A red blood cell quality detection system and method

Abstract

The present invention discloses a red blood cell quality detection system and method. The system includes a red blood cell processing device, a microscope and a red blood cell quality detection platform. The red blood cell processing device provided by the present invention has a simple structural design and low manufacturing cost, but plays a large role. By setting a suitable height, it is possible to avoid the overlap of red blood cells affecting the recognition and counting effect of the target detection model; by setting a suitable microchannel width and microscope magnification range, a suitable observation field of view can be obtained, and then a red blood cell image that is convenient for the subsequent target detection model to recognize can be obtained. The present invention uses a red blood cell quality index to represent the quality of red blood cells. The higher the index, the better the quality. The target detection model and the expression for calculating the red blood cell quality index provided are deployed on the red blood cell quality detection platform, which can quickly and simultaneously complete the recognition and counting of red blood cells of various morphological categories, and can finally calculate the red blood cell quality index according to the preset weight.

Description

Technical Field

[0001] The present invention belongs to the technical field of blood testing, and in particular relates to a red blood cell quality testing system and method. Background Art

[0002] Blood transfusion is an effective clinical treatment method often used in major surgeries or for the treatment of anemia. Generally, red blood cells stored in blood banks are used to supplement patients' red blood cells to help them maintain vital signs or recover quickly. The quality of stored red blood cells has a direct impact on the speed of recovery of patients receiving blood transfusion therapy. Therefore, it is very important to have a means to detect the quality of stored red blood cells.

[0003] At present, there is no unified standard for the evaluation of red blood cell quality. The commonly used traditional methods are manual observation through microscopes, image flow cytometry, electron microscopes or biochemical and physical analysis, which involve many reagents, many steps, expensive equipment, complex and time-consuming operations, often with subjective bias and labor-intensive. In recent years, with the development of microfluidics and artificial intelligence technology, relevant studies have applied these new technologies to improve the detection of red blood cell quality. For example, some studies have used microfluidics chips, microscopes and red blood cell image processing to achieve red blood cell quality detection, avoiding the use of expensive detection equipment such as image flow cytometry, but this method uses a lot of manual labor for red blood cell image processing, which is labor-intensive. Some improved methods propose an RBC morphology image segmentation model (RBC-MIST), which improves the image processing algorithm and saves labor to a certain extent, but requires expensive detection equipment such as image flow cytometry. Some improved methods use label-free imaging flow cytometers to collect red blood cell images, and then use the artificial intelligence deep learning ResNet50 model to identify and classify different red blood cell morphologies. This method aims to further improve the intelligence of red blood cell quality evaluation and reduce manual processing, but it also requires the use of expensive detection equipment such as image flow cytometers and the recognition accuracy is not high. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a red blood cell quality detection system and detection method, which solves the problems of high cost, low precision, subjective bias, complex and time-consuming operation and low intelligence level in the existing detection methods.

[0005] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is: a red blood cell quality detection system, including a red blood cell processing device, a microscope and a red blood cell quality detection platform;

[0006] The red blood cell processing device comprises a base and a cover plate; the cover plate is arranged above the base, a sample inlet and a sample outlet are arranged on the cover plate, and a microchannel, an observation area and a buffer chamber are arranged on the side of the cover plate facing the base;

[0007] The microchannel connects the sample inlet, the sample outlet, the observation area and the buffer chamber;

[0008] The red blood cell processing device is used to provide a red blood cell image sampling microenvironment, the microscope is used to collect red blood cell images, and the red blood cell quality detection platform is used to deploy a target detection model for identifying and counting red blood cells and an expression for calculating a red blood cell quality index.

[0009] Furthermore, the hole radius of the sample inlet and the sample outlet is 400 μm-600 μm.

[0010] The beneficial effects of the above further scheme are: the hole radius of the sample inlet and the sample outlet is wider, which is convenient for the entry and outflow of the red blood cell solution; the appropriate microchannel width range makes it easy for the microscope to clearly observe the morphological outline of the red blood cells under the appropriate magnification, and is just within the field of view of the microscope.

[0011] Furthermore: the width of the microchannel is 50 μm-120 μm.

[0012] Furthermore, the heights of the microchannel, the observation area and the buffer chamber are all 10 μm-22 μm, and the radius of the buffer chamber is 150 μm-200 μm.

[0013] Furthermore: the material of the substrate is glass, and the cover plate is a polydimethylsiloxane channel module.

[0014] The beneficial effects of the above further scheme are: the appropriate height range of the microchannel, observation area and buffer chamber can avoid the overlapping of red blood cells when flowing in the microchannel, so as to achieve the purpose of accurate identification and counting of red blood cells by the target detection model; the appropriate buffer chamber radius is conducive to adjusting the red blood cell solution to a flow rate that is easy to observe.

[0015] A method for detecting red blood cell quality comprises the following steps:

[0016] S1, mixing the stored red blood cells with PBS buffer to obtain a red blood cell solution;

[0017] S2. Place the red blood cell processing device on the stage of the microscope and place the observation area directly below the objective lens;

[0018] S3, introducing the red blood cell solution into the microchannel through the injection port;

[0019] S4. Collect red blood cell images through a microscope, and input the collected red blood cell images into a red blood cell quality detection platform to obtain a red blood cell quality index, thereby completing the blood quality detection.

[0020] Further: S1 specifically includes:

[0021] Mixing the stored red blood cells with PBS buffer at a volume ratio of 1:100-800 to prepare a red blood cell solution;

[0022] Wherein, the pH of the PBS buffer is 7.4.

[0023] Furthermore: in S2, the magnification of the microscope is set to 300-800 times.

[0024] The beneficial effects of the above further scheme are: setting a suitable range of the volume ratio of the stored red blood cells and the PBS buffer solution is so that the density of the number of red blood cells under the microscope observation field is suitable for the target detection algorithm to quickly identify and count; setting a suitable magnification range of the microscope is so that the morphological outline of the red blood cells can be clearly observed.

[0025] Further: S4 is specifically:

[0026] The high-speed camera on the microscope collects red blood cell images in the observation area, and the collected red blood cell images are input into the target detection model in the red blood cell quality detection platform to calculate the red blood cell quality index to complete the blood quality detection.

[0027] Further: In said S4, the expression for calculating the red blood cell quality index RBC_QI is specifically:

[0028]

[0029] Wherein, Type1 is the number of red blood cell morphology category one, Type2 is the number of red blood cell morphology category two, Type3 is the number of red blood cell morphology category three, Type4 is the number of red blood cell morphology category four, Type5 is the number of red blood cell morphology category five, and Type6 is the number of red blood cell morphology category six; α is the weight of red blood cell morphology category one, α∈[1,0.8); β is the weight of red blood cell morphology category two, β∈[0.8,0.6); γ is the weight of red blood cell morphology category three, γ∈[0.6,0.4); η is the weight of red blood cell morphology category four, η∈[0.4,0.2); ψ is the weight of red blood cell morphology category five, ψ∈[0.2,0); φ is the weight of red blood cell morphology category six, φ∈[0.2,0).

[0030] The beneficial effects of the present invention are:

[0031] (1) The red blood cell processing device provided by the present invention has a simple structural design, low manufacturing cost, and great function. By setting a suitable height, the red blood cell overlap is avoided to affect the recognition and counting effect of the target detection model; by setting a suitable microchannel width and microscope magnification range, a suitable observation field of view can be obtained, and then a red blood cell image that is convenient for the subsequent target detection model to recognize can be obtained.

[0032] (2) The target detection model and the expression for calculating the red blood cell mass index provided by the present invention are deployed on the red blood cell mass detection platform, which can quickly and simultaneously complete the recognition and counting of red blood cells of various morphological categories, and can finally calculate the red blood cell mass index according to the preset weight. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a schematic diagram of the system structure of the present invention;

[0034] Figure 2 A diagram of a red blood cell processing device according to the present invention;

[0035] Figure 3 is a cross-sectional view of the red blood cell processing device of the present invention;

[0036] Figure 4 is a flow chart of the method of the present invention;

[0037] Figure 5 This is a comparison chart of experimental results of recognizing red blood cells by the target detection model of the present invention and the YOLOv4 target detection model;

[0038] Figure 6 This is a diagram showing the effect of the target detection model of the present invention on identifying red blood cells;

[0039] Figure 7 This is a comparison chart of the number of each morphological category of red blood cells and the red blood cell quality index obtained in the experiment of the present invention.

[0040] Wherein: 1. Red blood cell processing device; 2. Microscope; 3. Red blood cell quality detection platform; 11. Base; 12. Cover plate; 13. Sample inlet; 14. Sample outlet; 15. Microchannel; 16. Observation area; 17. Buffer chamber. DETAILED DESCRIPTION

[0041] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0042] Embodiment 1:

[0043] like Figure 1 As shown, in one embodiment of the present invention, a red blood cell quality detection system includes a red blood cell processing device 1, a microscope 2 and a red blood cell quality detection platform 3;

[0044] Among them, Figure 2 and Figure 3 As shown, the red blood cell processing device 1 includes a base 11 and a cover plate 12; the cover plate 12 is arranged above the base 11, and is provided with an inlet 13 and an outlet 14 on the cover plate 12, and a microchannel 15, an observation area 16 and a buffer chamber 17 are arranged on the side of the cover plate 12 facing the base 11;

[0045] The microchannel 15 connects the sample inlet 13, the sample outlet 14, the observation area 16 and the buffer chamber 17;

[0046] The red blood cell processing device 1 is used to provide a red blood cell image sampling microenvironment, the microscope 2 is used to collect red blood cell images, and the red blood cell quality detection platform 3 is used to deploy a target detection model for identifying and counting red blood cells and an expression for calculating a red blood cell quality index.

[0047] The base 11 and the cover plate 12 are bonded and pressed together, the inlet 13 and the outlet 14 are respectively used for the flow in and out of the red blood cell solution, the microchannel 15 is used for the flow of the red blood cell solution, and the buffer chamber 17 is used to adjust the flow speed of the red blood cell solution so as to sample the red blood cell images at different flow rates in the observation area 16.

[0048] The hole radius of the sample inlet 13 and the sample outlet 14 is 400 μm-600 μm.

[0049] The width of the microchannel 15 is 50 μm-120 μm.

[0050] The heights of the microchannel 15 , the observation area 16 and the buffer chamber 17 are all 10 μm-22 μm.

[0051] In this embodiment, the hole radius of the injection port 13 and the outlet port 14 is 450 μm; the width of the microchannel 15 connecting the injection port 13, the outlet port 14 and the buffer chamber 17 is 60 μm; the heights of the microchannel 15, the observation area 16 and the buffer chamber 17 are all 12 μm; and the radius of the buffer chamber 17 is 150 μm.

[0052] The material of the substrate 11 is glass, and the cover plate 12 is a polydimethylsiloxane channel module. The polydimethylsiloxane channel module is made by the following method:

[0053] (1) Weighing 20 g of polydimethylsiloxane liquid prepolymer with a weighing instrument, and mixing the prepolymer with a curing agent at a mass ratio of 16:1 to obtain a mixture;

[0054] (2) Stir the obtained mixture thoroughly with a glass rod until a large amount of bubbles are generated, and place the mixture in a desiccator and evacuate the mixture for 20 minutes;

[0055] (3) The vacuumed mixture was slowly poured onto the silicon wafer mold and heated at 75° C. for 1 hour to complete the fabrication of the polydimethylsiloxane channel module.

[0056] Embodiment 2:

[0057] This embodiment is directed to a method for using a red blood cell quality detection system.

[0058] like Figure 4 As shown, in this embodiment, a method for detecting red blood cell quality includes the following steps:

[0059] S1, mixing the stored red blood cells with PBS buffer to obtain a red blood cell solution;

[0060] S2, placing the red blood cell processing device 1 on the stage of the microscope 2, and placing the observation area 16 directly below the objective lens;

[0061] S3, introducing the red blood cell solution into the microchannel 15 through the injection port 13;

[0062] S4. Collecting red blood cell images through the microscope 2, and inputting the collected red blood cell images into the red blood cell quality detection platform 3 to obtain the red blood cell quality index, thereby completing the blood quality detection.

[0063] The S1 is specifically:

[0064] The stored red blood cells were mixed with PBS buffer at a volume ratio of 1:100-800 to prepare a red blood cell solution.

[0065] In S2, the magnification of the microscope 2 is set to 300-800 times.

[0066] In the above-mentioned S3, the flow rate of the red blood cell solution is 1 μL / min.

[0067] The S4 is specifically:

[0068] The high-speed camera on the microscope 2 collects red blood cell images in the observation area 16, and the collected red blood cell images are input into the target detection model in the red blood cell quality detection platform 3, and the red blood cell quality index is calculated to complete the blood quality detection.

[0069] In the S4, the expression for calculating the red blood cell quality index RBC_QI is specifically:

[0070]

[0071] Wherein, Type1 is the number of red blood cell morphology category 1 (smooth disc, SDC), Type2 is the number of red blood cell morphology category 2 (wrinkled disc, CDC), Type3 is the number of red blood cell morphology category 3 (wrinkled disc, CDD), Type4 is the number of red blood cell morphology category 4 (wrinkled spherical, CSD), Type5 is the number of red blood cell morphology category 5 (wrinkled spherical, CSE), and Type6 is the number of red blood cell morphology category 6 (smooth spherical, SSE); α is the weight of red blood cell morphology category 1, α∈[1,0.8); β is the weight of red blood cell morphology category 2, β∈[0.8,0.6); γ is the weight of red blood cell morphology category 3, γ∈[0.6,0.4); η is the weight of red blood cell morphology category 4, η∈[0.4,0.2); ψ is the weight of red blood cell morphology category 5, ψ∈[0.2,0); φ is the weight of red blood cell morphology category 6, φ∈[0.2,0).

[0072] In this embodiment, the stored red blood cells and the PBS buffer solution are mixed at a volume ratio of 1:500, and the magnification of the microscope 2 is set to 400 times.

[0073] like Figure 5 As shown, the target detection model of the present invention has a high accuracy in identifying red blood cells and can avoid identifying repeated red blood cells.

[0074] The target detection model of the present invention is an improved target model YOLOv4. The traditional target detection model YOLOv4 has the problem of repeated recognition and counting when identifying red blood cells. Under normal circumstances, each red blood cell identified and counted has only one recognition frame, and a red blood cell that is repeatedly identified and counted has at least two or more recognition frames (i.e., a red blood cell is identified and counted at least twice or more). Therefore, to achieve accurate recognition and counting of a red blood cell that is repeatedly identified and counted, only one recognition frame with the highest confidence is required to be retained (confidence refers to the score of a red blood cell being determined to be a certain morphological category. The higher the confidence, the more accurate the red blood cell morphological category identified in the corresponding recognition frame), and the other recognition frames of the red blood cell are deleted. Here, it is assumed that A and B are recognition frames for identifying and counting red blood cells. The improved YOLOv4 determines whether the A and B recognition frames are repeated recognition frames in the following manner and performs a deletion operation.

[0075] (1) Calculate the center positions of the A and B recognition boxes using the following formulas;

[0076]

[0077] Among them, AX center AY center Respectively represent the pixel coordinates of the center position of the A recognition frame on the x-axis and y-axis; BX center 、BY center They respectively represent the pixel coordinate values ​​of the center position of frame A on the x-axis and y-axis; AXl, AXr, AYt, and AYb respectively represent the pixel coordinate values ​​of the left, right, top, and bottom of frame A; BXl, BXr, BYt, and BYb respectively represent the pixel coordinate values ​​of the left, right, top, and bottom of frame B.

[0078] (2) Calculate the center distance between the recognition boxes A and B using the following formula;

[0079]

[0080] Among them, XCZ center Indicates the distance between the center of the A recognition frame and the center of the B recognition frame on the x-axis; YCZ center Indicates the distance between the center of the A identification box and the center of the B identification box on the y-axis.

[0081] (3) The overlap degree of recognition frames A and B is calculated by the following formula. When the threshold condition is met, True is returned. Recognition frames A and B are determined to be duplicate recognition frames, and the frames with lower confidence are deleted, thereby solving the problem of duplicate recognition and counting. The threshold value can be in the range of δ∈[0.2,0.4], and the threshold σ x ∈[8,12],σ y ∈[8,12], in this embodiment, δ=0.3, σ x =10,σ y =10.

[0082]

[0083] Embodiment 3:

[0084] This embodiment is directed to an experiment using a method for detecting the quality of red blood cells.

[0085] The red blood cell processing device 1 prepared in the above-mentioned embodiment, the red blood cell quality detection platform 3 deployed with the improved YOLOv4 target detection model, and the microscope 2 are used to complete the red blood cell quality detection experiment. In order to verify that the present invention can effectively detect the quality of red blood cells, three groups of comparative experiments are set up. The first group of experiments sampled images and performed quality inspections after the red blood cells were stored for one week, the second group of experiments sampled images and performed quality inspections after the red blood cells were stored for three weeks, and the third group of experiments sampled images and performed quality inspections after the red blood cells were stored for six weeks. Each group of experiments was operated according to the following steps.

[0086] (1) Preparation of red blood cell solution

[0087] The stored red blood cells are mixed with PBS buffer at a volume ratio of 1:500 to prepare a red blood cell solution. The PBS buffer can be purchased.

[0088] (2) Red blood cell image acquisition

[0089] The red blood cell processing device 1 prepared in the above embodiment is placed on the stage of a microscope 2, the observation area 16 is placed directly below the objective lens, and the magnification of the microscope 2 is set to 400 times; the red blood cell solution is passed into the prepared red blood cell processing device 1, and the red blood cell image is collected by the high-speed camera on the microscope 2.

[0090] (3) The collected red blood cell images are input into the improved YOLOv4 target detection model on the red blood cell quality detection platform 3 to identify and count each red blood cell morphology category. The recognition effect is shown in Figure 6 Set the weight of red blood cell morphology category 1 to 1.0, the weight of red blood cell morphology category 2 to 0.8, the weight of red blood cell morphology category 3 to 0.6, the weight of red blood cell morphology category 4 to 0.4, the weight of red blood cell morphology category 5 to 0.2, and the weight of red blood cell morphology category 6 to 0.2.

[0091] In each group of experiments, 50 red blood cell images were collected for identification and counting. The number of each type of red blood cell morphology and red blood cell quality index obtained in each group of experiments are shown in Figure 7 The red blood cell quality index obtained in the first group of experiments was 84.53%, the red blood cell quality index obtained in the second group of experiments was 73.33%, and the red blood cell quality index obtained in the third group of experiments was 24.34%. It can be seen that the index of the first group of experiments was the highest, that is, the red blood cell quality of the first group of experiments was the best; the index of the third group of experiments was the lowest, that is, the red blood cell quality of the third group of experiments was the worst.

[0092] The beneficial effects of the present invention are as follows: the red blood cell processing device 1 provided by the present invention has a simple structure design and low manufacturing cost, but plays a great role. By setting a suitable height, the red blood cell overlap is avoided to affect the recognition and counting effect of the target detection model; by setting a suitable width of the microchannel 15 and the magnification range of the microscope 2, a suitable observation field of view can be obtained, and then a red blood cell image that is convenient for the subsequent target detection model to recognize can be obtained.

[0093] The target detection model and the expression for calculating the red blood cell mass index provided by the present invention are deployed on the red blood cell mass detection platform 3, which can quickly and simultaneously complete the recognition and counting of red blood cells of various morphological categories, and can finally calculate the red blood cell mass index according to the preset weight.

[0094] In the description of the present invention, it is necessary to understand that the orientation or positional relationship indicated by the terms "center", "thickness", "upper", "lower", "horizontal", "top", "bottom", "inner", "outer", "radial", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In addition, the terms "first", "second", and "third" are used only for descriptive purposes, and cannot be understood as indicating or implying the relative importance or the number of implicitly specified technical features. Therefore, the features defined by "first", "second", and "third" may explicitly or implicitly include one or more of the features.

Claims

1. A red blood cell quality detection system, characterized in that: It comprises a red blood cell processing device (1), a microscope (2) and a red blood cell quality detection platform (3); The red blood cell processing device (1) comprises a base (11) and a cover plate (12); the cover plate (12) is arranged above the base (11), a sample inlet (13) and a sample outlet (14) are arranged on the cover plate (12), and a microchannel (15), an observation area (16) and a buffer chamber (17) are arranged on a side of the cover plate (12) facing the base (11); The microchannel (15) connects the sample inlet (13), the sample outlet (14), the observation area (16) and the buffer chamber (17); The red blood cell processing device (1) is used to provide a red blood cell image sampling microenvironment, the microscope (2) is used to collect red blood cell images, and the red blood cell quality detection platform (3) is deployed with a target detection model for identifying and counting red blood cells and an expression for calculating a red blood cell quality index; The target detection model is an improved target model YOLOv4. The improved YOLOv4 determines whether the A and B recognition boxes are duplicate recognition boxes and performs a deletion operation in the following manner; (1) Calculate the center positions of the A and B recognition boxes using the following formulas; Among them, AX center AY center Respectively represent the pixel coordinates of the center position of the A recognition frame on the x-axis and y-axis; BX center 、BY center Respectively represent the pixel coordinate values ​​of the center position of the B identification frame on the x-axis and y-axis; AXl, AXr, AYt, AYb represent the pixel coordinate values ​​of the left, right, top, and bottom of the A identification frame; BXl, BXr, BYt, BYb represent the pixel coordinate values ​​of the left, right, top, and bottom of the B identification frame; (2) Calculate the center distance between the recognition boxes A and B using the following formula; Among them, XCZ center Indicates the distance between the center of the A recognition frame and the center of the B recognition frame on the x-axis; YCZ center Indicates the distance between the center of the A recognition frame and the center of the B recognition frame on the y-axis; (3) The overlap degree of recognition frames A and B is calculated by the following formula. When the threshold condition is met, True is returned. Recognition frames A and B are determined to be duplicate recognition frames, and the frame with lower confidence is deleted, thereby solving the problem of duplicate recognition and counting. The threshold σ x ∈[8,12],σ y ∈[8,12]; 2. The red blood cell quality detection system according to claim 1, characterized in that: The hole radius of the sample inlet (13) and the sample outlet (14) is 400 μm-600 μm.

3. The red blood cell quality detection system according to claim 1, characterized in that: The width of the microchannel (15) is 50 μm-120 μm.

4. The red blood cell quality detection system according to claim 1, characterized in that: The heights of the microchannel (15), the observation area (16) and the buffer chamber (17) are all 10 μm-22 μm, and the radius of the buffer chamber (17) is 150 μm-200 μm.

5. The red blood cell quality detection system according to claim 1, characterized in that: The material of the substrate (11) is glass, and the cover plate (12) is a polydimethylsiloxane channel module.

6. A method for detecting red blood cell quality, using the red blood cell quality detection system according to any one of claims 1 to 5, characterized in that: The following steps are involved: S1, mixing the stored red blood cells with PBS buffer to obtain a red blood cell solution; S2, placing the red blood cell processing device (1) on the stage of the microscope (2), and placing the observation area (16) directly below the objective lens; S3, introducing the red blood cell solution into the microchannel (15) through the injection port (13); S4. Collecting red blood cell images through a microscope (2), and inputting the collected red blood cell images into a red blood cell quality detection platform (3) to obtain a red blood cell quality index, thereby completing the blood quality detection.

7. The method for detecting red blood cell quality according to claim 6, characterized in that: The S1 is specifically: Mixing the stored red blood cells with PBS buffer at a volume ratio of 1:100-800 to prepare a red blood cell solution; Wherein, the pH of the PBS buffer is 7.

4.

8. The method for detecting red blood cell quality according to claim 6, characterized in that: In S2, the magnification of the microscope (2) is set to 300-800 times.

9. The method for detecting red blood cell quality according to claim 6, characterized in that: The S4 is specifically: A high-speed camera on a microscope (2) collects red blood cell images in an observation area (16), inputs the collected red blood cell images into a target detection model in a red blood cell quality detection platform (3), calculates a red blood cell quality index, and completes the blood quality detection.

10. The method for detecting red blood cell quality according to claim 9, characterized in that: In S4, the expression for calculating the red blood cell quality index RBC_QI is specifically: Wherein, Type1 is the number of red blood cell morphology category one, Type2 is the number of red blood cell morphology category two, Type3 is the number of red blood cell morphology category three, Type4 is the number of red blood cell morphology category four, Type5 is the number of red blood cell morphology category five, and Type6 is the number of red blood cell morphology category six; α is the weight of red blood cell morphology category one, α∈[1,0.8); β is the weight of red blood cell morphology category two, β∈[0.8,0.6); γ is the weight of red blood cell morphology category three, γ∈[0.6,0.4); η is the weight of red blood cell morphology category four, η∈[0.4,0.2); ψ is the weight of red blood cell morphology category five, ψ∈[0.2,0); φ is the weight of red blood cell morphology category six, φ∈[0.2,0).

Images