Microplate AI image recognition and acquisition device and image recognition algorithm

Abstract

The invention relates to the technical field of blood analysis, and provides a microplate AI image recognition and acquisition device and an image recognition algorithm, the device comprises a base, the base is provided with a light source module and a photographing module, the light source module comprises a light source overturning cover and a light source outer cover, the light source outer cover is through up and down, and the photographing module is arranged on the light source overturning cover. The light source overturning cover is arranged at the top of the light source outer cover in an openable and closable mode, a light source is arranged on the inner side of the light source plate, and a plurality of micro-plate strips used for containing platelet samples are installed in the light source outer cover; the photographing module comprises a camera and an XY bidirectional moving mechanism used for driving the camera, the camera and the XY bidirectional moving mechanism are located below the light source outer cover, and the photographing end of the camera faces the light source outer cover. According to the device, the automation degree of the detection process is improved, batch samples can be detected at a time, and the detection efficiency is greatly improved. According to the microplate AI image recognition algorithm, the analysis result is more accurate and convenient, and false positive and false negative problems are avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of blood analysis, and in particular to a microplate AI image recognition and acquisition device and an image recognition algorithm. Background Art

[0002] In recent years, with the continuous deepening of medical research, the recognition and detection demand for platelet antibodies has gradually increased. At present, the detection of platelet antibodies is mostly carried out by taking pictures and then recognizing images, but the existing image recognition method is inefficient and low in accuracy, and cannot effectively detect platelet conditions. It is prone to false positives and false negatives. These two false results may cause doctors to misdiagnose, leading to unnecessary treatment interventions and causing trouble to medical work. Therefore, a more accurate and efficient microplate image recognition acquisition device and image recognition method are needed. Summary of the invention

[0003] In order to solve the problems in the background technology, the present invention proposes a microplate AI image recognition and acquisition device, comprising a base, on which a light source module and a camera module are arranged, the light source module comprises a light source flip cover and a light source outer cover, the light source outer cover is through-through, the light source flip cover is openably arranged on the top of the light source outer cover, a light source is arranged on the inner side of the light source flip cover, and a plurality of microplate strips for placing platelet samples are installed in the light source outer cover; The camera module comprises a camera and an XY bidirectional moving mechanism for driving the camera. The camera and the XY bidirectional moving mechanism are located below the light source housing, and the camera shooting end faces the light source housing.

[0004] Preferably, one end of the light source flip cover is hinged to one end of the top of the light source outer cover via a rotating shaft.

[0005] Preferably, a driven pulley is coaxially fixed on the rotating shaft, a reduction cover opening motor is fixed to one side of the light source outer cover, an output end of the reduction cover opening motor is connected to a driving pulley, and the driven pulley and the driving pulley are connected via a synchronous belt transmission.

[0006] Preferably, a diffuse reflection plate is provided on one side of the light source close to the light source cover.

[0007] Preferably, the XY bidirectional moving mechanism comprises an X-axis linear motor, a Y-axis linear motor, and an X-axis guide rail and a Y-axis guide rail respectively arranged along the X direction and the Y direction, the Y-axis linear motor and the Y-axis guide rail are both fixedly connected to the base, a connecting block is slidably connected to the Y-axis guide rail, and the lead screw of the Y-axis linear motor is threadedly connected to the connecting block; The X-axis guide rail is fixed on the connecting block, a camera moving seat is slidably connected to the X-axis guide rail, and a lead screw of the X-axis linear motor is threadedly connected to the camera moving seat; The camera is fixed on the camera moving seat.

[0008] Preferably, the Y-axis guide rail is fixed to the base through a Y-axis fixing plate, the Y-axis linear motor is located at one end of the Y-axis fixing plate, and a first limit plate is fixed to the other end of the Y-axis fixing plate, the X-axis guide rail is fixed to the connecting block through an X-axis fixing plate, the X-axis linear motor is fixed to one end of the X-axis fixing plate, and a second limit plate is fixed to the other end of the X-axis fixing plate.

[0009] Preferably, the light source housing is provided with A plurality of strip racks are provided with a plurality of support cups along the length direction of the strip racks, and the micro strips are placed in the support cups.

[0010] A microplate AI image recognition algorithm, applied to the image recognition acquisition device according to any one of claims 1 to 7, comprises the following steps: S1. Open the flip cover of the light source, place the microplate strip containing the platelet sample into the outer cover of the light source, and close the flip cover of the light source; S2, the XY bidirectional moving mechanism is started, the camera is moved to the bottom of the micro-strip to be detected, and the center of the camera's image acquisition angle is located at the center of the reaction cup of the micro-strip, the camera takes pictures, obtains the interpretation image, and the interpretation library program in the computer reads the interpretation image by calling the interface; S3, taking the center of the judgment image as the center of the circle and R as the radius, demarcating a circular area, and taking the image of the circular area for recognition and interpretation; S4, the identification and interpretation steps are: S41, converting the image into HSV color space pixels, and naming the converted image P0; S42, setting the red blood cell HSV threshold, extracting red blood cell pixels from the image P0; S5. Perform morphological analysis and interpretation on the extracted red blood cell pixels. The analysis and interpretation steps are as follows: S51. Perform negative interpretation: determine whether the red blood cell aggregate morphology meets the negative standard. If so, judge whether the red blood cell aggregate morphology is negative or suspicious based on the set circular threshold and negative standard. If not, perform positive interpretation. S52, performing positive judgment: according to the positive type classification standard, taking the number of red blood cells as the main parameter, firstly performing morphological and magnitude analysis on the red blood cells in multiple images, and extracting the number of red blood cells under different positive type classifications as the classification threshold; When the image is interpreted after the classification threshold is obtained, the number of red blood cells in the image is compared with the classification threshold of the number of red blood cells, the positive level is interpreted, and the interpretation result is obtained.

[0011] Preferably, the step S51 is specifically as follows: S511, finding all contour points under the aggregated morphology of red blood cells; S512, finding a contour that is approximately circular from the contour points; S513, first analyzing a plurality of images to obtain a lower limit of the approximate degree of the approximate circle, and after obtaining the lower limit of the approximate degree, when performing image analysis again, if the approximate degree of the approximate circle in the image is higher than the lower limit, then the approximate circle is determined to be a negative determination area; S514, obtaining the radius of the minimum closed circle of the negative determination area; S515, first analyze the multiple images to obtain an empirical value range of the closed circle radius; S516, after obtaining the empirical value range of the closed circle radius, when analyzing the image again, if the radius of the minimum closed circle of the negative judgment area in the image is within the empirical value range, then determine whether there is a white area in this area; S517. If there is no white area, the result is determined to be negative; if there is a white area, the result is determined to be suspicious; at this point, the result is determined to be negative or suspicious.

[0012] Preferably, the positive level includes four levels, each classification threshold corresponds to a positive level, and the judgment result is one of positive level one to positive level four.

[0013] The present invention has the following beneficial effects: The microplate AI image recognition and acquisition device of the present invention can utilize a light source to illuminate a platelet sample, thereby increasing the brightness of the environment, thereby increasing the clarity of the image and ensuring the accuracy of subsequent image analysis. The device of the present invention improves the degree of automation of the detection process, and can perform batch sample detection at one time, greatly improving detection efficiency. The image is then analyzed using the microplate AI image recognition algorithm of the present invention, and the analysis results are more accurate and convenient, avoiding the occurrence of false positive and false negative problems. The device and method described in the present invention can greatly improve the detection efficiency and accuracy of platelet antibodies. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 The overall structure of the present invention is shown in FIG. Figure 1 ; Figure 2 The overall structure of the present invention is shown in FIG. Figure 2 ; Figure 3 It is a schematic diagram of image P of the present invention; Figure 4 It is a schematic diagram of image P0 of the present invention; Figure 5 This is a schematic diagram of an image after the red blood cell pixels are extracted from the image P0 according to the present invention.

[0015] Numbers in the figure: 1. Motor support plate; 2. Light source flip cover; 3. Light source; 4. Support cup; 5. Frame; 6. Light source cover; 7. Camera; 8. Camera upper cover; 9. Camera lower seat; 10. Camera moving seat; 11. Rotating axis; 12. Driven pulley; 13. Driving pulley; 14. Speed ​​reduction cover opening motor; 15. X-axis linear motor; 16. X-axis guide rail; 17. Y-axis linear motor; 18. Y-axis guide rail; 19. Base; 20. Connecting block; 21. X-axis fixed plate; 22. Y-axis fixed plate; 23. First limit plate; 24. Second limit plate; 25. Photo control board; 26. Photo adapter board; 27. Synchronous belt; 28. Light source baffle; 29. ​​Tensioning wheel. DETAILED DESCRIPTION

[0016] In order to make the present invention clearer and more understandable, the technical solution of the present invention is further described in detail below in combination with the accompanying drawings and embodiments. It should be understood that the given embodiment is only one of the implementation methods and does not represent all embodiments.

[0017] In this article, the terms "inside, outside, up, down, front, back" and so on are established based on the positional relationship shown in the drawings. Depending on the different drawings, the corresponding positional relationship may also change accordingly. Therefore, they cannot be understood as an absolute limitation on the scope of protection.

[0018] Combined with Figure 1 -Attached Figure 2 A microplate AI image recognition and acquisition device comprises a base 19, on which a light source module and a camera module are arranged, wherein the light source module comprises a light source flip cover 2 and a light source outer cover 6, wherein the light source outer cover 6 is through-through, wherein the light source flip cover 2 is openably arranged on the top of the light source outer cover 6, wherein a light source 3 is arranged inside the light source flip cover 2, wherein a plurality of microplate strips for placing platelet samples are installed inside the light source outer cover 6; The camera module includes a camera 7 and an XY bidirectional moving mechanism for driving the camera 7. The camera 7 and the XY bidirectional moving mechanism are located below the light source housing 6. The camera 7 shooting end faces the light source housing 6, so that the camera 7 can shoot the micro-strips. After the light source flip cover 2 is closed, the light source 3 can illuminate the sample of the micro-strips to provide light for the camera 7 to shoot.

[0019] Specifically, one end of the light source flip cover 2 is hinged to one end of the top of the light source housing 6 through a rotating shaft 11. More specifically, both ends of the rotating shaft 11 are rotatably connected to the light source housing 6, and one end of the light source flip cover 2 is fixedly connected to the rotating shaft 11. Specifically, a driven pulley 12 is coaxially fixed on the rotating shaft 11, a reduction cover opening motor 14 is fixed to one side of the light source housing 6, and a driving pulley 13 is connected to the output end of the reduction cover opening motor 14. The driven pulley 12 and the driving pulley 13 are connected through a synchronous belt 27. The driving pulley 13 is driven to rotate by the reduction cover opening motor 14, and the driving pulley 13 drives the driven pulley 12 to rotate through the synchronous belt 27. The driven pulley 12 drives the rotating shaft 11 to rotate, and the rotating shaft 11 drives the light source flip cover 2 to rotate. The opening and closing of the light source flip cover 2 can be achieved by the forward and reverse driving of the reduction cover opening motor 14.

[0020] More specifically, the deceleration cover opening motor 14 is fixed to the rear side of the light source housing 6 via a motor support plate 1 , and a tensioning wheel 29 is rotatably mounted on the motor support plate 1 . The tensioning wheel 29 contacts the outer wall of the synchronous belt 27 for tensioning the synchronous belt 27 .

[0021] Specifically, a diffuse reflection plate is provided on one side of the light source 3 close to the light source housing 6, and the diffuse reflection plate is used to make the illumination more uniform and natural. More specifically, the light source 3 and the diffuse reflection plate are fixed on the light source flip cover 2 through a light source baffle 28.

[0022] Specifically, the XY bidirectional moving mechanism includes an X-axis linear motor 15, a Y-axis linear motor 17, and an X-axis guide rail 16 and a Y-axis guide rail 18 respectively arranged along the X direction and the Y direction. The Y-axis linear motor 17 and the Y-axis guide rail 18 are both fixedly connected to the base 19. The X direction and the Y direction are the left and right directions and the front and back directions of the light source housing 6 with reference to the light source housing 6. A connecting block 20 is slidably connected to the Y-axis guide rail 18, and the lead screw of the Y-axis linear motor 17 is threadedly connected to the connecting block 20. The X-axis guide rail 16 is fixed on the connecting block 20, the camera moving seat 10 is slidably connected to the X-axis guide rail 16, and the lead screw of the X-axis linear motor 15 is threadedly connected to the camera moving seat 10; The camera 7 is fixed on a camera moving seat 10 .

[0023] The Y-axis linear motor 17 drives the connecting block 20 to slide along the Y-axis guide rail 18, so that the X-axis guide rail 16 connected to the connecting block 20 can move along the Y direction, so that the camera moving seat 10 can move along the Y direction, thereby realizing the Y-direction movement of the camera 7; the X-axis linear motor 15 drives the camera moving seat 10 to slide along the X-axis guide rail 16, thereby realizing the X-direction movement of the camera 7. The position of the camera 7 is adjusted by the XY bidirectional moving mechanism so that the camera 7 can be located below the micro-strip to be detected.

[0024] Specifically, the Y-axis guide rail 18 is fixed to the base 19 through the Y-axis fixing plate 22, the Y-axis linear motor 17 is located at one end of the Y-axis fixing plate 22, and the other end of the Y-axis fixing plate 22 is fixed with a first limiting plate 23, the X-axis guide rail 16 is fixed to the connecting block 20 through the X-axis fixing plate 21, the X-axis linear motor 15 is fixed to one end of the X-axis fixing plate 21, and the other end of the X-axis fixing plate 21 is fixed with a second limiting plate 24. The first limiting plate 23 and the second limiting plate 24 can limit the connecting block 20 and the camera moving seat 10 respectively, so as to prevent the connecting block 20 from being separated from the Y-axis guide rail 18 and the camera moving seat 10 from being separated from the X-axis guide rail 16.

[0025] More specifically, a camera lower seat 9 is fixed on the camera moving seat 10, and the camera 7 is installed in the camera lower seat 9. A camera upper cover 8 is fixed above the camera lower seat 9, and the camera upper cover 8 fixes the camera 7 in the camera lower seat 9. An opening corresponding to the position of the shooting end of the camera 7 is left on the camera upper cover 8 to prevent the camera upper cover 8 from blocking the shooting of the camera 7; the camera lower seat 9 and the camera upper cover 8 are used to install and protect the camera 7.

[0026] Specifically, a plurality of strip racks 5 are arranged in parallel in the light source housing 6, and a plurality of cups 4 are arranged on the strip racks 5 along the length direction, and the micro strips are placed in the cups 4. More specifically, the cups 4 can be eight-unit cups, and the eight-unit cups and the plurality of strip racks 5 can realize the placement of a plurality of micro strips, thereby increasing the placement amount of the micro strips, and there is no need to repeatedly place the micro strips into the light source housing 6, thereby improving the detection efficiency.

[0027] Specifically, it also includes a camera control board 25 and a camera adapter board 26. The camera control board 25 is electrically connected to the X-axis linear motor 15, the Y-axis linear motor 17, the light source 3 and the deceleration cover opening motor 14, and the camera control board 25 controls the start and stop and action parameters of the X-axis linear motor 15, the Y-axis linear motor 17 and the deceleration cover opening motor 14, and controls the switch of the light source 3. The camera adapter board 26 is used to transfer the motor line and the zero line of the X-axis linear motor 15 to the camera control board 25 through the cable, so as to ensure that the X-axis linear motor 15 can move better relative to the base 19.

[0028] Combined with Figure 3 -Attached Figure 5 , a microplate AI image recognition algorithm, applied to the image recognition acquisition device according to any one of claims 1 to 7, comprising the following steps: S1, open the light source flip cover 2, place the microplate strip containing the platelet sample into the light source cover 6, and close the light source flip cover 2; S2, the XY bidirectional moving mechanism is started, the camera 7 is moved to the bottom of the micro strip to be detected, and the center of the camera 7 image acquisition angle is located at the center of the reaction cup of the micro strip, the camera 7 takes a picture to obtain the interpretation image, and the interpretation library program in the computer reads the interpretation image by calling the interface; S3, taking the center of the judgment image as the center of the circle and R as the radius, demarcating a circular area, and taking the image of the circular area for recognition and interpretation; S4, the identification and interpretation steps are: S41, converting the image into HSV color space pixels, and naming the converted image P0; S42, setting the red blood cell HSV threshold, extracting red blood cell pixels from the image P0; S5. Perform morphological analysis and interpretation on the extracted red blood cell pixels. The analysis and interpretation steps are as follows: S51. Perform negative interpretation: determine whether the red blood cell aggregate morphology meets the negative standard. If so, judge whether the red blood cell aggregate morphology is negative or suspicious based on the set circular threshold and negative standard. If not, perform positive interpretation. S52, performing positive judgment: according to the positive type classification standard, taking the number of red blood cells as the main parameter, firstly performing morphological and magnitude analysis on the red blood cells in multiple images, and extracting the number of red blood cells under different positive type classifications as the classification threshold; When the image is interpreted after the classification threshold is obtained, the number of red blood cells in the image is compared with the classification threshold of the number of red blood cells, the positive level is interpreted, and the interpretation result is obtained.

[0029] Specifically, the step S51 is as follows: S511, finding all contour points under the aggregated morphology of red blood cells; S512, finding a contour that is approximately circular from the contour points; S513, firstly analyzing the multiple images to obtain the lower limit of the approximation of the approximate circle, that is, analyzing the multiple images to obtain the approximation value of the approximate circle, when an approximation value X appears and the subsequent multiple approximation values ​​are not less than X, X is the lower limit of the approximation; After obtaining the lower limit of the approximation, when performing image analysis again, if the approximation of the approximate circle in the image is higher than the lower limit, the approximate circle is determined to be a negative determination area; S514, obtaining the radius of the minimum closed circle of the negative determination area; S515, firstly analyzing the multiple images to obtain an empirical value range of the closed circle radius, that is, analyzing the multiple images to obtain multiple closed circle radius values, and determining a numerical range where most of the closed circle radius values ​​are located, and this numerical range is the empirical value range; S516, after obtaining the empirical value range of the closed circle radius, when analyzing the image again, if the radius of the minimum closed circle of the negative judgment area in the image is within the empirical value range, then determine whether there is a white area in this area; S517. If there is no white area, the result is determined to be negative; if there is a white area, the result is determined to be suspicious; at this point, the result is determined to be negative or suspicious.

[0030] It should be understood that multiple image analysis is only required when the classification threshold of the positive type, the lower limit of the approximate circle approximation, and the empirical value range of the closed circle radius are not obtained. Once they are obtained, there is no need to perform the relevant steps of analyzing multiple images.

[0031] Specifically, the positive level includes four levels, each classification threshold corresponds to a positive level, and the judgment result is one of positive level one to positive level four.

[0032] While embodiments of the present invention have been shown and described, it will be apparent to those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A microplate AI image recognition and acquisition device, comprising a base (19), characterized in that: The base (19) is provided with a light source module and a camera module, the light source module comprising a light source flip cover (2) and a light source outer cover (6), the light source outer cover (6) is connected from top to bottom, the light source flip cover (2) is openably arranged on the top of the light source outer cover (6), a light source (3) is provided on the inner side of the light source flip cover (2), and a plurality of microplate strips for placing platelet samples are installed in the light source outer cover (6); The camera module comprises a camera (7) and an XY bidirectional moving mechanism for driving the camera (7); the camera (7) and the XY bidirectional moving mechanism are located below the light source housing (6); and the shooting end of the camera (7) faces the light source housing (6).

2. A micro-plate AI image recognition and acquisition device according to claim 1, characterized in that: One end of the light source flip cover (2) is hinged to one end of the top of the light source outer cover (6) via a rotating shaft (11).

3. A micro-plate AI image recognition and acquisition device according to claim 2, characterized in that: A driven pulley (12) is coaxially fixed on the rotating shaft (11), a deceleration cover opening motor (14) is fixed on one side of the light source outer cover (6), an output end of the deceleration cover opening motor (14) is connected to a driving pulley (13), and the driven pulley (12) and the driving pulley (13) are connected in transmission via a synchronous belt (27).

4. A micro-plate AI image recognition and acquisition device according to claim 1, characterized in that: A diffuse reflection plate is provided on one side of the light source (3) close to the light source outer cover (6).

5. A micro-plate AI image recognition and acquisition device according to claim 1, characterized in that: The XY bidirectional movement mechanism comprises an X-axis linear motor (15), a Y-axis linear motor (17), and an X-axis guide rail (16) and a Y-axis guide rail (18) respectively arranged along the X direction and the Y direction, wherein the Y-axis linear motor (17) and the Y-axis guide rail (18) are both fixedly connected to a base (19), a connecting block (20) is slidably connected to the Y-axis guide rail (18), and a lead screw of the Y-axis linear motor (17) is threadedly connected to the connecting block (20); The X-axis guide rail (16) is fixed on the connecting block (20), a camera moving seat (10) is slidably connected to the X-axis guide rail (16), and a lead screw of the X-axis linear motor (15) is threadedly connected to the camera moving seat (10); The camera (7) is fixed on a camera moving seat (10).

6. A micro-plate AI image recognition and acquisition device according to claim 5, characterized in that: The Y-axis guide rail (18) is fixed to the base (19) via a Y-axis fixing plate (22); the Y-axis linear motor (17) is located at one end of the Y-axis fixing plate (22); a first limit plate (23) is fixed to the other end of the Y-axis fixing plate (22); the X-axis guide rail (16) is fixed to the connecting block (20) via an X-axis fixing plate (21); the X-axis linear motor (15) is fixed to one end of the X-axis fixing plate (21); and a second limit plate (24) is fixed to the other end of the X-axis fixing plate (21).

7. A micro-plate AI image recognition and acquisition device according to claim 1, characterized in that: A plurality of strip frames (5) are arranged in parallel inside the light source outer cover (6), a plurality of support cups (4) are arranged on the strip frames (5) along the length direction, and the micro strips are placed in the support cups (4).

8. A microplate AI image recognition algorithm, characterized by: The image recognition and acquisition device according to any one of claims 1 to 7 comprises the following steps: S1, open the light source flip cover (2), place the microplate strip containing the platelet sample into the light source cover (6), and close the light source flip cover (2); S2, the XY bidirectional moving mechanism is started, the camera (7) is moved to the bottom of the micro strip to be detected, and the center of the camera (7) image acquisition angle of view is located at the center of the reaction cup of the micro strip, the camera (7) takes a picture to obtain a judgment image, and the judgment library program in the computer reads the judgment image by calling the interface; S3, taking the center of the judgment image as the center of the circle and R as the radius, demarcating a circular area, and taking the image of the circular area for recognition and interpretation; S4, the identification and interpretation steps are: S41, converting the image into HSV color space pixels, and naming the converted image P0; S42, setting the red blood cell HSV threshold, extracting red blood cell pixels from the image P0; S5. Perform morphological analysis and interpretation on the extracted red blood cell pixels. The analysis and interpretation steps are as follows: S51. Perform negative interpretation: determine whether the red blood cell aggregate morphology meets the negative standard. If so, judge whether the red blood cell aggregate morphology is negative or suspicious based on the set circular threshold and negative standard. If not, perform positive interpretation. S52, performing positive judgment: according to the positive type classification standard, taking the number of red blood cells as the main parameter, firstly performing morphological and magnitude analysis on the red blood cells in multiple images, and extracting the number of red blood cells under different positive type classifications as the classification threshold; When the image is interpreted after the classification threshold is obtained, the number of red blood cells in the image is compared with the classification threshold of the number of red blood cells, the positive level is interpreted, and the interpretation result is obtained.

9. A microplate AI image recognition algorithm according to claim 8, characterized in that: The step S51 is specifically as follows: S511, finding all contour points under the aggregated morphology of red blood cells; S512, finding a contour that is approximately circular from the contour points; S513, first analyzing a plurality of images to obtain a lower limit of the approximate degree of the approximate circle, and after obtaining the lower limit of the approximate degree, when performing image analysis again, if the approximate degree of the approximate circle in the image is higher than the lower limit, then the approximate circle is determined to be a negative determination area; S514, obtaining the radius of the minimum closed circle of the negative determination area; S515, first analyze the multiple images to obtain an empirical value range of the closed circle radius; S516, after obtaining the empirical value range of the closed circle radius, when analyzing the image again, if the radius of the minimum closed circle of the negative judgment area in the image is within the empirical value range, then determine whether there is a white area in this area; S517. If there is no white area, the result is determined to be negative; if there is a white area, the result is determined to be suspicious; at this point, the result is determined to be negative or suspicious.

10. A microplate AI image recognition algorithm according to claim 9, characterized in that: The positive grade includes four grades, each classification threshold corresponds to a positive grade, and the judgment result is one of positive grade one to positive grade four.

Images