-described problem, the present invention provides an apparatus for supporting knee diagnosis based on the X-ray image data of the knee. From the X-ray image data, the upper side of the knee cartilage existing region is obtained. And an interval detection means for obtaining the lower side, and a range in the left-right direction in which the cartilage of the knee exists from the X-ray image data, and a region surrounded by the obtained upper side and lower side in the range in the left-right direction And an area acquisition means for acquiring the area of this area, and the interval detection means performs an edge extraction process on the X-ray image data to obtain a joint surface of the femur of the knee joint Means for detecting the femoral edge as the upper side, means for performing edge extraction processing on the X-ray image data, detecting a trailing edge indicating the rear edge of the joint surface of the tibia, and the X-ray image data Maximum brightness extraction Means for performing processing and detecting a leading edge indicating a leading edge of the joint surface of the tibia, and determining a middle line between the leading edge and the trailing edge as the lower side, and obtaining the area The means obtains a regression line of the lower side, obtains a range in which the lower side is located below the regression line as a lateral direction range where the cartilage of the knee exists, and is surrounded by the upper side and the lower side in this range. The knee joint diagnosis support apparatus is characterized in that the area of the determined region is obtained.

  (8) In order to solve the above-described problems, the present invention provides an apparatus for assisting diagnosis of a knee based on the X-ray image data of the knee, and the side surface of the femur and the side surface of the tibia in the X-ray image data. And a side detection means for determining, a center curve of the femur from the determined side of the femur, a center curve of the tibia from the determined side of the tibia, and a regression line of the center curve of the femur And a FTA acquisition means for obtaining a regression line of the central curve of the tibia and obtaining an FTA which is an intersection angle of the two regression lines.

  (9) In order to solve the above problems, the present invention provides a program for causing a computer to operate as a device for assisting diagnosis of a knee based on the X-ray image data of the knee. From the interval detection procedure for obtaining the upper side and the lower side of the knee cartilage existing region, and from the X-ray image data, obtain a range in the left-right direction in which the knee cartilage is present, The area surrounded by the obtained upper and lower sides is regarded as an area where cartilage is present, and an area acquisition procedure for acquiring the area of this area is executed, and the interval detection procedure performs edge extraction on the X-ray image data. Processing, detecting a femoral edge indicating the joint surface of the femur of the knee joint as the upper side, applying edge extraction processing to the X-ray image data, A procedure for detecting a trailing edge that indicates the leading edge of the tibial joint surface by performing a maximum luminance extraction process on the X-ray image data, and the leading edge edge and the trailing edge. A step of obtaining a middle line with an edge as the lower side, wherein the area obtaining step obtains a regression line of the lower side, and a range in which the lower side is located below the regression line is defined as a cartilage of the knee. Is obtained as a range in the left-right direction, and the area of the region surrounded by the upper side and the lower side in this range is obtained.

  (10) In order to solve the above problems, the present invention provides a program for causing a computer to operate as a device for assisting diagnosis of a knee based on the X-ray image data of the knee. A lateral detection procedure for determining the lateral surface of the femur and the lateral surface of the tibia, a central curve of the femur from the lateral surface of the determined femur, and a central curve of the tibia from the lateral surface of the determined tibia, Further, a regression line of the femoral center curve is obtained, a regression line of the tibia center curve is obtained, and an FTA acquisition procedure for obtaining an FTA which is an intersection angle of the two regression lines is executed. It is a program.

  As described above, according to the present invention, effective information regarding knee osteoarthritis can be provided to a doctor based on an X-ray image. Therefore, it is possible to more accurately make a diagnosis of knee osteoarthritis and judge the degree. In addition, since the degree of knee osteoarthritis can be objectively determined, more appropriate prevention and lifestyle guidance are possible, medical costs can be reduced, and the patient's QOL (Qualituy of Life) can be improved. It contributes.

  The best mode for carrying out the present invention will be described below with reference to the drawings.

1. Overall System FIG. 1 shows an overall configuration diagram of a system described in the present embodiment. This is a system in which digitized X-ray image data (also referred to as XP) is stored in the image database 10 and the stored XP is analyzed by the analysis server 12.

  It is preferable to download a conventional DICOM image in the image database 10, and it is also preferable to store image data obtained by digitizing analog XP data separately.

  The analysis server 12 processes these images and outputs an objective index related to the diagnosis of knee osteoarthritis. As a result, useful information is obtained when a doctor makes a diagnosis, which contributes to an objective diagnosis.

  Information obtained in the present embodiment is the area of the portion of the joint where the cartilage exists and the FTA.

2. Processing Flow FIG. 2 shows a flowchart representing the flow and outline of the processing of the present embodiment.

  The processing shown here is processing for X-ray image data (XP) converted into digital data, and is processing performed by the analysis server 12. More specifically, an X-ray image data analysis program is stored in a storage device in the analysis server 12, and the processor of the analysis server 12 executes the X-ray image data analysis program, so that Processing is executed.

2.1 Noise Removal First, in step S2-1, a median filter is applied to the X-ray image data three times to remove noise. An explanatory diagram of this processing is shown in FIG. This median filter uses a 3 × 3 median filter, but may be another 4 × 4 or 5 × 5.

  In FIG. 3, an explanatory diagram of the operation of the median filter is shown. As shown in this explanatory diagram, the median filter is a filter that substitutes an intermediate value (median) of pixels in a predetermined template around the target pixel for the target pixel, and is widely used for noise removal and the like. . In the example of FIG. 3, nine pixel values of 3 × 3 around the pixel having the pixel value “12” are extracted, and an intermediate value (median) “6” of these values is substituted for the target pixel. This process is performed on each pixel to remove noise.

  Although the median filter is used in the present embodiment, various conventionally known filters can be used.

  This step S2-1 corresponds to a preferred example of the noise removal step in the claims.

  The processing of step S2-1 is also executed by the analysis server 12 using a computer program. Since median filters and other noise removal programs have been widely known, such means (programs) are created. This is easy for a person skilled in the art.

  Further, X-ray image data to be processed is taken out from the image database 10, and various known formats can be used for the image data. In addition to DICOM images and bitmap images (such as BMP) widely used in the medical field, various formats that can be handled by a computer can be used.

2.2 Left / Right Determination Next, in step S2-2, the left / right of the X-ray image data is checked. An explanatory diagram of this processing is shown in FIG. In this left / right check process, first, the edge of the X-ray image data is extracted. Then, the width of the two bones is acquired. The two bones are the tibia and the ribs. Then, the right foot or the left foot is determined from the thickness of the two bones. In the example of FIG. 4, a thick bone (ie, tibia) is shown on the left, and a thin bone (ie, ribs) is shown on the right, so that it can be determined that it is the left foot.

  In the present embodiment, imaging is performed such that the femur is positioned above the X-ray image data and the tibia and ribs are disposed below. In other words, “upper” refers to the direction of the femur, that is, the lumbar side, and “lower” refers to the direction of the tibia and fibula, that is, the foot side.

  In addition, photographing is performed by irradiating a human body with X-rays from the front and placing an X-ray film on the back of the human body.

  From such a positional relationship, if the tibia is on the left and the tibia is on the right, it can be determined that it is X-ray image data of the left foot.

  Such image processing is also executed by a program stored in the analysis server 12 and a processor that executes the program. Since processing for detecting thick / thin areas is a technique conventionally known in the field of image processing, it is easy for those skilled in the art to create such a program. In particular, in the present embodiment, it has been found that the tibia and the rib always exist almost below the X-ray image data. Therefore, the portion is scanned in the horizontal direction, and the portion of the bone is determined from the change in luminance and the value. It is very easy to judge the thickness of the program and to create the program.

2.3 Determination of Reference Area Next, in step S2-3, the reference area of the knee joint is determined. This reference area means an inspection area for cartilage, and the thickness (degree of wear) of the cartilage in this reference area is calculated. An explanatory diagram showing the state of this processing is shown in FIGS.

  Now, only the edge component in the vertical direction is extracted from the X-ray image data using a Roberts filter. This is shown in FIG. The center of gravity of the set of pixels determined to be the edge is obtained, and a predetermined range around the center of gravity is determined as the reference region (see FIG. 6).

FIG. 5 shows an edge portion extracted by the Roberts filter. An edge part is a set of pixels recognized as an edge. Since the center of gravity of the set of pixels is considered to be substantially the center of the joint portion, a certain range around the center of gravity is used as a reference area. This is shown in FIG.
Note that the Roberts filter is a filter used for, for example, contour extraction. An explanatory diagram of the Roberts filter employed in the present embodiment is shown in FIG. As shown in this figure, the Roberts filter of this embodiment is a filter that performs a predetermined calculation using three pixels near the lower right 3 of the target pixel and substitutes the result into the target pixel. Specifically, as shown in FIG. 10, first, the difference du between the target pixel and the lower right pixel thereof, and the difference dv between the pixel right next to the target pixel and the lower pixel of the target pixel are obtained. Ask. Next, the value f is obtained by taking the square root of the sum of the square of du and the square of dv. The obtained f is set as a new value of the target pixel.
That is, the new value of each pixel is a value having the maximum value of the deviation of the neighboring pixel on the upper left and the deviation of the other two neighboring pixels. As a result, the Roberts filter highlights the pixels in which the luminance change occurs along the diagonal axis and extracts the contour.

  Now, FIG. 6 is a diagram for explaining a reference area employed in the present embodiment. In this figure, an area represented as two squares is a reference area used when examining the joint interval. The upper and lower parallel straight lines are reference areas for calculating FTA (femorobital angle).

  These are areas where a predetermined range centered on the center of gravity is considered to be appropriate from experience. The calculation of the joint interval and the calculation of the FTA will be described in detail later.

  The calculation of the center of gravity in the determination process of the reference area is conventionally performed widely in image processing and other technical fields, and it is easy for those skilled in the art to create a program for calculating the center of gravity. Also, the process of setting a predetermined area centered on the center of gravity as a reference area only sets the reference area at a predetermined distance corresponding to the center of gravity, so creating such a program Easy. Further, it is preferable to store parameters such as the reference area set in the middle of the processing, intermediate data, and the like in the storage means in the analysis server 12. Therefore, it is easy for those skilled in the art to configure the means for performing the reference region determination.

2.4 Interval detection (Cartilage area detection 1)
Next, in step S2-4, the joint interval is detected. Specifically, the lower edge of the femur and the upper edge of the tibia are obtained. The distance between the lower edge of the femur and the upper edge of the tibia is one of important diagnostic information.

  The state of these measurements is shown in FIG. As shown in FIG. 7, this interval detection process is executed in the following flow. An enlarged view of the X-ray image data of FIG. 7 is shown in FIG.

  (1) First, the upper bone, that is, the joint surface edge of the femur is detected. Similar to the processing described in step S2-3, this basically detects the luminance change in the vertical direction of the X-ray image data and extracts the edge. Among these edges, the uppermost edge is recognized as an edge indicating the joint surface of the femur (referred to as a femur edge in the text) (see FIG. 8).

  (2) Next, the lower bone, that is, the joint surface of the tibia is detected. In general, in X-ray photography, the joint surface of the lower tibia is slightly tilted from the perpendicular to the imaging surface, so the X-ray image data of the knee joint shows an image as if the joint surface was observed obliquely. It is.

  Accordingly, when the edge is detected in the same manner as the femur for the tibia, the edge (rear edge) on the side far from the viewpoint of the joint surface can be detected. This is called a trailing edge in the text (see FIG. 8).

  Next, a portion of the tibia of the X-ray image data having a maximum luminance value is traced and extracted. This is considered to be the edge on the side close to the viewpoint of the joint surface of the tibia. This is called a leading edge in the text, and the extracted edge is called a leading edge in the text (see FIG. 8).

  The midline of the two edges of the leading edge and the trailing edge taken out in this way is taken, and this midline edge is regarded as an edge indicating the joint surface of the tibia (see FIG. 8).

  (3) In this way, a femoral edge indicating the joint surface of the femur and a midline edge indicating the joint surface of the tibia are obtained. Then, the distance between the two edges (the femur edge and the midline edge) is recognized as the joint interval. That is, cartilage is thought to exist during this time.

  In the present embodiment, one of the physical quantities to be finally obtained is the area of the portion where the cartilage portion exists, and these two edges are the upper side and the lower side of the existing portion. Then, the area is obtained in the next step S2-5.

  In this way, the joint interval is detected. Prior to all the processes, all the images are subjected to a median filter and a smoothing filter.

  In the processing described above, processing for extracting an edge and processing for tracing and extracting a pixel having a maximum luminance value are techniques that have been conventionally known for image processing. It is easy to create a program that does. In addition, since the process of obtaining the middle line of two edges is one of basic image processes, it is easy to create a program for executing the process. Therefore, it is easy for those skilled in the art to create a means for executing the interval detection process.

2.5 Area acquisition (Cartilage area detection part 2)
Next, in step S2-5, the area of the part where the cartilage exists, which is one of the effective information obtained in the present embodiment, is obtained. The upper side and the lower side (that is, the boundary between the upper direction and the lower direction) of this part (region) were obtained in step S2-4 described above.

  Therefore, next, the left and right boundaries of this portion (region) are obtained. An explanatory diagram showing the state of this processing operation is shown in FIG. An enlarged view of the photograph of FIG. 9 is shown in FIG.

  First, regression analysis is performed on the edge indicating the joint surface of the tibia obtained above to obtain a regression line. The obtained regression line is compared with the edge indicating the joint surface of the tibia, and the edge is located below the regression line (the direction in which the tibia exists, conversely, the direction of the femur is called the upward direction) The range is recognized as a portion where the cartilage exists, and the left and right sides of the range are regarded as the left side and the right side of the portion where the cartilage exists.

  Specifically, a vertical line passing through the intersection of the regression line and the edge is drawn, and this vertical line becomes a straight line that defines the left and right boundaries.

  In general, there are two regions where the edge is downward from the regression line, and the left side and the right side (the above-described vertical line) of the two parts are respectively obtained. This area is represented in white in the photograph of FIG. In FIG. 10, the left and right ranges of the two regions are indicated by arrows.

  A knee joint generally has an anterior cruciate ligament, a posterior cruciate ligament, etc. at the center, and a meniscus or the like that cushions the left and right sides. In the present embodiment, since the amount of cartilage in the portion serving as the cushion is to be inspected, as described above, the areas are obtained for the left and right portions excluding the central portion.

  The area of the portion where the cartilage thus obtained is obtained and the value is output. The area can be obtained by various conventional algorithms. It can be obtained by counting the number of pixels belonging to a predetermined region on the image or by other methods. Therefore, it is easy for those skilled in the art to configure a means for executing this “area acquisition” operation.

2.6 Side detection (FTA detection part 1)
Next, in step S2-6, the side surfaces of the femur and tibia are detected in order to calculate the FTA using the linear reference region among the reference regions obtained in step S2-3.

  First, a detection point (a point where the amount of change in luminance is large) is found on the upper linear reference region, and is identified as one point on the side surface of the femur. From this detected detection point, the side of the femur can be obtained by tracking (also called “trace”) a point where the luminance change is large. Since there are two left and right side surfaces, two detection points are also found, and the left and right side surfaces can be obtained by tracking the edges starting from each of the two points. In addition, since it is image processing, although it is a side "surface", it cannot be overemphasized that it becomes a music "line" showing a side surface on an image.

  This is shown in the explanatory diagram of FIG.

  By the way, the portion close to the joint in this side surface has a slightly irregular shape and is not used in the calculation of FTA. In order to remove the portion close to the joint, the side surface is differentiated twice, and the portion after the portion where the sign of the twice-differentiated value is changed is removed. By such removal, it is possible to take out only the side surface of the portion that coincides with the extension direction of the femur.

  In this way, the side of the femur is obtained. Although only the femur is shown in FIG. 11, the same process is performed on the tibia to obtain the side surface of the tibia.

In this way, the side detection process is generally
(1) Detection of initial detection point for tracking (trace) using reference area (2) Tracking edge part from initial detection point (3) Removal of irregular part Since these are processes conventionally known in the field of image processing, it is easy for those skilled in the art to configure a program for performing these processes. By executing such a program by the analysis server 12 (a processor thereof), a means for executing the side surface detection process can be configured.

2.7 FTA acquisition (FTA detection part 2)
Center curve Next, in step S2-7, the center curves of the obtained left and right side surfaces are obtained.

  Since obtaining an intermediate curve between two curves is conventionally performed widely in image processing, it is easy for those skilled in the art to perform such processing. A regression line of the center curve thus obtained is obtained, and a straight line (femoral axis) indicating the extension direction of the femur is obtained. This is shown in FIG.

  Note that the “center curve” is sometimes called a “middle line”. Both mean intermediate curves between the two curves.

  Further, not only the femur but also the same process is performed on the tibia to obtain a central curve of the tibia. A regression line of the central curve thus obtained is obtained to obtain a straight line (tibia axis) indicating the extension direction of the tibia (see FIG. 12).

  In such a process of detecting the femoral axis and the tibia axis, the image is subjected to a median filter and a smoothing filter.

Detection of FTA The angle formed by the femoral axis and the tibia axis thus obtained is determined, and the FTA (femoral tibia angle) is detected.

  Since obtaining an angle formed by a line segment on an image is a general process, it is easy for those skilled in the art to construct such a program, and the analysis server 12 (processor) analyzes this program. By executing, a means for executing the FTA acquisition process can be configured.

In this way, the process of FTA acquisition is roughly
(1) Calculation of center curve (2) Flow of obtaining the angle between the center curve of the femur (femoral axis) and the center curve of the tibia (tibial axis). Since the processing is known in the field of image processing, it is easy for those skilled in the art to configure a program for performing these processing. By executing such a program by the analysis server 12 (processor thereof), means for executing FTA calculation processing can be configured.

Summary As described above, according to the present embodiment, the area of the area where the cartilage exists and the FTA can be automatically calculated from the X-ray image data of the knee. Therefore, it provides useful judgment materials in the diagnosis of a doctor and contributes to a more objective diagnosis.

Other Examples In the above-described example, as shown in FIG. 1, the image database 10 in which X-ray image data is stored and the analysis server 12 that executes actual image processing and the like are configured separately. Of course, it is also preferable to implement with one computer.

1 is an overall configuration diagram of a system described in the present embodiment. It is a flowchart showing the flow and outline | summary of the process of this Embodiment. It is explanatory drawing which shows the mode of noise removal. It is explanatory drawing explaining the processing operation | movement of right-and-left determination. It is explanatory drawing showing the mode of the determination process of a reference area. It is explanatory drawing which shows the specific example of a reference area. It is explanatory drawing showing the operation | movement which detects the space | interval of a joint. It is an enlarged view of the photograph of FIG. It is explanatory drawing showing the mode of the area acquisition (area detection operation | movement) of the cartilage in a joint. It is an enlarged view of the photograph of FIG. It is explanatory drawing showing the mode of the operation | movement of the area acquisition of the cartilage in a joint. It is explanatory drawing showing operation | movement of the detection of a center curve and the detection of FTA. It is explanatory drawing of a Roberts filter.

Explanation of symbols

10 Image database 12 Analysis server

Claims (5)

In a method for supporting knee diagnosis based on knee X-ray image data,
An interval detecting step for obtaining an upper side and a lower side of the knee cartilage existing region from the X-ray image data;
From the X-ray image data, a lateral range where the knee cartilage exists is obtained, and an area surrounded by the obtained upper and lower sides in the lateral range is regarded as an area where cartilage exists. An area acquisition step for acquiring the area of the region;
Including
The interval detection step includes
Performing an edge extraction process on the X-ray image data, detecting a femoral edge indicating a joint surface of a femur of a knee joint as the upper side;
Performing an edge extraction process on the X-ray image data and detecting a trailing edge indicating a trailing edge of the joint surface of the tibia;
Performing a maximum brightness extraction process on the X-ray image data, and detecting a leading edge indicating a leading edge of the joint surface of the tibia;
Determining a middle line of the leading edge and the trailing edge as the lower side;
Including
The area acquisition step includes
Obtaining a regression line of the lower side, obtaining a range in which the lower side is located below the regression line as a lateral range where the cartilage of the knee exists, and a region surrounded by the upper side and the lower side in this range A method for assisting diagnosis of a knee joint, characterized in that an area of the knee is obtained. The knee joint diagnosis support method according to claim 1,
A reference region determination step of performing vertical edge extraction processing on the X-ray image data of the knee, obtaining a center of gravity of the obtained edge, and determining a predetermined range centered on the center of gravity as a reference region;
And performing the process according to claim 1 in the reference region. In a method for supporting knee diagnosis based on knee X-ray image data,
A side detection step for obtaining a side surface of the femur and a side surface of the tibia in the X-ray image data;
A femoral center curve is obtained from the obtained femoral side surface, a tibia center curve is obtained from the obtained tibia side surface, a regression line of the femoral center curve is obtained, and a tibia center curve is obtained. An FTA acquisition step of obtaining a regression line and obtaining an FTA that is an intersection angle of the two regression lines;
In the knee joint diagnosis support method characterized by including :
The X-ray image data of the knee is subjected to vertical edge extraction processing, the center of gravity of the obtained edge is obtained, and the horizontal line segment is arranged at a predetermined distance above the center of gravity so as to intersect the femur, Refer to the arranged line segment as a reference area on the femur side, arrange a horizontal line segment to intersect the tibia at a predetermined distance below the center of gravity, and use the arranged line segment as a reference area on the tibia side Region determination step,
Including
The side detection step includes
From the intersection of the reference region on the femur side and the femur, the edge that is a pixel having a large luminance change in the X-ray image data is tracked to detect the side surface of the femur,
A knee joint diagnosis supporting method for detecting a side surface of the tibia by tracking an edge which is a pixel having a large luminance change in the X-ray image data from an intersection of the reference region on the tibia side and the tibia. In an apparatus for supporting knee diagnosis based on knee X-ray image data,
An interval detecting means for obtaining an upper side and a lower side of the knee cartilage existing region from the X-ray image data;
From the X-ray image data, a lateral range where the knee cartilage exists is obtained, and an area surrounded by the obtained upper and lower sides in the lateral range is regarded as an area where cartilage exists. Area acquisition means for acquiring the area of the region;
Including
The interval detecting means includes
Means for performing edge extraction processing on the X-ray image data and detecting a femoral edge indicating a joint surface of a femur of a knee joint as the upper side;
Means for performing an edge extraction process on the X-ray image data and detecting a trailing edge indicating a trailing edge of the joint surface of the tibia;
Means for performing a maximum luminance extraction process on the X-ray image data and detecting a leading edge indicating a leading edge of the joint surface of the tibia;
Means for determining a middle line of the leading edge and the trailing edge as the lower side;
Including
The area acquisition means includes
Obtaining a regression line of the lower side, obtaining a range in which the lower side is located below the regression line as a lateral range where the cartilage of the knee exists, and a region surrounded by the upper side and the lower side in this range An apparatus for assisting diagnosis of knee joint, characterized in that the area of the knee is determined. In a program for operating a computer as a device for supporting knee diagnosis based on knee X-ray image data,
In the computer,
From the X-ray image data, an interval detection procedure for obtaining an upper side and a lower side of the knee cartilage existing region;
From the X-ray image data, a lateral range where the knee cartilage exists is obtained, and an area surrounded by the obtained upper and lower sides in the lateral range is regarded as an area where cartilage exists. An area acquisition procedure for acquiring the area of the area;
And execute
The interval detection procedure includes:
A procedure for performing edge extraction processing on the X-ray image data and detecting a femoral edge indicating a joint surface of a femur of a knee joint as the upper side;
A procedure of performing edge extraction processing on the X-ray image data and detecting a trailing edge indicating a trailing edge of the joint surface of the tibia;
Performing a maximum luminance extraction process on the X-ray image data, and detecting a leading edge indicating a leading edge of the joint surface of the tibia;
A procedure for determining a middle line between the leading edge and the trailing edge as the lower side;
Including
The area acquisition procedure includes:
Obtaining a regression line of the lower side, obtaining a range in which the lower side is located below the regression line as a lateral range where the cartilage of the knee exists, and a region surrounded by the upper side and the lower side in this range A program characterized by determining the area of a computer.

Images