Cell morphology analysis equipment, cell identification method and storage medium

Abstract

A cell morphology analysis device, a cell identification method and a storage medium, wherein a digital camera captures a cell image of a blood sample under a low-power objective lens, a processor identifies and locates suspected preset type cells in the cell image, obtains an identification result, determines the located number of suspected preset type cells based on the identification result and the target number, and after completing the identification and positioning of the located number of suspected preset type cells, the digital camera captures the identified and located suspected preset type cells under a high-power objective lens, and the processor then identifies whether the captured suspected preset type cells are preset type cells, so as to count the number of preset type cells captured under the high-power objective lens and obtain a statistical value. When the statistical value satisfies: target number ≤ statistical value, the capture is stopped. At the end, the processor identifies at least the target number of preset type cells to meet clinical needs for preset type cells.

Description

Technical Field

[0001] The present invention relates to medical equipment technology, and in particular to a cell morphology analysis device, a cell identification method and a storage medium. Background Art

[0002] Blood cell morphology analysis mainly analyzes the cell morphology in a blood sample to classify the cells in the blood sample and even identify abnormal cells in the blood sample. For example, blood cell morphology analysis can be used to analyze a certain number of preset cell types in a blood sample, such as a certain number of white blood cells. The process is as follows:

[0003] For example, a low-magnification objective lens in a cell morphology analysis device captures a cell image of a blood sample, identifies and locates cells in the cell image, and when the number of identified and located cells is a certain ratio higher than the target number (such as 1.5N, where N is the target number), the low-magnification objective lens is switched to a high-magnification objective lens for shooting.

[0004] However, there are certain problems in the implementation of this treatment measure of floating up a certain proportion: for blood samples with many interfering cells, even if they float up a certain proportion, the number of preset cell types photographed under a high-power objective lens will still not reach the target number, and cannot meet the clinical demand for preset cell types. Summary of the Invention

[0005] The embodiments of the present invention provide a cell morphology analysis device, a cell identification method and a storage medium, which can capture at least a target number of cells of a preset type under a high-power objective lens, thereby meeting clinical needs for cells of the preset type.

[0006] The technical solution of the embodiment of the present invention is achieved by the following method:

[0007] In one aspect, an embodiment of the present invention provides a cell morphology analysis device, comprising: a digital imaging device, a control device, a processor, and an output device; the digital imaging device comprises: a low-power objective lens, a high-power objective lens, and a digital camera;

[0008] The digital camera is configured to capture cell images of the blood sample under the low-magnification objective lens;

[0009] The processor is configured to obtain a target number, the target number being the number of cells of a preset type that need to be photographed under the high-power objective lens in the cell morphology analysis device, the processor identifying and locating suspected cells of the preset type in the cell image to obtain an identification result, and based on the identification result and the target number, determining the number of cells of the suspected preset type identified and located under the low-power objective lens, and after completing the identification and positioning of the number of cells of the suspected preset type, instructing the digital camera to stop photographing under the low-power objective lens;

[0010] The control device is configured to switch the low-power objective lens to the high-power objective lens after the processor completes the identification and positioning of the number of cells suspected of the preset type;

[0011] The digital camera is further configured to photograph the suspected preset type of cells identified and located under the low-magnification objective lens under the high-magnification objective lens;

[0012] The processor is further configured to identify whether the cells of the suspected preset type photographed under the high-power objective lens are the cells of the preset type, count the number of cells of the preset type photographed under the high-power objective lens to obtain a statistical value, and instruct the digital camera to stop photographing under the high-power objective lens when the statistical value satisfies: target number ≤ statistical value;

[0013] The output device is used to output cell information of cells suspected of a preset type that are identified as cells of a preset type.

[0014] In another aspect, an embodiment of the present invention provides a cell identification method, which is applied to a cell morphology analysis device. The method includes:

[0015] Obtaining a target number, where the target number is the number of cells of a preset type that the cell morphology analysis device needs to photograph under a high-power objective lens;

[0016] Take cell images of blood samples using a low-magnification objective lens;

[0017] Identifying and locating cells of suspected preset types in the cell image to obtain an identification result, and determining the number of cells of suspected preset types identified and located under a low-power objective lens based on the identification result and the target number;

[0018] After completing the identification and positioning of a number of cells suspected of a preset type, stopping the photographing under the low-magnification objective lens;

[0019] Switching the low-power objective lens to the high-power objective lens;

[0020] photographing the suspected preset type of cells identified and located under the low-magnification objective lens under the high-magnification objective lens;

[0021] Identifying whether the suspected preset type cells photographed under the high-power objective lens are the preset type cells, and counting the number of the preset type cells photographed under the high-power objective lens to obtain a statistical value;

[0022] When the statistical value satisfies: target number ≤ statistical value, stop shooting under the high-power objective lens;

[0023] The cell information of the suspected predetermined type cell identified as the predetermined type cell is output.

[0024] In another aspect, an embodiment of the present invention provides a cell identification method, applied to a cell morphology analysis device, the method comprising:

[0025] Obtaining a target number, wherein the target number is the number of cells of a preset type that need to be photographed under a high-power objective lens in the cell morphology analysis device;

[0026] Acquire cell images of blood samples taken under a low-magnification objective lens;

[0027] Identifying cells of suspected preset types in the cell image, and sequentially identifying and locating the cells of suspected preset types;

[0028] Obtain information that all suspected preset type cells at the current moment are identified as preset type cells to determine whether it is necessary to locate and identify the next suspected preset type cell;

[0029] When it is necessary to locate and identify the next suspected preset type cell, locate and identify the next suspected preset type cell;

[0030] When there is no need to locate and identify the next suspected preset type of cell, stop photographing under the low-magnification objective lens;

[0031] Switching the low-power objective lens to the high-power objective lens;

[0032] Acquiring the suspected preset type cells identified and located under the low-magnification objective lens and photographed under the high-magnification objective lens, identifying whether the suspected preset type cells photographed under the high-magnification objective lens are the preset type cells, and stopping photographing under the high-magnification objective lens when the number of the preset type cells meets the target number;

[0033] The cell information of the suspected predetermined type cell identified as the predetermined type cell is output.

[0034] In another aspect, an embodiment of the present invention provides a cell morphology analysis device, comprising:

[0035] a memory configured to store executable instructions;

[0036] The processor is configured to execute any one of the above-mentioned cell identification methods when running the executable instructions stored in the memory.

[0037] On the other hand, an embodiment of the present invention provides a storage medium storing executable instructions, which is configured to cause a processor to execute the executable instructions to implement any of the above-mentioned cell identification methods.

[0038] In an embodiment of the present invention, a cell morphology analysis device includes: a digital imaging device, a control device, a processor and an output device. The digital imaging device includes a low-magnification objective lens, a high-magnification objective lens and a digital camera. The digital camera captures a cell image of a blood sample under the low-magnification objective lens. The processor identifies and locates suspected preset type cells in the cell image to obtain an identification result. Based on the identification result and the target number of preset type cells that need to be photographed under the high-magnification objective lens, the number of suspected preset type cells identified and located under the low-magnification objective lens is determined. After completing the identification and positioning of the located number of suspected preset type cells, the digital camera photographs the suspected preset type cells identified and located under the high-magnification objective lens. The processor then identifies whether the suspected preset type cells photographed under the high-magnification objective lens are the preset type cells, so as to count the number of preset type cells photographed under the high-magnification objective lens to obtain a statistical value. When the statistical value satisfies: the target number ≤ the statistical value, the digital camera is instructed to stop photographing under the high-magnification objective lens, thereby ending the identification of the suspected preset type cells photographed under the high-magnification objective lens. At the end, the processor recognizes at least the target number of preset cell types, indicating that no matter how many or few interfering cells are in the blood sample, at least the target number of preset cell types can be captured under a high-power objective lens, meeting the clinical demand for preset cell types. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0040] Figure 1 This is a schematic diagram of an optional structure of a cell morphology analysis device provided in an embodiment of the present invention;

[0041] Figure 2 This is an optional connection diagram of the digital imaging device provided by an embodiment of the present invention;

[0042] Figure 3 This is a schematic diagram of identification and positioning under a low-magnification objective lens provided by an embodiment of the present invention;

[0043] Figure 4 and Figure 5 Yes Figure 3 The image obtained by taking a high-power objective lens to the identification and positioning results shown;

[0044] Figure 6 This is an optional schematic diagram for displaying cell information provided by an embodiment of the present invention;

[0045] Figure 7This is another optional structural diagram of the cell morphology analysis device provided in an embodiment of the present invention;

[0046] Figures 8 to 11 These are images of different types of cells taken under a high-power objective lens provided by an embodiment of the present invention;

[0047] Figure 12 This is another optional structural diagram of the cell morphology analysis device provided in an embodiment of the present invention;

[0048] Figure 13 This is an optional schematic diagram of an image displayed by a cell morphology analysis device provided in an embodiment of the present invention;

[0049] Figure 14 This is an optional flow chart of the cell identification method provided by an embodiment of the present invention;

[0050] Figure 15 is another optional flow chart of the cell identification method provided in an embodiment of the present invention;

[0051] Figure 16 This is another optional structural diagram of the cell morphology analysis device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to make the objectives, technical solutions and advantages of the present invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings. The present invention should not be understood as being limited to the embodiments provided. On the contrary, the contents described in the embodiments of the present invention make the present invention comprehensive and complete, and convey the concepts of the embodiments of the present invention to those skilled in the art. Therefore, other embodiments obtained by ordinary technicians in this field without making creative efforts are all within the scope of protection of the present invention.

[0053] It should be noted that, in the embodiments of the present invention, the terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a method or server comprising a series of elements includes not only the elements explicitly stated, but also other elements not explicitly listed, or also includes elements inherent to the implementation of the method or server. In the absence of further restrictions, an element defined by the phrase "comprising a ..." does not exclude the presence of other related elements (such as steps in the method or units in the server, for example, a unit can be a part of a circuit, a part of a processor, a part of a program or software, etc.) in the method or server comprising the element.

[0054] For example, the cell identification method provided in the embodiment of the present invention includes a series of steps, but the cell identification method provided in the embodiment of the present invention is not limited to the recorded steps. Similarly, the cell morphology analysis equipment provided in the embodiment of the present invention includes a series of devices, but the cell morphology analysis equipment provided in the embodiment of the present invention is not limited to including the devices explicitly recorded, and may also include devices required to obtain relevant information or perform processing based on the information. It should be noted that in the following description, reference is made to "one embodiment", which describes a subset of all possible embodiments, but it is understood that "one embodiment" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict.

[0055] Figure 1 This is an optional structural diagram of a cell morphology analysis device provided in an embodiment of the present invention. Figure 1 The cell morphology analysis device 100 shown may include: a digital imaging device 101, a control device 102, a processor 103 and an output device 104. The digital imaging device 101 includes: a low-power objective lens, a high-power objective lens and a digital camera.

[0056] High-power and low-power objectives are two types of objective lenses used in cell morphology analysis equipment. Their magnifications are relative, with high-power objectives having a greater magnification than low-power objectives. When using these two types of objective lenses, a low-power objective is usually used first to identify and locate cells, and then a high-power objective is used to photograph the identified and located cells, allowing for further cell identification under the high-power objective. Both high-power and low-power objectives can magnify blood samples, but they have different magnifications. The specific magnification used depends on the objective lens used in the cell morphology analysis equipment, such as a 100X high-power objective and a 10X low-power objective.

[0057] After the high-power objective lens and the low-power objective lens amplify the blood sample, the digital camera is configured to capture the cell image of the blood sample under the low-power objective lens, or configured to capture the cell image under the high-power objective lens. In particular, the digital camera is configured to capture the suspected preset type of cells identified and located under the low-power objective lens under the high-power objective lens. The connection relationship diagram of the high-power objective lens, the low-power objective lens and the digital camera in the digital imaging device is shown as follows: Figure 2As shown, a high-magnification objective lens 1011 and a low-magnification objective lens 1012 are connected to a digital camera 1014 via an eyepiece 1013. Cells in a blood sample are located using the high-magnification objective lens 1011 and the low-magnification objective lens 1012. The digital camera 1014 captures the image located by the high-magnification objective lens 1011 or the low-magnification objective lens 1012 through the eyepiece 1013, thereby achieving either high-magnification or low-magnification photography. In this embodiment, the digital camera 1014 and the eyepiece 1013 can be replaced by an electronic eyepiece, utilizing the image acquisition function of the electronic eyepiece to achieve photography.

[0058] When a digital camera is shooting under a low-power or high-power objective lens, it can capture multiple images. The fields of view corresponding to these multiple images can be different, allowing the digital camera to capture more cell images under the low-power or high-power objective lens. Field of view control under the low-power or high-power objective lens can be achieved by the control device 102. One field of view control method is to adjust the relative position of the blood sample and the low-power or high-power objective lens by the control device 102, such as adjusting the position of at least one of the blood sample, the low-power or high-power objective lens to achieve the adjustment of the relative position. It should be noted that the different fields of view can be completely different fields of view or partially different fields of view. A completely different field of view can quickly acquire more and more comprehensive cell images. Although a partial difference in field of view is slower than a complete difference in field of view in acquiring more and more comprehensive cell images, it can acquire cell images of the same cell under different fields of view, thereby performing multiple recognition and analysis on the cell images of the same cell under different fields of view, thereby improving accuracy.

[0059] The processor 103 is configured to obtain a target number, identify and locate suspected cells of a preset type in a cell image captured under a low-power objective lens, obtain an identification result, determine the number of cells of the suspected preset type identified and located under the low-power objective lens based on the identification result and the target number, and instruct the digital camera to stop capturing the image under the low-power objective lens after completing the identification and location of the located number of cells of the suspected preset type. It should be noted that the suspected cells of the preset type in the cell image captured under the low-power objective lens refer to cells that are considered to be nucleated cells under the low-power objective lens, but do not represent true nucleated cells. For example, under the low-power objective lens, but not limited to, white blood cells, nucleated red blood cells, smear cells, large / giant platelets, sediment, platelet aggregates, and dust may be identified as suspected cells of the preset type.

[0060] The target number is the number of preset cell types that need to be photographed under a high-power objective lens. The preset cell types may include but are not limited to at least one type of cell among white blood cells, nucleated red blood cells, smear cells, large / giant platelets, sediment, platelet aggregates and dust. For example, the preset cell types include at least one of nucleated red blood cells and white blood cells.

[0061] During the actual sample analysis process, the preset cell type can be adjusted according to the sample analysis requirements. For example, if the sample analysis requirements require analysis of platelets, the preset cell type is set to platelets. In this embodiment, the setting and adjustment of the preset cell type can be carried out with the help of human-computer interaction instructions, and the currently specified preset cell type is carried in the human-computer interaction instructions. The human-computer interaction instructions include but are not limited to control instructions issued by an external control device, user voice instructions, user gesture instructions, and instructions directly input by the user in the interface of the cell morphology analysis device (such as the preset cell type input directly by the user in the interface of the cell morphology analysis device), etc.

[0062] After determining the preset cell types, the processor 103 needs to further determine the number of preset cell types that need to be photographed under the high-power objective lens (i.e., the target number). For different preset cell types, the corresponding target numbers may be different. For example, for white blood cells and nucleated red blood cells, the analysis of these two types of cells corresponds to different disease conditions, so it is necessary to set the corresponding target numbers for these two types of cells respectively.

[0063] The setting and adjustment of the target number is similar to the setting and adjustment of the preset cell type. The setting and adjustment of the target number can also be performed using human-computer interaction instructions, which carry the currently specified target number. For the format of the human-computer interaction instructions, please refer to the above description. In addition, the human-computer interaction instruction can carry both the currently specified preset cell type and the currently specified target number of cells of the preset cell type, thereby achieving a one-time setting of the preset cell type and target number. The target number can also be set by other means, such as setting it before the cell morphology analysis device leaves the factory. In this way, the preset cell type and corresponding target number can be directly used after the cell morphology analysis device leaves the factory.

[0064] The above content explains that one function of the target number is to limit the number of cells of a preset type that need to be photographed under a high-power objective lens, and another function of the target number is to assist in obtaining the number of suspected cells of a preset type identified and located under a low-power objective lens (i.e., the located number of the above-mentioned suspected cells of the preset type). For example, the process by which the processor determines the located number of suspected cells of the preset type is: identifying and locating the suspected cells of the preset type in the cell image captured under a low-power objective lens, obtaining an identification result, and determining the located number of suspected cells of the preset type identified and located under a low-power objective lens based on the identification result and the target number.

[0065] Among them, the processor can identify and locate suspected preset type cells in the cell image captured under the low-power objective lens and obtain the target number simultaneously or sequentially. The purpose of identification and positioning is mainly to identify and locate suspected preset type cells in the cell image, especially to identify and locate whether there are preset type cells. The recognition result obtained based on this identification and positioning is mainly used to indicate whether there are preset type cells among the suspected preset type cells identified and located in the cell image. Further, if there are preset type cells, it is also used to indicate how many preset type cells are identified and located. For example, the recognition result includes at least an estimated number, which is used to indicate the number of preset type cells identified and located under the low-power objective lens. The form of the estimated number includes but is not limited to the following methods:

[0066] One form of the estimated number is: the estimated number can be the number of preset type cells cumulatively identified at the current moment, that is, the estimated number is the sum of the number of preset type cells identified at the current moment and the number of preset type cells identified previously. Correspondingly, based on the recognition result and the target number, a way to determine the number of suspected preset type cells located is: if the current estimated number is greater than or equal to the target number, then the number of suspected preset type cells currently cumulatively identified is determined as the number of suspected preset type cells located. For example, if the current cumulative number of suspected preset type cells identified is 260, the current cumulative estimated number of preset type cells identified is 200, and the target number is also 200, then the number of preset type cells identified under the low-power objective lens at this time meets the requirement for continued identification under the high-power objective lens, then the number of suspected preset type cells located can be 260, and the determination of the number of suspected preset type cells located is completed at the same time as the identification and positioning of the suspected preset type cells is completed.

[0067] Another form of estimated number is: the estimated number is the number of cells of a preset type identified and estimated in a cell image captured in at least one field of view under a low-magnification objective lens, and the estimated number is used as a benchmark for the number of cells of the preset type identified from the cell image captured under the low-magnification objective lens to calculate the number of suspected cells of the preset type. For example, if 60 suspected cells of the preset type are identified and located from the cell image, and there are 30 cells of the preset type, this means that 30 cells of the preset type can be identified from the 60 suspected cells of the preset type. If the target number is 100, the number of suspected cells of the preset type located is: 100*60 / 30. Based on this, the calculation formula for the corresponding number of suspected cells of the preset type when the estimated number is used as the number benchmark is: target number*N / estimated number, where N is the number of suspected cells of the preset type identified and located when the estimated number is obtained.

[0068] By determining the number of cells suspected of being of the preset type, the number of cells of the preset type identified and located under a low-power objective lens can meet the target number requirement, so that images of the target number of cells of the preset type can be captured as much as possible under a high-power objective lens.

[0069] After determining the number of cells suspected of a preset type, the processor 103 controls the photography under the low-power objective lens based on the number of cells suspected of a preset type. For example, after the processor completes the identification and location of the number of cells suspected of a preset type, it instructs the digital camera to stop photography under the low-power objective lens. At this point, the control device 102 can switch the low-power objective lens to a high-power objective lens, thereby switching the blood sample from the field of view of the low-power objective lens to the field of view of the high-power objective lens, and continuing to use the digital camera to photograph the blood sample under the field of view of the high-power objective lens. Compared to photography with a low-power objective lens, the digital camera photographs the cells suspected of a preset type identified and located under the low-power objective lens under the high-power objective lens.

[0070] That is to say, the purpose of shooting under a low-power objective lens is to identify and locate the suspected preset type of cells from the blood sample, and the purpose of shooting under a high-power objective lens is to magnify the suspected preset type of cells identified and located at a high magnification to obtain images of the suspected preset type of cells, such as Figure 3 The figure shows the processor identifying and locating suspected preset type cells in the cell image taken under the low-magnification objective lens. Figure 3 The black dots in the figure represent the suspected preset type cells that have been identified and located. Figure 4 and Figure 5 Under high magnification objective lens Figure 3 An image is obtained by capturing cells of suspected preset types.

[0071] Correspondingly, the processor 103 is further configured to identify whether the cells of the suspected preset type photographed under the high-power objective lens are cells of the preset type, count the number of cells of the preset type photographed under the high-power objective lens, obtain a statistical value, and instruct the digital camera to stop photographing under the high-power objective lens when the statistical value satisfies: target number ≤ statistical value. It will be understood that, generally, the statistical value ≤ the number of cells located.

[0072] In this embodiment, all suspected preset type cells identified and located by the processor 103 may not all be preset type cells. For example, if the preset type cells are white blood cells, the identified and located nucleated red blood cells do not belong to the preset type cells. Therefore, the processor 103 needs to further analyze the suspected preset type cells photographed under the high-power objective lens to determine whether the suspected preset type cells are the preset type cells. For the processor to determine whether the suspected preset type cells are the preset type cells, the existing identification method for white blood cells, nucleated red blood cells and other cells can be referred to, which will not be elaborated in this embodiment.

[0073] The above statistical value is used to indicate the number of cells of the preset type identified under the high-power objective lens at the current moment, and the initial value of the statistical value can be 0. Every time the processor identifies a suspected preset type cell as a preset type cell, it will update the previous statistical value in the following manner: current statistical value = previous statistical value + the number of cells of the preset type identified this time. If the statistical value obtained by the processor meets the preset conditions, such as the target number ≤ statistical value, and further the target number ≤ statistical value ≤ positioning number, it means that the cell morphology analysis device has captured at least the target number of cells of the preset type. At this time, the digital camera can be instructed to stop shooting, and at the same time as instructing the digital camera to stop shooting, the processor needs to update the statistical value to the initial value, so that the next time the processor identifies the preset type cells in the blood sample, it will still count from the initial value, thereby improving the counting accuracy.

[0074] The output device 104 is used to output cell information of suspected preset type cells identified as preset type cells, wherein the cell information of suspected preset type cells of preset type cells indicates that the suspected preset type cells are preset type cells, such as the cell information of suspected preset type cells of preset type cells includes: images of suspected preset type cells of preset type cells, especially images of suspected preset type cells under a high-power objective lens.

[0075] In this embodiment, one form of the output device is: the output device can be an output interface (i.e., an electrical interface) for outputting digital / analog signals, capable of outputting signals to an external display device (i.e., a display device other than the independent cell morphology analysis device), such as being capable of outputting signals to an external display screen or projector, so as to output cell information of cells suspected of being cells of the preset type through the external display screen or projector, such as Figure 6 As shown, the cell information is transmitted to the projector via the output device, and the cell information is projected into a projection area by the projector. When the output device is implemented as an output interface, the output interface can be configured to connect to different display devices according to the usage environment. For example, at least two display screens can be connected to enable simultaneous detection of multiple people.

[0076] Another form of the output device is: the output device may include an output interface and a display screen, wherein the display screen is connected to the output interface and is used to receive the signal output by the output interface and display the cell information accordingly, such as Figure 7As shown, the cell morphology analysis device has a display screen, and the cell information is output to the display screen for display via an output interface. In addition, the output interface of the output device can also transmit the cell information to an external display device, so that the cell information can be displayed on the cell morphology analysis device while also being displayed on the external display device. Thus, the cell information can be displayed simultaneously on different display devices to meet the viewing needs of different users. For example, the external display device can be a terminal located in a doctor's office, so that the doctor can view the cell information of cells suspected of being a preset type of cell while the cell morphology analysis device is performing identification and analysis.

[0077] In an embodiment of the present invention, a cell morphology analysis device includes: a digital imaging device, a control device, a processor and an output device. The digital imaging device includes a low-magnification objective lens, a high-magnification objective lens and a digital camera. The digital camera captures a cell image of a blood sample under the low-magnification objective lens. The processor identifies and locates suspected preset type cells in the cell image to obtain an identification result. Based on the identification result and the target number of preset type cells that need to be photographed under the high-magnification objective lens, the number of suspected preset type cells identified and located under the low-magnification objective lens is determined. After completing the identification and positioning of the located number of suspected preset type cells, the digital camera photographs the suspected preset type cells identified and located under the high-magnification objective lens. The processor then identifies whether the suspected preset type cells photographed under the high-magnification objective lens are the preset type cells, so as to count the number of preset type cells photographed under the high-magnification objective lens to obtain a statistical value. When the statistical value satisfies: the target number ≤ the statistical value, the digital camera is instructed to stop photographing under the high-magnification objective lens, thereby ending the identification of the suspected preset type cells photographed under the high-magnification objective lens. At the end, the processor recognizes at least the target number of preset cell types, indicating that no matter how many or few interfering cells are in the blood sample, at least the target number of preset cell types can be captured under a high-power objective lens, meeting the clinical demand for preset cell types.

[0078] Existing cell morphology analysis equipment may fail to capture the target number of cells of a preset type, especially when there are a large number of suspected cells of the preset type. In this case, the existing cell morphology analysis equipment may continue to use a low-magnification objective lens to identify the cells, and then use a high-magnification objective lens to capture the target number of cells of the preset type, which increases the time cost. The cell morphology analysis equipment of this embodiment can capture at least the target number of cells of the preset type under the high-magnification objective lens, regardless of whether there are many or few interfering cells in the blood sample, thus reducing the time cost compared to existing cell morphology analysis equipment.

[0079] The following describes how to determine the number of positioning positions in this embodiment. In this embodiment, the methods for determining the number of positioning positions include but are not limited to the following methods:

[0080] One way to determine the number of locations is to obtain the probability information p of each suspected preset type cell being a preset type cell during the cell image recognition process. i , according to the probability information p of each suspected preset type cell being a preset type cell i and the number of targets, and calculate the number of localizations of suspected preset cell types.

[0081] The probability information is used to characterize the probability that the suspected preset type of cell is the preset type of cell. The probability information can be directly presented in the form of a probability value, such as the probability information has a value range of [0, 1]. Figure 4 The probability of smear cells being the preset type of cells is 0.023. Figure 5 is a white blood cell, and the probability that it is a preset type of cell is 0.997. Figures 8 to 11 They are sediment cells, platelets, nucleated red blood cells and platelet aggregates, and the probability of each being a preset type of cell is 0.006, 0.012, 0.047 and 0.002. The processor then calculates the number of locations of suspected preset type cells based on these probability information and the target number.

[0082] The processor determines the probability information p of each suspected preset type cell as a preset type cell. i An optional calculation method for calculating the number of suspected preset type cells is: according to the probability information p of each suspected preset type cell being a preset type cell i , calculate the estimated number of cells of a given type N is the number of suspected preset type cells identified and located at the current moment. Based on the estimated number of preset type cells and the target number, the number of located cells of the suspected preset type is calculated.

[0083] In this embodiment, N can be the number of suspected preset type cells identified in the cell image captured at the current moment, that is, N is the estimated number calculated based on a single cell image, and is calculated by the formula The number of cells of a predetermined type identified in a single cell image can be determined. A corresponding optional method for calculating the number of localized cells of a suspected predetermined type is: target number*N / estimated number.

[0084] In this embodiment, N may also be the number of suspected preset type cells cumulatively identified and located at the current moment, that is, the total number of suspected preset type cells identified in the cell image captured at the current moment and the cell images captured previously. To determine the total number of preset type cells cumulatively identified at the current moment, the corresponding optional method for calculating the number of suspected preset type cells located is:

[0085] Accumulate the probability information of all suspected preset type cells obtained at the current moment as preset type cells, and obtain the estimated number of preset type cells accumulated at the current moment When the estimated number S≥λ*target number is satisfied, the digital camera is instructed to stop shooting under the low-power objective lens. The current moment N is the positioning number, and λ is a constant. The purpose of setting λ is to avoid the situation in which, in certain special samples, even when the estimated number S reaches the target number, the number of cells of the preset type actually obtained does not reach the target number, thereby causing the test results to not meet clinical needs. In this embodiment, the value of λ satisfies the relationship: 1≤λ≤1.25, then λ can select a value between [1, 1.25], such as at least one of λ=1, λ=1.15, λ=1.2, and λ=1.25. The value of λ can also be increased in actual clinical needs, such as the maximum value of λ can be 1.5, that is, λ can select a value between [1, 1.5].

[0086] When the estimated number S ≥ λ * target number, it means that the processor has cumulatively identified at least the target number of cells of the preset type. At this time, N when calculating the estimated number can be determined as the positioning number, so that the identification and positioning of the suspected preset type cells of the positioning number can be completed at the same time as the positioning number is calculated.

[0087] In the process of calculating the number of locations based on the estimated number of cells of the preset type accumulated at the current moment, there are two situations: the first situation is that the sum of the estimated number of cells of the preset type accumulated before switching the field of view and the estimated number of cells of the preset type in the cell image of the current field of view is less than λ*target number; the second situation is that the sum of the estimated number of cells of the preset type accumulated before switching the field of view and the estimated number of cells of the preset type in the cell image of the current field of view is greater than or equal to λ*target number. The following optional processing methods can be used to handle these two situations:

[0088] Optional processing method for situation 1: If the sum of the estimated number of cells of the preset type accumulated before switching the field of view and the estimated number of cells of the preset type in the cell image of the current field of view is less than λ*target number, the processor 103 completes the identification and location of the suspected cells of the preset type in the cell image of the current field of view, updates the estimated number of cells of the preset type accumulated at the current moment, and instructs the control device 102 to control the low-magnification objective lens to switch to the next field of view. The digital camera continues to photograph the blood sample in the next field of view under the low-magnification objective lens.

[0089] For example, if the sum of the estimated numbers is 150 and the target number is 200, the value range of λ is [1, 1.25], and the value range of λ can even be extended to [1, 1.5]. Within this value range, no matter which value λ is used, the sum of the estimated numbers is less than λ*target number. The processor then updates the estimated number of cells of the preset type accumulated at the current moment to 150, and then the control device controls the low-magnification objective lens to switch the field of view, so as to continue to obtain the cell image of the blood sample in the next field of view through the digital camera for identification and positioning.

[0090] An optional processing method for situation two: the processor 103 calculates the difference between the estimated number before switching the field of view and the λ*target number, identifies and locates the suspected preset type cells in the cell image of the current field of view in turn, and obtains the estimated number of preset type cells in the current field of view at the current moment. If the estimated number of preset type cells in the current field of view at the current moment is greater than or equal to the difference, the identification and positioning is stopped, thereby stopping the identification and positioning of the remaining suspected preset type cells after at least the target number of preset type cells are identified, thereby reducing the identification and positioning time while meeting the requirement of the target number of preset type cells.

[0091] For example, if the processor has identified 190 cells of the preset type and the λ*target number is 200, then the difference between the estimated number before switching the field of view and the λ*target number is 10. Every time the processor obtains the probability information that a suspected preset type cell is a preset type cell, it will calculate the estimated number of cells of the preset type at the current moment in the current field of view (note that it is the current moment in the current field of view, not the cumulative number). If the estimated number of cells of the preset type at the current moment in the current field of view is greater than or equal to the difference, the identification and positioning of the remaining suspected cells of the preset type will be stopped. One thing to note here is that the estimated number of cells of the preset type at the current moment in the current field of view is the sum of the probability information identified at the current moment in the current field of view, and the sum may have a decimal part. Therefore, when the estimated number of cells of the preset type at the current moment in the current field of view is less than but close to the difference, it is still necessary to continue identification until it is greater than or equal to the difference.

[0092] Another optional processing method for situation two: the processor 103 identifies and locates all suspected preset type cells in the cell image of the current field of view, and obtains the estimated number of preset type cells accumulated at the current moment. The estimated number of preset type cells accumulated at the current moment is the sum of the estimated number of preset type cells accumulated before switching the field of view and the estimated number of preset type cells in the current field of view. If the estimated number of preset type cells accumulated at the current moment is ≥λ*target number, the digital camera is instructed to stop shooting under a low-power microscope after locating all suspected preset type cells in the cell image of the current field of view.

[0093] During this optional processing method, each time at least one suspected cell of a preset type is identified and located, the processor calculates the cumulative estimated number of cells of the preset type. If the cumulative estimated number of cells of the preset type is ≥ λ*target number, the processor continues to identify and locate the remaining suspected cells of the preset type. Although this increases the identification and location time, it increases the number of cells of the preset type identified and located, ensuring that at least the target number of cells of the preset type can be imaged under the high-power objective. For example, if λ*target number is 200, and the processor has identified 201.02 cells, the processor will continue to identify and locate the remaining suspected cells of the preset type.

[0094] In this embodiment, the processor determines the probability information p of each suspected preset type cell as a preset type cell. i Another optional calculation method for calculating the number of locations of suspected preset type cells is as follows: the processor 103 obtains the probability information p of each suspected preset type cell in the cell image of the first field of view being a preset type cell. i , according to the probability information p that each suspected preset type cell in the cell image of the first field of view is a preset type cell i and the number of targets, and calculate the number of localizations of suspected preset cell types.

[0095] The first field of view is a field of view under a low-power objective lens. The first field of view is not limited to the first field of view captured by the low-power objective lens, but the last field of view captured by the low-power objective lens is excluded as much as possible. The reason for excluding the last field of view is that if the last field of view is used as a criterion, it is necessary to obtain multiple cell images under low-power objective lenses beforehand, which will reduce processing efficiency. In this embodiment, the first field of view can be any field of view except the last field of view captured by the low-power objective lens, or before implementing this embodiment, multiple different blood samples are analyzed to determine a reference field of view that can represent the cell distribution of the blood samples. Relative to this reference field of view, the suspected preset type cells identified and located in other fields of view and the probability information of the suspected preset type cells being the preset type cells are similar. In this way, the differences between different fields of view can be ignored. Therefore, when implementing this embodiment, the reference field of view is used as the first field of view.

[0096] An optional method for the processor 103 to calculate the number of locations of suspected preset type cells based on the probability information and the target number in the first field of view is: for each suspected preset type cell obtained in the cell image of the first field of view, the probability information p of the preset type cell is respectively i Accumulate and get the estimated number of cells of the preset type in the first field of view Where N is the number of suspected preset type cells identified and located in the first field of view; the number of suspected preset type cells located = target number * N / S.

[0097] For example, the number of suspected preset type cells identified and located in the first field of view is 4, and the estimated number is 2, which means that there are 2 preset type cells among the 4 suspected preset type cells. If the target number is 4, 8 suspected preset type cells need to be identified and located before 4 preset type cells can be identified. Therefore, the calculation formula for the number of suspected preset type cells located can be determined as the above-mentioned location number = target number * N / S.

[0098] The above optional calculation method describes the calculation of the positioning number based on the cell image of one field of view. In the actual implementation process, the positioning number can also be calculated based on the cell images of two fields of view or more. The following is an example of the cell images of two fields of view, where the cell images of the two fields of view are cell images of two different fields of view. The corresponding processor calculates the probability information p of each suspected preset type cell according to the preset type cell. i The process of calculating the number of suspected preset cell types based on the number of targets is as follows:

[0099] The processor 103 obtains probability information that each cell suspected of a preset type in the cell image of the first field of view is a cell of the preset type; the control device 102 switches the field of view of the low-power objective lens so that the processor 103 obtains probability information that each cell suspected of a preset type in the cell image of the second field of view is a cell of the preset type;

[0100] The processor 103 identifies and locates all suspected preset type cells based on the first field of view and the second field of view, and the probability information p that all cells of the preset type are cells of the preset type. i and the number of targets, and calculate the number of localizations of suspected preset cell types.

[0101] The first field of view and the second field of view are each a field of view under a low-power objective lens. The first field of view and the second field of view are not limited to being the first field of view taken with the low-power objective lens, but are preferably excluded from being the last field of view taken with the low-power objective lens. The reason for excluding the last field of view is that if the last field of view is used as a criterion, multiple cell images taken with low-power objective lenses must be acquired beforehand, which reduces processing efficiency. In this embodiment, the first field of view and the second field of view can be any field of view other than the last field of view taken with the low-power objective lens, or, before implementing this embodiment, multiple different blood samples are analyzed to determine two reference fields of view that can represent the cell distribution of the blood samples. Relative to these two reference fields of view, the probability information of the suspected preset type cells identified and located in other fields of view and the suspected preset type cells being the preset type cells are similar. In this way, the differences between the different fields of view can be ignored. Therefore, when implementing this embodiment, these two reference fields of view are used as the first field of view and the second field of view.

[0102] Alternatively, the first field of view and the second field of view are two adjacent fields of view when photographed under a low-power objective lens, but these two adjacent fields of view are not necessarily the first field of view and the second field of view when photographed under a low-power objective lens. For example, according to the above description of the first field of view, a field of view when photographed with a low-power objective lens is selected as the first field of view, and after the first field of view is determined, the field of view next to the first field of view is used as the second field of view. When the processor determines the number of locations based on the first field of view and the second field of view, the processor 103 can calculate the number of locations of suspected preset type cells based on the probability information of each suspected preset type cell in the cell image of the first field of view being a preset type cell. If the number of locations of suspected preset type cells calculated based on the first field of view does not reach (such as less than) the target number, the processor 103 will continue to obtain the probability information of each suspected preset type cell in the cell image of the second field of view being a preset type cell, and recalculate the number of locations of suspected preset type cells based on the probability information of each suspected preset type cell identified and located under the first field of view and the second field of view being a preset type cell.

[0103] An optional method for the processor 103 to calculate the number of locations of suspected preset type cells based on the probability information and target number in the first field of view and the second field of view is: based on the probability information p of all suspected preset type cells identified and located in the first field of view and the second field of view, they are respectively the cells of the preset type. i , calculate the estimated number of cells of the preset type at the current moment Wherein N is the number of suspected preset type cells identified and located by the first field of view and the second field of view; then the number of located suspected preset type cells = target number * N / S.

[0104] That is, the estimated number of cells of the preset type is obtained by using the probability information of each suspected preset type cell in the first field of view and the second field of view being the preset type cell. Then, the located number of suspected preset type cells is obtained based on the estimated number, the number of suspected preset type cells located when the estimated number is obtained, and the target number.

[0105] When the processor 103 obtains the number of localized cells based on the cell images in at least one field of view, two special cases may occur: the first special case is that the sum of the number of cells of the suspected preset type before switching the field of view and the number of cells of the suspected preset type in the cell image in the current field of view is less than the number of localized cells; the second special case is that the sum of the number of cells of the suspected preset type before switching the field of view and the number of cells of the suspected preset type in the cell image in the current field of view is greater than or equal to the number of localized cells. The following optional processing methods can be used to handle these two special cases:

[0106] An optional processing method for special case one is: the processor 103 identifies and locates all suspected preset type cells in the cell image of the current field of view, and obtains the number of suspected preset type cells at the current moment. The number of suspected preset type cells at the current moment is the sum of the number of suspected preset type cells obtained before switching the field of view and the number of suspected preset type cells identified from the cell image of the current field of view. If the number of suspected preset type cells at the current moment is less than the located number, after identifying and locating all suspected preset type cells in the cell image of the current field of view, the control device is instructed to control the low-power objective lens to switch to the next field of view; the digital camera is also configured to continue to photograph the blood sample in the next field of view under the low-power objective lens.

[0107] For example, at the current moment, the number of suspected preset type cells is 150, and the located number is 200. At this time, the number of suspected preset type cells at the current moment is less than the located number, indicating that the processor has not completed the identification of the located number of suspected preset type cells. Then, the processor 103 instructs the control device to control the low-power objective lens to switch the field of view, so as to continue to obtain the cell image of the blood sample in the next field of view through the digital camera for identification and positioning.

[0108] An optional processing method for special case two is: the processor 103 identifies and locates the suspected preset type cells in the cell image of the current field of view in sequence, and obtains the number of all suspected preset type cells at the current moment (that is, the cumulative number of all suspected preset type cells identified at the current moment). If the number of all suspected preset type cells at the current moment reaches the positioning number, the identification and positioning of the remaining suspected preset type cells in the cell image of the current field of view is stopped, and the digital camera is instructed to stop shooting under the low-power objective lens, so that after the positioning number of suspected preset type cells is identified and positioned, the identification and positioning of the remaining suspected preset type cells is stopped, thereby reducing the identification and positioning time while meeting the requirement of positioning the number of suspected preset type cells.

[0109] For example, the number of all suspected preset type cells identified by the processor at the current moment is 220, and the number of located cells is 200, which means that the processor has identified the located number of suspected preset type cells. At this time, the processor stops identifying and locating the remaining suspected preset type cells.

[0110] Another optional processing method for special case 2 is: the processor 103 identifies and locates the suspected preset type of cells in the cell image of the current field of view, obtains the number of all suspected preset type of cells at the current moment, and if the number of all suspected preset type of cells at the current moment reaches the positioning number, after identifying and locating all suspected preset type of cells in the cell image of the current field of view, instructs the digital camera to stop capturing images under the low-power objective lens. In other words, compared with the optional processing method for special case 2 described above, it differs in that even if the number of all suspected preset type of cells at the current moment reaches the positioning number, it is still necessary to identify and locate all remaining suspected preset type of cells. Although this processing method increases the identification and positioning time, it increases the number of preset type of cells that are identified and located, ensuring that at least the target number of preset type of cells can be captured under the high-power objective lens.

[0111] For the three optional calculation methods for the above-mentioned located number of suspected preset type cells, an optional method for the processor of this embodiment to obtain the probability information that each suspected preset type cell is a preset type cell is: the processor 103 calls the probability analysis model to identify the cell image to obtain the probability information that each suspected preset type cell is a preset type cell.

[0112] For example, the probability analysis model is a deep neural network model obtained by training the cell images of historical blood samples and whether each suspected preset type of cell in the historical blood samples is a preset type of cell. For example, the deep neural network model can be an AlexNet model, and the probability analysis model is obtained by training the existing AlexNet model on the cell images of historical blood samples and whether each suspected preset type of cell in the historical blood samples is a preset type of cell. After the probability analysis model is trained, the processor 103 inputs the cell image of the currently captured blood sample into the probability analysis model to obtain the probability information of each suspected preset type of cell being a preset type of cell output by the probability analysis model.

[0113] In this embodiment, a threshold value can also be set in the probability analysis model. If the probability calculated by the probability analysis model is less than the threshold value, the probability analysis model will adjust the probability less than the threshold value to a fixed value (such as 0) and output it, and the probabilities greater than or equal to the threshold value will all be output according to the actual calculation results of the probability analysis model. For example, the set threshold value can be but not limited to 0.5. If the probability calculated by the probability analysis model is 0.1, the probability analysis model outputs 0. If the probability calculated by the probability analysis model is 0.6, the probability analysis model outputs 0.6. That is, the probability analysis model has a probability adjustment function, and the probability less than the threshold value is adjusted and output. In this way, during the estimated number calculation process, the impact of the probability less than the threshold value on the estimated number can be reduced, thereby improving the accuracy of the estimated number. This is because the suspected preset type cells with a probability less than the threshold value are very likely not the preset type cells, and adjusting their probability to 0 will automatically ignore this type of suspected preset type cells, thereby improving the accuracy of the estimated number.

[0114] Of course, the probability analysis model can also adopt other deep neural network models, which will not be illustrated one by one in this embodiment. However, it should be noted that the pixels of the cell image of the historical blood sample and the cell image of the current blood sample taken under a low-magnification objective lens are the same. For example, the pixels of the cell image can be but are not limited to 224*224.

[0115] In this embodiment, in addition to obtaining probability information based on the probability analysis model, the processor 103 can also obtain probability information in other ways. For example, the processor 103 compares the cell image with a preset reference image to obtain probability information that each suspected preset type of cell is a preset type of cell.

[0116] The preset reference image can be an image with a known percentage of cells of a preset type. The percentage of cells of the preset type can be obtained by calculating the area occupied by cells of the preset type in the preset reference image, and taking the ratio of the area to the area of ​​the preset reference image as the percentage of cells of the preset type. The percentage of cells of the preset type is regarded as the probability information of cells of the preset type in the preset reference image. One point that needs to be pointed out here is that the probability information of cells of the preset type in the preset reference image here is a total probability information of cells of the preset type. For example, the preset reference image can be an image without suspected cells of the preset type, then the corresponding probability information of cells of the preset type is approximately 0, or the preset reference image is an image in which suspected cells of the preset type are almost full, such as the probability information of cells of the preset type is close to 100%.

[0117] After obtaining a cell image of the blood sample under a low-power objective lens, the area of ​​the suspected preset type cells in the cell image that are cells of the preset type is obtained by identifying at least one aspect such as color and cell size in the cell image. Based on the currently obtained area, the area of ​​the preset type cells in the preset reference image, and the probability information of the preset type cells in the preset reference image, the probability information of each suspected preset type cell in the cell image at the current moment being a preset type cell is obtained. For example, if the ratio M between the currently obtained area and the area of ​​the preset type cells in the preset reference image is calculated, the probability information of each suspected preset type cell in the cell image at the current moment being a preset type cell = the probability information of the preset type cell in the preset reference image * M.

[0118] In this embodiment, another way for the processor 103 to determine the located number of suspected preset type cells is: the processor 103 locates and identifies the suspected preset type cells in the cell image in turn to identify whether the suspected preset type cells are the preset type cells, and counts the number of cells identified as the preset type under the low-magnification objective lens. When the number of cells identified as the preset type under the low-magnification objective lens reaches the target number, the digital camera is instructed to stop shooting under the low-magnification objective lens, and the number of suspected preset type cells that have been identified and located when the target number is reached is used as the located number of suspected preset type cells.

[0119] That is, in the process of identifying and locating suspected preset type cells under a low-power objective lens, the processor 103 further identifies whether the suspected preset type cells are preset type cells. If the probability information obtained by the processor 103 that the suspected preset type cells are preset type cells is greater than a preset threshold (such as but not limited to 0.8), it means that the suspected preset type cells are preset type cells, and a preset type cell is located. When the target number of preset type cells are identified and located, the number of suspected preset type cells that have been identified and located is the located number of suspected preset type cells, indicating that the target number of preset type cells has been identified from the located number of suspected preset type cells.

[0120] Because the range of the blood sample captured by the low-magnification objective lens in one field of view is limited, the processor 103 may find that the sum of the number of cells identified as the preset type in the previous field of view and the number of cells identified as the preset type in the cell image of the current field of view is less than the target number during one identification and positioning process. In this case, the processor 103 performs the following operations:

[0121] The processor 103 identifies and locates all suspected preset type cells in the cell image of the current field of view, and obtains the number of cells identified as preset type at the current moment. The number of cells identified as preset type at the current moment is the sum of the number of cells identified as preset type before switching the field of view and the number of cells identified as preset type in the cell image of the current field of view; if the number of cells identified as preset type at the current moment is less than the target number, after locating all suspected preset type cells in the cell image of the current field of view, the control device is instructed to control the low-power objective lens to switch to the next field of view; the digital camera is also configured to continue to shoot the blood sample in the next field of view under the low-power objective lens.

[0122] That is to say, if the cumulative number of all preset types of cells identified at the current moment is less than the target number, it is necessary to continue shooting under the field of view of the low-magnification objective lens to identify the target number of preset types of cells under the low-magnification objective lens. At this time, the control device will switch the field of view of the low-magnification objective lens, and the digital camera will continue to shoot the blood sample in the next field of view of the low-magnification objective lens.

[0123] If the number of cells identified as the preset type at the current moment is less than the target number, the processor 103 may stop identifying the remaining suspected cells of the preset type or continue identifying the remaining suspected cells of the preset type. This embodiment does not limit the operation.

[0124] Regarding the above cell morphology analysis device, another optional structure of the cell morphology analysis device is as follows: Figure 12 As shown, the apparatus may further include: a first memory 105. The first memory 105 is configured to store cell images of the blood sample captured under a low-magnification objective lens; an output device 104 includes a first display screen, and the first display screen is configured to display the cell images of the blood sample, so that the cell images of the blood sample can be viewed, and the user can decide whether to re-examine the blood sample through viewing.

[0125] The cell morphology analysis device may further include a second memory configured to store only images of cells suspected of being a predetermined type that were determined to be cells of the predetermined type and captured under a high-power objective lens; and a corresponding output device including a second display screen configured to display images of cells suspected of being a predetermined type that were determined to be cells of the predetermined type and captured under a high-power objective lens, thereby facilitating viewing of the cells suspected of being a predetermined type.

[0126] The second memory and the first memory can be the same memory, or even two storage spaces of the same memory, or two independent memories, so that the cell morphology analysis device can simultaneously store the cell image of the blood sample taken under a low-power objective lens and the image of the suspected preset type of cell determined to be a preset type of cell taken under a high-power objective lens. The corresponding first display screen and the second display screen can be two display areas of a display screen, which simultaneously display the cell image of the blood sample and the image of the suspected preset type of cell determined to be a preset type of cell in a split-screen manner, such as Figure 13 A display method shown can display the cell image of the blood sample and at least one image of a suspected preset type of cell that is determined to be a preset type of cell at the same time, or the cell morphology analysis device can display the cell image of the blood sample and the image of the suspected preset type of cell that is determined to be a preset type of cell in a time-sharing manner through a single display screen, or one of the first display screen and the second display screen is a display screen of the cell morphology analysis device, and the other is a display screen independent of the cell morphology analysis device. In this way, the cell image of the blood sample and the image of the suspected preset type of cell that is determined to be a preset type of cell can also be displayed simultaneously.

[0127] In addition, in the above-mentioned cell morphology analysis equipment, the digital camera can sequentially photograph the suspected preset type cells that are identified and located under a high-power objective lens; the corresponding processor simultaneously identifies whether the suspected preset type cells are the preset type cells during the process of the digital camera photographing the suspected preset type cells; after determining the target number of preset type cells, the digital camera is instructed to stop photographing under the high-power objective lens, thereby realizing the simultaneous photographing under the high-power objective lens and identifying the preset type cells, thereby improving efficiency.

[0128] On this basis, the cell morphology analysis equipment can also include: a third memory, the third memory is configured to store images of preset type cells and suspected preset type cells of non-preset type cells photographed under a high-power objective lens; the output device includes a third display screen, the third display screen is configured to display images of preset type cells and suspected preset type cells of non-preset type cells photographed under a high-power objective lens.

[0129] In this embodiment, the cell morphology analysis device can simultaneously include at least two of the above-mentioned first memory, second memory and third memory. These at least two memories can be two storage spaces of the same memory, or two independent memories, so that the cell morphology analysis device can store different types of images at the same time. The corresponding output device can include at least two of the above-mentioned first display screen, second display screen and third display screen. These at least two display screens can be two display areas of a display screen, and different types of images can be displayed simultaneously in a split-screen manner, or the cell morphology analysis device can display different types of images in a time-sharing manner through a display screen, or one of the at least two display screens is the display screen of the cell morphology analysis device, and the other is a display screen independent of the cell morphology analysis device. In this way, different types of images can also be displayed simultaneously.

[0130] See also Figure 14 , which shows an optional flow chart of a cell identification method provided by an embodiment of the present invention. The cell identification method is applied to the above-mentioned cell morphology analysis device and may include the following steps:

[0131] 201: Obtain a target number, where the target number is the number of cells of a preset type that the cell morphology analysis device needs to capture under a high-power objective lens. For example, suspected cells of a preset type in a blood sample are cells with nuclei in the blood sample, including but not limited to white blood cells, nucleated red blood cells, smear cells, large / giant platelets, sediment, and platelets, etc., while the preset cell type is at least one type of cell selected from these suspected cells of the preset type. For example, the preset cell type includes at least one of nucleated red blood cells and white blood cells. The preset cell type and the target number of the preset cell type can be adjusted via human-computer interaction instructions. For details, please refer to the relevant description in the above-mentioned device embodiment, which will not be elaborated in this embodiment.

[0132] 202: Take a cell image of a blood sample under a low-magnification objective.

[0133] 203: Identify suspected preset type cells in the localized cell image, obtain an identification result, and determine the number of suspected preset type cells identified and located under a low-power objective lens based on the identification result and the target number.

[0134] Among them, the processor can identify and locate suspected preset type cells in the cell image captured under the low-magnification objective lens and obtain the target number simultaneously or sequentially. The purpose of identification and positioning is mainly to identify and locate suspected preset type cells in the cell image, especially to identify and locate whether there are preset type cells. The recognition result obtained based on this identification and positioning is mainly used to indicate whether there are preset type cells among the suspected preset type cells identified and located in the cell image. Further, if there are preset type cells, it is also used to indicate how many preset type cells are identified and located. For example, the recognition result includes at least an estimated number, which is used to indicate the number of preset type cells identified and located under the low-magnification objective lens. For a description of the estimated number, please refer to the above-mentioned device embodiment.

[0135] Moreover, when shooting under a low-magnification objective lens, the field of view of the low-magnification objective lens covers a limited range of blood samples. Therefore, when shooting under a low-magnification objective lens, it is necessary to switch the field of view of the low-magnification objective lens. Every time a cell image of a blood sample is taken, the cell image needs to be identified and located to determine the number of cells suspected of being of a preset type.

[0136] In this embodiment, one way to determine the number of cells suspected of being of a preset type is to obtain the probability information p of each cell suspected of being of a preset type during the cell image recognition process. i ; According to the probability information p of each suspected preset type cell being a preset type cell i and the target number, and calculate the number of cells suspected of being of a preset type. Optional calculation methods include but are not limited to the following methods:

[0137] An optional calculation method is: according to the probability information p of each suspected preset type cell being a preset type cell i , calculate the estimated number of cells of a given type Based on the estimated number of cells of the preset type and the target number, the number of cells suspected of the preset type is calculated. For example, the probability information of all cells suspected of the preset type obtained at the current moment are accumulated to obtain the estimated number of cells of the preset type accumulated at the current moment. When the estimated number S≥λ*target number is satisfied, stop shooting under the low-magnification objective lens. The current moment N is the positioning number, where λ is a constant, preferably 1≤λ≤1.25, and the value of λ can even be increased, such as 1≤λ≤1.5.

[0138] Under this optional calculation method, the cell identification method provided in this embodiment may also include: if the sum of the estimated number of preset type cells accumulated before switching the field of view and the estimated number of preset type cells in the current field of view cell image is less than λ*target number, after the identification and positioning of the suspected preset type cells in the current field of view cell image is completed, the estimated number of preset type cells accumulated at the current moment is updated, and the low-power objective lens is controlled to switch to the next field of view to continue photographing the blood sample in the next field of view under the low-power objective lens.

[0139] And / or, the cell identification method provided in this embodiment may further include: calculating the difference between the estimated number before switching the field of view and the λ*target number, sequentially identifying and locating suspected preset type cells in the cell image of the current field of view, obtaining the estimated number of preset type cells in the current field of view at the current moment, and stopping identification and locating if the estimated number of preset type cells in the current field of view at the current moment is greater than or equal to the difference;

[0140] And / or, the cell identification method provided in this embodiment may also include: identifying and locating all suspected preset type cells in the cell image of the current field of view, and obtaining the estimated number of preset type cells accumulated at the current moment, the estimated number of preset type cells accumulated at the current moment is the sum of the estimated number of preset type cells accumulated before switching the field of view and the estimated number of preset type cells in the current field of view; if the estimated number of preset type cells accumulated at the current moment is ≥λ*target number, then stop shooting under a low-power microscope after locating all suspected preset type cells in the cell image of the current field of view.

[0141] Another optional calculation method is to obtain the probability information p of each suspected preset type cell in the cell image of the first field of view being a preset type cell. i According to the probability information p that each suspected preset type cell in the cell image of the first field of view is a preset type cell i and the number of targets, and calculate the number of localizations of suspected preset cell types.

[0142] For example, the probability information p of each suspected preset type cell obtained in the cell image of the first field of view is respectively the preset type cell i Accumulate and get the estimated number of cells of the preset type in the first field of view Where N is the number of suspected preset type cells identified and located in the first field of view; the number of suspected preset type cells located = target number * N / S.

[0143] Another optional calculation method is: obtaining the probability information of each suspected preset type cell in the cell image of the first field of view being a cell of the preset type; switching the field of view to obtain the probability information of each suspected preset type cell in the cell image of the second field of view being a cell of the preset type; and calculating the probability information p of all the suspected preset type cells identified and located in the first field of view and the second field of view being cells of the preset type. i and the number of targets, and calculate the number of localizations of suspected preset cell types.

[0144] For example, all the suspected preset type cells identified and located based on the first field of view and the second field of view are the probability information p of the preset type cells. i , get the estimated number of cells of the preset type at the current moment Where N is the number of cells of the suspected preset type identified and located by the first field of view and the second field of view; the number of cells of the suspected preset type located = the number of targets * N / S.

[0145] For the above two methods of determining the located number of suspected preset type cells based on the cell image under the field of view, the above cell identification method provided by this embodiment may also include: identifying and locating all suspected preset type cells in the cell image of the current field of view, and obtaining the number of suspected preset type cells at the current moment. The number of suspected preset type cells at the current moment is the sum of the number of suspected preset type cells obtained before switching the field of view and the number of suspected preset type cells identified from the cell image of the current field of view. If the number of suspected preset type cells at the current moment is less than the located number, after identifying and locating all suspected preset type cells in the cell image of the current field of view, control the low-power objective lens to switch to the next field of view to continue photographing the blood sample in the next field of view under the low-power objective lens.

[0146] And / or, the above-mentioned cell identification method provided in this embodiment may also include: identifying and locating the suspected preset type cells in the cell image of the current field of view in sequence, and obtaining the number of all suspected preset type cells at the current moment (that is, the number of all suspected preset type cells cumulatively identified at the current moment); if the number of all suspected preset type cells at the current moment reaches the positioning number, stop identifying and locating the remaining suspected preset type cells in the cell image of the current field of view, and stop shooting under the low-magnification objective lens.

[0147] And / or, the above-mentioned cell identification method provided in this embodiment may also include: identifying and locating suspected preset type cells in the cell image of the current field of view, obtaining the number of all suspected preset type cells at the current moment, if the number of all suspected preset type cells at the current moment reaches the positioning number, identifying and locating all suspected preset type cells in the cell image of the current field of view and stopping shooting under the low-magnification objective lens.

[0148] One method for calculating the probability information that each of the suspected preset cell types is a cell of the preset type is to call a probability analysis model to identify the cell image to obtain the probability information that each of the suspected preset cell types is a cell of the preset type. The probability analysis model can be a deep neural network model, such as an AlexNet model, and the pixel size of the cell image can be, but is not limited to, 224*224. Another method for calculating the probability information is to compare the cell image with a preset reference image to obtain the probability information that each of the suspected preset cell types is a cell of the preset type.

[0149] In this embodiment, another way to determine the located number of suspected preset type cells is: locate and identify the suspected preset type cells in the cell image in sequence; identify whether the suspected preset type cells are the preset type cells, and count the number of cells identified as the preset type under a low-power objective lens; when the number of cells identified as the preset type under the low-power objective lens reaches the target number, stop shooting under the low-power objective lens, and use the number of suspected preset type cells that have been identified and located when the target number is reached as the located number of suspected preset type cells.

[0150] Because the range of the blood sample captured by a low-magnification objective lens in one field of view is limited, during one identification and positioning process, there may be a situation where the sum of the number of cells identified as the preset type in the previous field of view and the number of cells identified as the preset type in the cell image of the current field of view is less than the target number. To address this situation, the cell identification method provided in this embodiment further performs the following steps:

[0151] All suspected preset type cells in the cell image of the current field of view are identified and located to obtain the number of cells identified as the preset type at the current moment. The number of cells identified as the preset type at the current moment is the sum of the number of cells identified as the preset type before switching the field of view and the number of cells identified as the preset type in the cell image of the current field of view. If the number of cells identified as the preset type at the current moment is less than the target number, after locating all suspected preset type cells in the cell image of the current field of view, control the low-power objective lens to switch to the next field of view, and continue to shoot the blood sample in the next field of view under the low-power objective lens.

[0152] For a detailed description of the above-mentioned calculation of the number of locations of cells suspected of a preset type, please refer to the above-mentioned device embodiment, which will not be elaborated in this embodiment.

[0153] 204: After completing the identification and positioning of a number of cells suspected of being of a preset type, stop photographing under the low-magnification objective lens.

[0154] 205: Switch from a low-magnification objective lens to a high-magnification objective lens, and use the high-magnification objective lens to photograph the suspected preset cell type identified and located using the low-magnification objective lens. The difference between photographing with a low-magnification objective lens and photographing with a digital camera using a high-magnification objective lens is that photographing with a low-magnification objective lens photographs the suspected preset cell type identified and located using the low-magnification objective lens. In other words, photographing with a low-magnification objective lens is used to identify and locate the suspected preset cell type in the blood sample, while photographing with a high-magnification objective lens is used to magnify the identified and located suspected preset cell type at a high magnification to obtain an image of the suspected preset cell type.

[0155] 206: Identify whether the suspected preset type cells photographed under the high-power objective lens are the preset type cells, and count the number of the preset type cells photographed under the high-power objective lens to obtain a statistical value.

[0156] The statistical value is used to indicate the number of cells of the preset type identified under the high-power objective lens at the current moment. The initial value of the statistical value can be 0. When the processor identifies a suspected preset type of cell as a preset type of cell, it will update the previous statistical value in the following way: current statistical value = previous statistical value + number of preset type cells identified this time.

[0157] 207: When the statistical value satisfies the condition: target number ≤ statistical value, cease capturing under the high-power objective lens. If the statistical value satisfies the condition: target number ≤ statistical value, it indicates that the cell morphology analysis device has captured at least the target number of cells of the preset type. Capturing can be stopped at this point, and the statistical value is updated to the initial value simultaneously with the cessation of capturing. This ensures that the next identification of the preset cell type in the blood sample will still start from the initial value, thereby improving counting accuracy. It is understood that, generally, the statistical value ≤ the number of locations. Furthermore, the condition: target number ≤ statistical value ≤ number of locations can be used as a condition for terminating capturing.

[0158] 208: Outputting cell information of cells suspected of being of a predetermined type that are identified as cells of a predetermined type. The cell information of cells suspected of being of a predetermined type indicates that the cells suspected of being of a predetermined type are cells of a predetermined type. For example, the cell information of cells suspected of being of a predetermined type includes an image of the cells suspected of being of a predetermined type, particularly an image of the cells suspected of being of a predetermined type under a high-power objective lens. For details on how to output the cell information, please refer to the above-described device embodiments and will not be further elaborated in this embodiment.

[0159] In addition, the cell identification method provided in this embodiment may further include: storing the cell image of the blood sample captured under a low-magnification objective lens; and displaying the cell image of the blood sample.

[0160] and / or

[0161] Only images of cells suspected of being of a preset type that are determined to be cells of a preset type and taken under a high-power objective lens are stored; and images of cells suspected of being of a preset type that are determined to be cells of a preset type and taken under a high-power objective lens are displayed.

[0162] and / or

[0163] Under a high-power objective lens, the identified and located cells of suspected preset type are photographed sequentially; during the photographing process of the suspected preset type cells, it is simultaneously identified whether the suspected preset type cells are cells of the preset type; after determining the target number of cells of the preset type, photographing under the high-power objective lens is stopped, thereby achieving simultaneous photographing under the high-power objective lens and identifying the preset type cells, thereby improving efficiency. On this basis, the cell identification method may also include: storing images of the preset type cells and non-preset type cells of suspected preset type cells photographed under the high-power objective lens; and displaying images of the preset type cells and non-preset type cells of suspected preset type cells photographed under the high-power objective lens.

[0164] See also Figure 15 , which shows another optional flow chart of the cell identification method provided by an embodiment of the present invention. The cell identification method is also applied to the above-mentioned cell morphology analysis device and may include the following steps:

[0165] 301: Obtaining a target number, wherein the target number is the number of cells of a preset type that need to be photographed under a high-power objective lens in a cell morphology analysis device.

[0166] 302: Acquire a cell image of a blood sample taken under a low-magnification objective lens.

[0167] 303: Identify cells of suspected preset types in the cell image, and identify and locate the cells of suspected preset types in sequence.

[0168] 304: Obtain information that all suspected cells of the preset type are identified as cells of the preset type at the current moment to determine whether it is necessary to locate and identify the next suspected cell of the preset type.

[0169] 305: When it is necessary to locate and identify the next suspected cell of the preset type, locate and identify the next suspected cell of the preset type.

[0170] 306: When there is no need to locate and identify the next suspected preset type of cell, stop shooting under the low-magnification objective lens and switch the low-magnification objective lens to the high-magnification objective lens.

[0171] 307: Acquire the suspected preset type cells identified and located under the low magnification objective lens and photographed under the high magnification objective lens, determine whether the suspected preset type cells photographed under the high magnification objective lens are the preset type cells, and stop photographing under the high magnification objective lens when the number of cells of the preset type meets the target number. For example, if the number of cells of the preset type is the same as the target number, or the number of cells of the preset type is greater than the target number, stop photographing under the high magnification objective lens.

[0172] 308: Outputting cell information of cells suspected of a preset type that are identified as cells of a preset type.

[0173] Figure 15 The cell identification method shown is relative to the above Figure 14 The difference of the cell identification method shown is that: it is explained from the perspective of the sequential identification and positioning process of suspected preset type cells. The specific process is that after obtaining a cell image of a blood sample taken under a low-magnification objective lens, the suspected preset type cells in the cell image are sequentially identified and positioned to determine the information that all suspected preset type cells at the current moment are identified as preset type cells, and to determine whether to identify and position the next suspected preset type cell, such as whether to identify and position the next suspected preset type cell in the currently acquired cell image, or whether to switch the field of view of the low-magnification objective lens to continue to obtain another cell image under the low-magnification objective lens, and sequentially identify and position the suspected preset type cells in the other cell image. The information that all suspected preset type cells at the current moment are identified as preset type cells can indicate the number of preset type cells at the current moment, such as the estimated number mentioned above. For specific details, please refer to the relevant description in the above-mentioned device embodiment.

[0174] In this embodiment, an optional method for obtaining information about cells of a preset type to determine whether the next suspected cell of the preset type needs to be located and identified is:

[0175] Locate the current suspected preset type cell, obtain the probability information of the current suspected preset type cell being identified as the preset type cell and the probability information accumulated before the current suspected preset type cell is identified; update the accumulated probability information based on the probability information of the current suspected preset type cell being identified as the preset type cell and the probability information accumulated before the current suspected preset type cell is identified, and determine whether it is necessary to locate the next suspected preset type cell based on the accumulated probability information.

[0176] The probability information of a suspected preset type cell being identified as a preset type cell can characterize whether the suspected preset type cell is a preset type cell. The probability information can be obtained, but is not limited to, through the above-mentioned probability analysis model. After obtaining the current probability information and the previously accumulated probability information, the accumulated probability information can be updated by calculating the sum of the probability information. The accumulated probability information can indicate the current cumulative number of identified preset type cells, such as the above-mentioned estimated number, so as to determine whether the next suspected preset type cell needs to be located based on the estimated number. If the estimated number ≥ λ * target number, it indicates that the next suspected preset type cell does not need to be located; otherwise, the next suspected preset type cell needs to be located.

[0177] In this embodiment, another optional method for obtaining information on preset type cells to determine whether the next suspected preset type cell needs to be located and identified is: locating the current suspected preset type cell, identifying whether the current suspected preset type cell is the preset type cell; updating the number of suspected preset type cells identified as preset type cells to determine whether the next suspected preset type cell needs to be located and identified.

[0178] The purpose of locating the current suspected preset type cell is to identify whether the current suspected preset type cell is a preset type cell. If the probability information that the suspected preset type cell is a preset type cell is greater than a preset threshold (such as but not limited to 0.8), it means that the suspected preset type cell is a preset type cell, and a preset type cell is located. The number of suspected preset type cells identified as preset type cells is updated (i.e., the number of suspected preset type cells). When the number of suspected preset type cells indicates that the target number of preset type cells are identified and located, the number of suspected preset type cells that have been identified and located is the located number of suspected preset type cells, indicating that the target number of preset type cells have been identified from the located number of suspected preset type cells, and the positioning and identification of the suspected preset type cells can be stopped. Otherwise, the positioning and identification of the next suspected preset type cell is required.

[0179] Through the above-mentioned cell recognition method, at least the target number of preset type cells can be identified when the shooting under the high-power objective lens is completed. This means that no matter how many or few interfering cells are in the blood sample, at least the target number of preset type cells can be photographed under the high-power objective lens, meeting the clinical demand for preset type cells.

[0180] See also Figure 16 , which shows another optional structural schematic diagram of the cell morphology analysis device provided by an embodiment of the present invention, Figure 16The cell morphology analysis device 400 shown may include a processor 401 and a memory 402, and may also include at least one network interface 403 and a user interface 404. The various components in the cell morphology analysis device 400 are coupled together via a bus system 405. It is understood that the bus system 405 is used to achieve connection and communication between these components. In addition to including a data bus, the bus system 405 also includes a power bus, a control bus, and a status signal bus. However, for the sake of clarity, the bus system 405 is not described in detail. Figure 16 Various buses are labeled as bus system 405 .

[0181] The user interface 404 may include a display, a keyboard, a mouse, a trackball, a click wheel, keys, buttons, a touch pad, or a touch screen.

[0182] It is understood that memory 402 can be volatile memory or non-volatile memory, or can include both volatile and non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), ferromagnetic random access memory (FRAM), flash memory, magnetic surface memory, optical disk, or compact disc read-only memory (CD-ROM); magnetic surface memory can be magnetic disk memory or tape memory. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), synchronous static random access memory (SSRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct rambus random access memory (DRRAM).The memory 402 described in the embodiments of the present invention is intended to include these and any other suitable types of memory.

[0183] The memory 402 in this embodiment of the present invention includes, but is not limited to, ternary content addressable memory (TCAM) and static random access memory (SRAM) capable of storing various types of data, such as received cell images, to support the operation of the cell morphology analysis device 400. Examples of such data include any computer program used to operate on the cell morphology analysis device 400, such as an operating system 4021 and application programs 4022, stored images, target numbers, and location numbers. The operating system 4021 includes various system programs, such as a framework layer, a core library layer, and a driver layer, for implementing various basic services and processing hardware-based tasks. The application program 4022 may include executable instructions that, when executed, execute the cell recognition method described above.

[0184] The methods disclosed in the above embodiments of the present invention can be implemented by processor 401. Processor 401 can be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by hardware integrated logic circuits in processor 401 or by software operations. The above processor 401 can be a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Processor 401 can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of the present invention can be directly implemented and executed by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium located in memory 402. Processor 401 reads information from memory 402 and, in conjunction with its hardware, completes the corresponding steps described above.

[0185] In an exemplary embodiment, the present invention further provides a storage medium having executable instructions stored thereon, configured to cause a processor to implement the above-mentioned cell identification method when executing the executable instructions.

[0186] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, devices, or storage media. Therefore, embodiments of the present invention may take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Furthermore, embodiments of the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including magnetic disk storage and optical storage, etc.) containing computer-usable program code.

[0187] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program operations. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the operations performed by the processor of the computer or other programmable data processing device produce the operations for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0188] These computer program operations may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the operations stored in the computer readable memory produce an article of manufacture including an operating device that implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0189] These computer program operations can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the operations executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0190] The above description is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A cell morphology analysis device, comprising: digital imaging devices, control devices, processors, and output devices; The digital imaging device comprises: a low-power objective lens, a high-power objective lens and a digital camera; The digital camera is configured to capture cell images of the blood sample under the low-magnification objective lens; The processor is configured to obtain a target number, the target number being the number of cells of a preset type that need to be photographed under the high-power objective lens in the cell morphology analysis device, the processor identifying and locating suspected cells of the preset type in the cell image to obtain an identification result, and based on the identification result and the target number, determining the number of cells of the suspected preset type identified and located under the low-power objective lens, and after completing the identification and positioning of the number of cells of the suspected preset type, instructing the digital camera to stop photographing under the low-power objective lens; The control device is configured to switch the low-power objective lens to the high-power objective lens after the processor completes the identification and positioning of the number of cells suspected of the preset type; The digital camera is further configured to photograph the suspected preset type of cells identified and located under the low-magnification objective lens under the high-magnification objective lens; The processor is further configured to identify whether the cells of the suspected preset type photographed under the high-power objective lens are the cells of the preset type, count the number of cells of the preset type photographed under the high-power objective lens to obtain a statistical value, and instruct the digital camera to stop photographing under the high-power objective lens when the statistical value satisfies: target number ≤ statistical value; The output device is used to output cell information of cells suspected of a preset type that are identified as cells of a preset type.

2. The device according to claim 1, wherein The processor is configured to obtain probability information of each suspected preset type cell being a preset type cell during the cell image recognition process. , according to the probability information of each suspected preset type cell being a preset type cell and the target number, and calculating the localization number of the suspected preset type of cells.

3. The device according to claim 2, wherein The processor is configured to determine the probability information of each suspected preset type cell being a preset type cell. , calculate the estimated number of cells of a preset type , N is the number of suspected preset type cells identified and located at the current moment. Based on the estimated number of the preset type cells and the target number, the number of located cells of the suspected preset type is calculated.

4. The device according to claim 3, wherein The processor is configured to accumulate the probability information of all suspected preset type cells obtained at the current moment being preset type cells, and obtain the estimated number of preset type cells accumulated at the current moment. , when the estimated number is met When the number of targets is reached, the digital camera is instructed to stop shooting under the low-magnification objective lens, N at the current moment is the number of positioning, and λ is a constant.

5. The device according to claim 4, wherein 1≤λ≤1.

25.

6. The device according to claim 4, wherein The processor is configured to: if the sum of the estimated number of cells of the preset type accumulated before switching the field of view and the estimated number of cells of the preset type in the cell image of the current field of view is less than target number, after the suspected preset type of cells in the current field of view cell image are identified and located, the estimated number of preset type of cells accumulated at the current moment is updated, and the control device is instructed to control the low-power objective lens to switch to the next field of view; The digital camera is further configured to continue photographing the blood sample in the next field of view under the low-power objective lens.

7. The apparatus according to claim 4 or 6, wherein: The processor is configured to calculate the estimated number and The difference between the target numbers is used to identify and locate the suspected preset type cells in the cell image of the current field of view in sequence to obtain the estimated number of preset type cells in the current field of view at the current moment. If the estimated number of preset type cells in the current field of view at the current moment is greater than or equal to the difference, identification and positioning is stopped.

8. The apparatus according to claim 4 or 6, wherein: The processor is configured to identify and locate all suspected preset type cells in the cell image of the current field of view, and obtain the estimated number of preset type cells accumulated at the current moment, wherein the estimated number of preset type cells accumulated at the current moment is the sum of the estimated number of preset type cells accumulated before switching the field of view and the estimated number of preset type cells in the current field of view. If the estimated number of preset type cells accumulated at the current moment is The target number is determined by the digital camera, which instructs the digital camera to stop shooting under the low-magnification objective lens after locating all cells of the suspected preset type in the cell image of the current field of view.

9. The device according to claim 2, wherein The processor is configured to obtain probability information that each suspected preset type of cell in the cell image of the first field of view is a preset type of cell. , according to the probability information that each suspected preset type cell in the cell image of the first field of view is a preset type cell and the target number, and calculating the localization number of the suspected preset type of cells.

10. The device according to claim 9, wherein The processor is configured to generate probability information of each suspected preset type cell obtained in the cell image of the first field of view being a preset type cell. Accumulate and get the estimated number of cells of the preset type in the first field of view , where N is the number of suspected preset type cells identified and located in the first field of view; Then the number of cells suspected of being of the preset type located = the number of targets .

11. The device according to claim 2, wherein The processor is configured to obtain probability information that each suspected preset type of cell in the cell image of the first field of view is a cell of the preset type; The control device is further configured to switch the field of view of the low-power objective lens so that the processor obtains probability information that each suspected preset type of cell in the cell image of the second field of view is a preset type of cell; The processor is configured to identify and locate all suspected preset type cells based on the first field of view and the second field of view, and to provide probability information that all cells of the preset type are cells of the preset type. and the target number, and calculating the localization number of the suspected preset type of cells.

12. The apparatus according to claim 11, wherein The processor is configured to identify and locate all suspected preset type cells based on the first field of view and the second field of view, and to provide probability information that all cells of the preset type are cells of the preset type. , calculate the estimated number of cells of the preset type at the current moment , where N is the number of cells of suspected preset type identified and located by the first field of view and the second field of view; Then the number of cells suspected of being of the preset type located = the number of targets .

13. The apparatus according to any one of claims 9 to 12, wherein: The processor is configured to identify and locate all suspected preset type cells in the cell image of the current field of view, obtain the number of suspected preset type cells at the current moment, and the number of suspected preset type cells at the current moment is the sum of the number of suspected preset type cells obtained before switching the field of view and the number of suspected preset type cells identified from the cell image of the current field of view. If the number of suspected preset type cells at the current moment is less than the located number, after identifying and locating all suspected preset type cells in the cell image of the current field of view, instruct the control device to control the low-power objective lens to switch to the next field of view; The digital camera is further configured to continue photographing the blood sample in the next field of view under the low-power objective lens.

14. The apparatus according to any one of claims 9 to 12, wherein: The processor is configured to sequentially identify and locate cells of suspected preset types in the cell image of the current field of view, and obtain the number of all suspected cells of the preset type at the current moment. If the number of all suspected cells of the preset type at the current moment reaches the located number, the processor stops identifying and locating the remaining suspected cells of the preset type in the cell image of the current field of view, and instructs the digital camera to stop shooting under the low-magnification objective lens.

15. The apparatus according to any one of claims 9 to 12, wherein: The processor is configured to identify and locate suspected preset type cells in the cell image of the current field of view, obtain the number of all suspected preset type cells at the current moment, and if the number of all suspected preset type cells at the current moment reaches the positioning number, after identifying and locating all suspected preset type cells in the cell image of the current field of view, instruct the digital camera to stop shooting under the low-magnification objective lens.

16. The apparatus according to any one of claims 2 to 6 and 9 to 12, wherein: The processor is configured to call a probability analysis model to identify the cell image to obtain probability information that each suspected preset type of cell is a preset type of cell.

17. The apparatus according to claim 16, wherein The probability analysis model is a deep neural network model.

18. The apparatus according to claim 17, wherein The deep neural network model is the AlexNet model, and the pixels of the cell image are .

19. The apparatus according to any one of claims 2 to 6 and 9 to 12, wherein: The processor is configured to compare the cell image with a preset reference image to obtain probability information that each suspected preset type of cell is a preset type of cell.

20. The apparatus of claim 1, wherein The processor is configured to sequentially locate and identify cells of suspected preset types in the cell image to identify whether the cells of suspected preset types are cells of preset types, count the number of cells identified as cells of preset types under a low-magnification objective lens, and when the number of cells identified as cells of preset types under a low-magnification objective lens reaches the target number, instruct the digital camera to stop shooting under the low-magnification objective lens, and use the number of suspected cells of preset types that have been identified and located when the target number is reached as the located number of the suspected cells of preset types.

21. The apparatus according to claim 20, wherein The device further includes: if the sum of the number of cells identified as the preset type in the previous field of view and the number of cells identified as the preset type in the cell image of the current field of view is less than the target number, the device includes: The processor is configured to identify and locate all cells of a suspected preset type in the cell image of the current field of view, and obtain the number of cells identified as the preset type at the current moment, wherein the number of cells identified as the preset type at the current moment is the sum of the number of cells identified as the preset type before switching the field of view and the number of cells identified as the preset type in the cell image of the current field of view; if the number of cells identified as the preset type at the current moment is less than the target number, after locating all cells of the suspected preset type in the cell image of the current field of view, instruct the control device to control the low-power objective lens to switch to the next field of view; The digital camera is further configured to continue photographing the blood sample in the next field of view under the low-power objective lens.

22. The apparatus according to any one of claims 1 to 6, 9 to 12, 20 to 21, wherein: The device further comprises: a first memory configured to store the cell image of the blood sample captured under the low-magnification objective lens; The output device includes a first display screen configured to display a cell image of the blood sample.

23. The apparatus according to any one of claims 1 to 6, 9 to 12, 20 to 21, wherein: The device further comprises: The second memory is configured to store only images of cells suspected of a preset type that are determined to be cells of a preset type and are taken under a high-power objective lens; The output device includes a second display screen configured to display images of suspected cells of a preset type that are determined to be cells of a preset type and are taken under a high-power objective lens.

24. The apparatus according to any one of claims 1 to 6, 9 to 12, 20 to 21, wherein: The digital camera is configured to sequentially photograph the identified and located cells of suspected preset type under the high-power objective lens; The processor is configured to simultaneously identify whether the suspected preset type of cells are the preset type of cells during the process of the digital camera photographing the suspected preset type of cells; and instruct the digital camera to stop photographing under the high-power objective lens after determining the target number of preset type of cells.

25. The apparatus of claim 24, wherein: The device further comprises: a third memory configured to store images of cells of a preset type and cells of a non-preset type suspected to be cells of a preset type captured under the high-power objective lens; The output device includes a third display screen configured to display images of cells of a preset type and cells of a non-preset type that are suspected to be cells of a preset type, which are captured under the high-power objective lens.

26. The apparatus according to any one of claims 1 to 6, 9 to 12, 20 to 21, wherein: The preset cell types are white blood cells and / or nucleated red blood cells.

27. A cell identification method, applied to the cell morphology analysis device according to claim 1, the method comprising: Obtaining a target number, where the target number is the number of cells of a preset type that the cell morphology analysis device needs to photograph under a high-power objective lens; Take cell images of blood samples using a low-magnification objective lens; Identifying and locating cells of suspected preset types in the cell image to obtain an identification result, and determining the number of cells of suspected preset types identified and located under a low-power objective lens based on the identification result and the target number; After completing the identification and positioning of a number of cells suspected of a preset type, stopping the photographing under the low-magnification objective lens; Switching the low-power objective lens to the high-power objective lens; photographing the suspected preset type of cells identified and located under the low-magnification objective lens under the high-magnification objective lens; Identifying whether the suspected preset type cells photographed under the high-power objective lens are the preset type cells, and counting the number of the preset type cells photographed under the high-power objective lens to obtain a statistical value; When the statistical value satisfies: target number ≤ statistical value, stop shooting under the high-power objective lens; The cell information of the suspected predetermined type cell identified as the predetermined type cell is output.

28. The method according to claim 27, wherein Determining the number of cells of the suspected preset type identified and located under a low-power objective lens based on the recognition result and the target number includes: In the cell image recognition process, the probability information of each suspected preset type cell being a preset type cell is obtained. ; According to the probability information that each suspected preset type cell is a preset type cell and the target number, and calculating the localization number of the suspected preset type of cells.

29. The method according to claim 28, wherein The method comprises: According to the probability information that each suspected preset type cell is a preset type cell , calculate the estimated number of cells of a preset type ; The localized number of cells of the suspected predetermined type is calculated based on the estimated number of cells of the predetermined type and the target number.

30. The method according to claim 29, wherein The method comprises: Accumulate the probability information of all suspected preset type cells obtained at the current moment as preset type cells, and obtain the estimated number of preset type cells accumulated at the current moment ; When the estimated number is met When the target number is reached, the shooting under the low-power objective lens is stopped, and N at the current moment is the positioning number, wherein λ is a constant, 1≤λ≤1.

25.

31. The method according to claim 28, wherein The method comprises: Obtaining probability information that each suspected preset type cell in the cell image of the first field of view is a preset type cell ; According to the probability information that each suspected preset type cell in the cell image of the first field of view is a preset type cell and the target number, calculating the localization number of the suspected preset type of cells; or The method comprises: Obtaining probability information that each suspected preset type of cell in the cell image of the first field of view is a preset type of cell; Switching the field of view to obtain probability information that each suspected preset type of cell in the cell image of the second field of view is a cell of the preset type; Probability information based on the first visual field and the second visual field identifying and locating all suspected preset type cells as cells of the preset type and the target number, and calculating the localization number of the suspected preset type of cells.

32. The method according to claim 31, wherein The method comprises: The probability information of each suspected preset type cell obtained in the cell image of the first field of view being a preset type cell Accumulate and get the estimated number of cells of the preset type in the first field of view , where N is the number of suspected preset type cells identified and located in the first field of view; Then the number of cells suspected of being of the preset type located = the number of targets .

33. The method according to claim 31, wherein The method comprises: Probability information based on the first visual field and the second visual field identifying and locating all suspected preset type cells as cells of the preset type , get the estimated number of cells of the preset type at the current moment , where N is the number of suspected preset type cells identified and located by the first field of view and the second field of view; then the number of suspected preset type cells located = the number of targets .

34. The method of claim 27, wherein: Determining the number of cells of the suspected preset type identified and located under a low-power objective lens based on the recognition result and the target number includes: Positioning and identifying cells of suspected preset types in the cell image in sequence; Identifying whether the suspected preset type of cells are the preset type of cells, and counting the number of cells identified as the preset type under a low-power objective lens; When the number of cells identified as the preset type under the low-magnification objective lens reaches the target number, shooting under the low-magnification objective lens is stopped, and the number of suspected preset type cells that have been identified and located when the target number is reached is used as the located number of the suspected preset type cells.

35. A cell identification method, applied to the cell morphology analysis device according to claim 1, the method comprising: Obtaining a target number, wherein the target number is the number of cells of a preset type that need to be photographed under a high-power objective lens in the cell morphology analysis device; Acquire cell images of blood samples taken under a low-magnification objective lens; Identifying cells of suspected preset types in the cell image, and sequentially identifying and locating the cells of suspected preset types; Obtain information that all suspected preset type cells at the current moment are identified as preset type cells to determine whether it is necessary to locate and identify the next suspected preset type cell; When it is necessary to locate and identify the next suspected preset type cell, locate and identify the next suspected preset type cell; When there is no need to locate and identify the next suspected preset type of cell, stop photographing under the low-magnification objective lens; Switching the low-power objective lens to the high-power objective lens; Acquiring the suspected preset type cells identified and located under the low-magnification objective lens and photographed under the high-magnification objective lens, identifying whether the suspected preset type cells photographed under the high-magnification objective lens are the preset type cells, and stopping photographing under the high-magnification objective lens when the number of the preset type cells meets the target number; The cell information of the suspected predetermined type cell identified as the predetermined type cell is output.

36. The method according to claim 35, wherein The obtaining of information that all suspected cells of the preset type are identified as cells of the preset type at the current moment to determine whether the next suspected cell of the preset type needs to be located and identified includes: Locating the current suspected preset type cell, obtaining probability information of the current suspected preset type cell being identified as the preset type cell and probability information accumulated before the current suspected preset type cell is identified; The accumulated probability information is updated based on the probability information of the current suspected preset type cell being identified as the preset type cell and the probability information accumulated before the current suspected preset type cell is identified, and whether the next suspected preset type cell needs to be located is determined based on the accumulated probability information.

37. The method according to claim 35, wherein The obtaining of information that all suspected cells of the preset type are identified as cells of the preset type at the current moment to determine whether the next suspected cell of the preset type needs to be located and identified includes: Locating the current suspected preset type cell and identifying whether the current suspected preset type cell is the preset type cell; The number of suspected preset type cells identified as preset type cells is updated to determine whether the next suspected preset type cell needs to be located and identified.

38. A cell morphology analysis device, comprising: a memory configured to store executable instructions; The processor is configured to execute the cell identification method described in any one of claims 27 to 34 and / or the cell identification method described in any one of claims 35 to 37 when running the executable instructions stored in the memory.

39. A storage medium storing executable instructions, configured to cause a processor to execute the executable instructions to implement the cell identification method described in any one of claims 27 to 34 and / or execute the cell identification method described in any one of claims 35 to 37.

Images