KR20100125889A - Diganosis system of rheumatoid arthritis and it's method - Google Patents

Abstract

PURPOSE: A system and a method for diagnosing RA(Rheumatoid Arthritis) are provided to measure JSW(Joint Space Width) by automatically selecting ROI(Region Of Interest) from an X-ray. CONSTITUTION: A JSW(Joint Space Width) measuring module(110) measures a JSW from an X-ray image. A database(120) stores JSW information of each group from statistics about a normal group, and seropositive RA and seronegative RA groups. A diagnosis module(130) diagnoses RA by comparing the measured JSW with the JSW of each group stored in the database.

Description

Diagnosis System of Rheumatoid Arthritis And It's Method

The present invention relates to a rheumatoid arthritis (RA) diagnosis system and method, and more particularly, to automatically select a region of interest (ROI) from an X-ray image, and to determine a joint space width (Joint Space Width, The present invention relates to a rheumatoid arthritis diagnosis system and method for evaluating rheumatoid arthritis immediately upon treatment by measuring JSW) and comparing with databases of three groups of normal group, seropositive RA group and seropositive RA group.

Rheumatoid arthritis (RA) is an inflammatory autoimmune disease that causes cartilage destruction and erosion due to synovial inflammation, resulting in narrowing of joint gaps or deforming joints.

X-ray findings in subchondral bone tissue cause osteoporosis, and rheumatoid arthritis disease is reported in about 1% of the population, and it is a disease frequently occurring in women aged 30 to 50 years.

The cause of rheumatoid arthritis has not yet been clearly identified, but in addition to reports that about 60% of disease predisposition is due to genetic factors, it is known that there are various causes such as infection by microorganisms or environmental factors.

In general, Rheumatoid Factor (RF), anti cyclic citrullinated peptide antibody, hematological findings, rapid reactor protein, and radiological findings are used as diagnostic tests.

Since rheumatoid arthritis is generally regarded as early rheumatism within 2 years after onset and irreversible damage to articular cartilage occurs within this period, it is very important to start early treatment with early detection.

Various studies are in progress for the early detection of rheumatoid arthritis as described above, and it is necessary to develop a diagnostic system that can easily and easily determine rheumatoid arthritis using X-ray images.

The present invention has been made to solve the above problems, an object of the present invention is to automatically select the ROI (Region Of Interest) from the X-ray image, to measure the joint space width (JSW) The present invention provides a diagnostic system for rheumatoid arthritis that can evaluate rheumatoid arthritis immediately after treatment, compared to a database of three groups, a normal group, a seropositive test and a seropositive test.

Rheumatoid arthritis diagnostic system according to the present invention to achieve the above object is a system for diagnosing rheumatoid arthritis from an X-Ray image, measuring the joint space width (Joint Space Width, JSW) from the X-Ray image A database storing JSW information of each group extracted from statistics for the JSW measurement module, the normal group, the seropositive RA and the seronegative RA group, and the JSW measured from the JSW measurement module. Comprising a diagnostic module for diagnosing rheumatoid arthritis by comparing the JSW information for each group stored in the.

In the case of the hand, since there are 14 joints, up to 14 intervals (hereinafter, referred to as “JSW”) between the measured hand joints may be measured.

Here, the JSW measurement module includes an image preprocessing unit for extracting only a hand image from the X-ray image, and ROI (Region) from phalangeal region (PR) information of a finger extracted using a step wedge function. Of interest) is characterized in that it comprises a ROI measuring unit for automatically extracting and JSW measuring unit for measuring the JSW from the ROI region.

Here, the JSW measurement module is a module formed in software by a program. As described above, the image preprocessor, the ROI measurement unit, and the JSW measurement unit may be classified according to functions, but may be integrated into one software. Is self-explanatory.

More specifically, the image preprocessing unit removes background noise and patient information from the X-ray image, extracts only a pure hand image through binarization of an image blurred by diffusion, and The ROI measuring unit extracts only the PR information by using the step wedge function, extracts branch path of phalanx (BPP) information by using a method corresponding to an intermediate point from the PR information. Computing a profile plot of the extracted BPP and automatically extracting joint (DIP, PIP, MCP) positions using local minimum points, extracting ROI from the extracted joint positions, The JSW measurement unit is characterized in that for measuring the JSW from the ROI.

Here, the JSW measurement can automatically measure the distance between the joints directly using a ROI having a clear joint boundary, and can be measured by manually selecting both ends of the JSW on the ROI.

In addition, the diagnostic module analyzes the JSW measured from the JSW measurement module and the average JSW values and patterns of the normal group, the seropositive RA and the seronegative RA group stored in the database, and match the patterns. Judging the group is characterized by diagnosing rheumatoid arthritis.

Meanwhile, the method for diagnosing rheumatoid arthritis according to the present invention includes calculating average hand joint interval (JSW) information for each group of a normal group, seropositive RA, and negative RA group and storing them in a database. And measuring JSW from the X-Ray image through the JSW measurement module and diagnosing rheumatoid arthritis by comparing the measured JSW value with average JSW information for each group stored in the database.

In the measuring of the JSW, only the hand image is extracted from the X-Ray image, and a ROI (Region Of Interest) is automatically generated from the extracted joint region (PR) information using a step wedge function. Extraction, and measuring the JSW from the ROI region.

As described above, the rheumatoid arthritis diagnosis system and method according to the present invention automatically select ROI (Region Of Interest) from X-Ray images to measure JSW, normal group, seropositive RA group and serologic negative. The superior effect of assessing rheumatoid arthritis occurs immediately after treatment compared to the database for the 3 groups of the RA group.

Hereinafter, specific embodiments of the present invention will be described in detail with reference to the drawings.

1 is a schematic block diagram of a system for diagnosing rheumatoid arthritis according to a preferred embodiment of the present invention.

1, the rheumatoid arthritis diagnosis system according to the present invention is a JSW measurement module 110 for measuring JSW from the X-Ray image and the normal group, seropositive RA (seropositive RA) and negative RA (seronegative RA) Diagnostic module 130 for diagnosing rheumatoid arthritis by comparing the JSW information for each group extracted from the statistics for the group 120 and the JSW measured from the JSW measurement module and the JSW information for each group stored in the database It may be configured to include.

The present invention is characterized by using the information of the hand joint interval from the X-Ray image used directly in the clinic for the clinical use of the diagnosis of early rheumatoid arthritis.

To extract these parameters, we developed an algorithm that automatically selects a region of interest (ROI) from X-ray images and then measures JSW, an important target for the diagnosis of early rheumatoid arthritis.

Looking at the algorithm briefly, as shown in FIG. 2, the human hand is sequentially divided into three phalanges, metacarpals, carpals, radials, and ulna from the most distal end of the finger. You can see that it is configured.

The fingers consist of three bones, called phalanges, each consisting of proximal phalanx, middle phalanx, and distal phalanx, but two fingers without thumb Consists of.

The basal bone is connected to a bone called metacarpal, which is called an MCP (MetacarPophalangeal) joint. The PIP (Proximal InterPhalangeal) joint is named between the femoral bone and the femur, and the DIP (Distal InterPhalangeal) joint is named between the middle and the distal bones. It is named as an IP (InterPhangeal) joint.

First, a database is constructed by measuring JSW from the original X-Ray two-handed images of the normal group, seropositive RA and seronegativ RA.

That is, information such as the average JSW value and distribution of each group is extracted from the X-ray images of the normal group, seropositive RA and seronegativ RA, and stored in a database. Here, the X-ray information of the normal group, seropositive RA and seronegativ RA may use data that has already been determined by a medical institution such as a hospital. Since is reduced, it is desirable to extract information such as average JSW value and distribution of each group by acquiring as many groups of X-ray images as possible.

Hereinafter, a method of measuring JSW from the X-Ray image will be described with reference to FIG. 3.

The JSW measurement module first removes the background noise, and extracts only the pure hand image by binarization of the image from which the image is blurred by diffusion.

Next, a grid pattern with regular intervals in the longitudinal direction is extracted on the extracted hand image, and when the value of the corresponding grid is larger than the average value using the average value of each grid, the phalanx region (PR), If it is small, the PR information of the finger is extracted by using a step wedge function, which is processed as noise, and the median value of the PR information, that is, the PR information of the finger corresponds to the center point of the phalanges, so that the closest adjacent to the longitudinal axis direction is obtained. When connecting the center point, this information corresponds to finger information, which can be viewed as branch path of phalanx (BPP) information.

The joint position is searched through a profile plot indicating intensity information of the pixel from the BPP information of the extracted five fingers. In other words, joints located between three phalanges and metacarpals show local sudden changes (minimum values). Using this information, three (two thumb fingers) joints (DIP, PIP) , MCP) location is automatically extracted. ROI (Region Of Interest) is automatically extracted by using such joint position information and its vertical local BPP information.

This ROI was applied differently depending on the type of joint. That is, a specific region is extracted using the extracted joint position information and a parallel line existing at an arbitrary pixel distance from the local BPP, which corresponds to the ROI. The arbitrary pixel distance is 60 pixels from the joint position for the X-Ray hand image with a resolution of 2614 × 2164, 39 pixels from the local BPP, 75 pixels from the joint position, and 45 from the local BPP. The pixel and MCP joints were 96 pixels from the joint position and 60 pixels from the local BPP.

4 illustrates ROI regions of 14 joints extracted from the X-ray hand image by the above process.

Referring to FIG. 4, the distance between the joints (JSW) from the extracted ROI region may be measured, and JSW is measured as shown in FIG. 5.

Here, the JSW can not only automatically measure the JSW value from the ROI region, but also can be measured by manually clicking both ends of the JSW.

Since the ROI region is an image formed by a specific scale, since the joints of the ROI region are clearly distinguished, the distance JSW between the joints can be accurately measured.

6 is a graph of JSW results for 12 samples calculated for each group.

Here, (a) is a qualitative group, (b) is a seronegative RA group, and (c) is a seropositive RA group.

Referring to FIG. 6, it can be seen from the inter-joint spacing (JSW) results that all JSWs are reduced in the rheumatoid positive RA group and the symmetrical characteristics of both hands. It can be seen that the PIP and MCP joints, which are frequently used in the diagnosis of rheumatoid arthritis, are reduced. This is consistent with rheumatoid arthritis findings currently used for diagnosis in rheumatology. However, the results of the rheumatoid seronegative RA (RAE) group showed reduced JSW results compared to the normal group, but larger JSW results also exist and the shape is asymmetrical, and the deviation is large. This shows a tendency to osteoarthritis (OA) but not rheumatoid arthritis.

In addition, these results are expected to obtain more accurate results through the use of more and more clinical data (image X-Ray).

The results of the hand joint spacing for each finger and joint calculated as described above are compared to a program tool for measuring JSW from X-Ray images taken from any patient who visits the hospital and the database constructed as described above for rheumatoid arthritis immediately. Can be evaluated. In addition, the use of more samples is expected to enable early diagnosis of rheumatoid arthritis, as well as the diagnosis of rheumatoid arthritis in advance.

While the present invention has been described in connection with what is presently considered to be practical exemplary embodiments, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention.

1 is a schematic block diagram of a system for diagnosing rheumatoid arthritis according to a preferred embodiment of the present invention.

Figure 2 is an anatomical schematic of the bones of the hand.

Figure 3 is a flow chart schematically showing a method for diagnosing rheumatoid arthritis according to a preferred embodiment of the present invention.

4 illustrates ROI regions of 14 joints extracted from the X-ray hand image by the above process.

5 is an exemplary diagram of measuring JSW using FIG. 4.

6 is a graph of JSW results calculated for each group.

* Description of the symbols for the main parts of the drawings *

110: JSW measurement module 120: database

130: diagnostic module

Claims (7)

A system for diagnosing rheumatoid arthritis from an X-ray image, A JSW measurement module for measuring a gap between the hand joints (JSW) from the X-ray image; A database for storing JSW information for each group extracted from statistics for the normal group, seropositive RA and seronegative RA groups; And a diagnostic module for diagnosing rheumatoid arthritis by comparing the JSW measured from the JSW measurement module and JSW information for each group stored in the database. The method of claim 1, The JSW measurement module An image preprocessor extracting only a pure hand image from the X-ray image; A ROI measurement unit for automatically extracting a region of interest (ROI) from phalanx region (PR) information extracted using a step wedge function; Rheumatoid arthritis diagnosis system, characterized in that it comprises a JSW measuring unit for measuring the JSW from the ROI region. 3. The method of claim 2, The image preprocessor Rheumatoid arthritis diagnosis system, characterized in that to remove the background noise and patient information from the X-ray image, and to extract only the pure hand image through the binarization of the image (blur) by blur (diffusion). 3. The method of claim 2, The ROI measuring unit Only the PR information is extracted using the step wedge function, and the branch path of the phalanx (BPP) information is extracted using the method corresponding to the midpoint from the PR information, and the extracted Rheumatoid, characterized by calculating a profile plot of BPP and automatically extracting joint (DIP, PIP, MCP) locations using local minimum points, and extracting ROIs from the extracted joint locations. Arthritis Diagnosis System. The method of claim 1, The diagnostic module Rheumatoid arthritis is determined by analyzing the average JSW values and patterns of the JSW measured from the JSW measurement module and the normal group, seropositive RA and seronegative RA groups stored in the database. Rheumatoid arthritis diagnostic system, characterized in that the diagnosis. Calculating average hand joint interval (JSW) information for each group of the normal group, seropositive RA, and seronegative RA groups, and storing them in a database; Measuring JSW from the X-Ray image through the JSW measurement module; Rheumatoid arthritis diagnostic method comprising the step of diagnosing rheumatoid arthritis by comparing the measured JSW value and the average JSW information for each group stored in the database. The method of claim 6, Measuring the JSW Only hand images are extracted from the X-Ray image, and ROI (Region Of Interest) is automatically extracted from the extracted joint region (Phalangeal Region, PR) information using a step wedge function, and from the ROI region. A method for diagnosing rheumatoid arthritis, characterized by measuring JSW.

Images