Suspected blood parasite detection and AI training methods

Abstract

In the suspected blood parasite detection and AI training method, an infected blood sample is collected; the infected blood sample is infected by parasites; preprocessing the blood sample to obtain a microscopic examination sample; setting magnification times of the microscope, and shooting the microscopic examination sample to obtain a picture; in the pictures, images of suspected parasites are obtained; marking the parasites as suspected parasites, and obtaining a marked image; performing AI training by using the marked image to obtain suspected parasite feature data; the suspected parasite feature data is combined with the corresponding AI algorithm, and the parasite recognition capability is achieved. Preprocessing the blood sample to obtain a microscopic examination sample; setting magnification times of the microscope, and shooting the microscopic examination sample to obtain a microscopic examination sample image; an AI recognition algorithm is matched with the suspected parasite feature data, the microscopic examination sample image is analyzed, and suspected parasite in the image is recognized; the suspected parasite feature data is obtained by training suspected parasite images.

Description

Technical Field

[0001] The present application belongs to the technical field of analysis of formed components based on microscopic magnified images, and in particular relates to a method for assessing the possibility of parasitic infection, a method for detecting suspected blood parasites, and an AI training method and computing processing, storage device or detection device thereof. Background Art

[0002] In the prior art, the diagnosis of parasitic infection usually requires the use of a microscope with sufficient magnification.

[0003] For example, Babesia is a blood parasite. The main mode of infection and transmission is the bite of hard ticks. After the tick bites, the Babesia sporozoites are released into the host's blood. The sporozoites penetrate the red blood cells and divide to form schizonts and merozoites. The red blood cells rupture and the parasites overflow. The parasites continue to invade and destroy new red blood cells to repeat the above development process. At the same time, it can also be transmitted through the blood and vertically through the placenta. According to morphology, it can be divided into two categories: small Babesia and large Babesia.

[0004] Babesia microti: mainly Babesia gibbsii, 1-3 μm in size, with a typical feature of a single ring structure such as a ring, and also dots and small rods. Multiple parasites can be infected in one red blood cell. Figure 1 and Figure 2 The figure shows the infection of Babesia microti in red blood cells. Figure 1 and Figure 2 This is a magnified microscopic image under an oil microscope. The small dots in the red blood cells in the image are Babesia microti.

[0005] Babesia macrocarpa: mainly Babesia canis, Babesia voroni, Babesia rosenbergii, with a size of 3-7 μm, and a typical pear-seed shape, such as a melon seed, with a pointed end and a blunt end. There are also various shapes such as ring-shaped and amoeba-like. Multiple parasites can be infected in one red blood cell. Figure 3 The picture shows the infection of Babesia major in red blood cells. Figure 3 This is a magnified microscopic image under an oil microscope. The pear-shaped dots in the red blood cells and outside the cells are Babesia major.

[0006] In the formed component detection device, the higher the magnification of the microscope, the higher the cost. Diagnosis of Babesia usually requires a microscope with a magnification of more than 60 times, such as a 100-times oil immersion lens. How to output the detection results of Babesia with a low magnification microscope, such as a microscope with a magnification of about 40 times, is a problem to be solved. Summary of the invention

[0007] In the present application, the inventor proposed a technical solution that uses a relatively low-magnification microscope combined with suspected parasite characteristic data to output a test result that is more suspected of Babesia infection.

[0008] The AI ​​training method for suspected blood parasite detection can obtain suspected parasite characteristic data, and with the help of suspected parasite characteristic data and corresponding detection methods, it can output suspected blood parasite detection results, provide more in-depth information for clinicians, and provide a more accurate guide for whether to conduct a larger magnification test. It can make full use of existing hardware conditions to obtain more in-depth and extensive information for clinical reference.

[0009] The solution of the present application to solve the above-mentioned technical problems is an AI training method for detecting suspected blood parasites, comprising: collecting infected blood samples; the infected blood samples are infected with parasites; pre-treating the blood samples to obtain microscopic samples; setting the magnification of the microscope, photographing the microscopic samples, and obtaining pictures; marking the images of suspected parasites in the pictures as suspected parasites to obtain marked images; performing AI training with the marked images to obtain suspected parasite feature data; combining the suspected parasite feature data with the corresponding AI algorithm to have the ability to identify parasites.

[0010] It may be that the parasite is Babesia; Babesia major; Babesia minor; Plasmodium.

[0011] It may be that the parasite is Babesia major, and the magnification is greater than 10 times and less than 60 times; the parasite is Babesia microphylla, and the magnification is greater than 10 times and less than 60 times; the parasite is Plasmodium, and the magnification is greater than 10 times and less than 60 times.

[0012] It can be that if the parasite is Babesia major, the magnification is set to be greater than 80 times, an image is taken, and the suspected parasite is confirmed to be Babesia major; if the parasite is Babesia microphylla, the magnification is set to be greater than 90 times, an image is taken, and the suspected parasite is confirmed to be Babesia microphylla; if the parasite is Plasmodium, the magnification is set to be greater than 100 times, an image is taken, and the suspected parasite is confirmed to be Babesia microphylla.

[0013] The solution of the present application to solve the above-mentioned technical problems can also be a suspected blood parasite detection method, which comprises pre-treating a blood sample to obtain a microscopic sample; setting the magnification of the microscope, photographing the microscopic sample, and obtaining a microscopic sample image; using an AI recognition algorithm in combination with suspected parasite feature data to analyze the microscopic sample image and identify suspected parasites in the image; the suspected parasite feature data is obtained by training with suspected parasite images.

[0014] It may be that the parasite is Babesia; Babesia major; Babesia minor; Plasmodium.

[0015] It may further include obtaining the number of suspected parasites in the microscopic sample image, obtaining the volume corresponding to the microscopic sample image, and calculating the content of the suspected parasites per unit volume.

[0016] It may further include obtaining the number of suspected parasites in the microscopic sample image, obtaining the number of red blood cells corresponding to the microscopic sample image, and calculating the ratio of the number of suspected parasites to the number of red blood cells.

[0017] The solution of the present application to solve the above technical problem can also be a computing and processing device or a detection device, which is used to run all or part of the above method; the memory of the computing and processing device includes the above-mentioned suspected blood parasite feature data set.

[0018] The solution of the present application to solve the above technical problem can also be a data storage device, which stores all or part of the program code used for the above method; and stores the above-mentioned suspected blood parasite feature data set.

[0019] The technical effects of the above technical solution include: using low-magnification microscopic equipment to train the ability to detect suspected infections, which can significantly reduce the cost of screening equipment, reduce the cost of suspected screening, and enable widespread screening and testing, enabling commercial testing to be carried out.

[0020] The technical effects of the above technical solution include: using samples infected with parasites to form effective samples, the training efficiency is higher. Using infected samples to train suspected parasite feature data sets can improve the accuracy of identifying suspected parasite infections.

[0021] The technical effects of the above technical solution include: establishing multiple suspected parasite feature data sets, and having the ability to screen multiple parasites. The method in this application can be used to screen suspected parasites of small size.

[0022] The technical effects of the above technical solution include: different magnification ranges are suitable for parasites of different sizes.

[0023] The technical effects of the above technical solution include: a suspected blood parasite detection method that can complete the screening of small-sized parasites with the help of a low-magnification microscope, and combined with the suspected content, it can partially confirm the infection.

[0024] The technical effects of the above technical solution include: the content of suspected parasites in unit volume can be used to quantitatively evaluate the suspected parasites, the output results are more credible, and it is easier to distinguish the severity of the infection.

[0025] The technical effects of the above technical solution include: a method for evaluating the possibility of blood parasite infection, which can output the ratio of the number of suspected parasites to red blood cells, does not affect the results for samples of different concentrations, and conducts a quantitative evaluation of suspected parasites, which has a higher reference value.

[0026] The technical effects of the above technical solution include: a method for evaluating the possibility of blood parasite infection, which can calculate the content of suspected parasites in a unit volume and accurately quantify the results, making the results more reliable. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 is a schematic diagram of erythrocytes infected with Babesia microti at low magnification; Figure 2 This is a schematic diagram of erythrocytes infected with Babesia microti at high magnification; Figure 3 This is a schematic diagram of red blood cells infected with Babesia major; Figure 4 This is a schematic diagram of an AI training method for suspected blood parasite detection; Figure 5 This is a schematic diagram of an AI training method for suspected blood parasite detection; Figure 6 It is a black and white schematic diagram of red blood cells infected with Babesia major; Figure 7 It is a schematic diagram of the training red blood cells infected with Babesia microti; Figure 8 It is a schematic diagram of a method for detecting suspected blood parasites; Fig. 9 It is a schematic diagram of a method for detecting suspected blood parasites. DETAILED DESCRIPTION

[0028] The content of this application is further described in detail below in conjunction with the accompanying drawings. It should be noted that the following is a description of the preferred embodiments of the present invention and does not constitute any limitation to the present invention. The description of the preferred embodiments of the present invention is only an illustration of the general principles of the present invention. The numbers such as "first", "second" and "A" and "B" involved in the present invention are only for the convenience of explanation and do not represent the order relationship in time or space. The combination of letters and numbers "TA", "TB" and "H" involved in the present invention are only for the convenience of explanation, and the specific meaning is determined by the specific words referred to.

[0029] like Figure 4 A suspected blood parasite detection AI training method includes: collecting infected blood samples; the infected blood samples are infected by parasites; pre-treating the blood samples to obtain microscopic samples; setting the microscope magnification, photographing the microscopic samples, and obtaining pictures; marking the images of suspected parasites in the pictures as suspected parasites to obtain marked images; performing AI training with the marked images to obtain suspected parasite feature data; combining the suspected parasite feature data with the corresponding AI algorithm to have the ability to identify parasites.

[0030] The parasite is Babesia; the parasite is Babesia major; the parasite is Babesia minor; the parasite is Plasmodium.

[0031] The parasite is Babesia major, and the magnification is greater than 10 times and less than 60 times.

[0032] The parasite is Babesia microti, and the magnification is greater than 10 times and less than 60 times. Figure 7 , is a microscopic enlarged image of Babesia microphylla at a magnification of 40 times. Babesia microphylla: colored in light pink, light blue, and light purple; size: 1 to 3.3 μm; shape is dot-like or ring-like; intracellular: dot-like, at any position; inside / outside the edge of the red blood cell, similar to Heinz bodies.

[0033] The target objects in the present application may also have more types, as long as they are targets that can be observed by a 100x oil objective but cannot be observed by a 40x objective, such as targets related to leukocyte virus infection.

[0034] The parasite is Plasmodium, and the magnification is greater than 10 times and less than 60 times.

[0035] The parasite is Babesia major, and the magnification is set to be greater than 80 times, and an image is taken to confirm that the suspected parasite is Babesia major.

[0036] The parasite is Babesia microphylla, and the magnification is set to be greater than 90 times, and an image is taken to confirm that the suspected parasite is Babesia microphylla.

[0037] The parasite is Plasmodium, and the magnification is set to be greater than 100 times, and an image is taken to confirm that the suspected parasite is Babesia microti.

[0038] like Figure 8 , a suspected blood parasite detection method, pre-processing a blood sample to obtain a microscopic sample; setting the microscope magnification, photographing the microscopic sample, and obtaining a microscopic sample image; using an AI recognition algorithm in conjunction with suspected parasite feature data, analyzing the microscopic sample image, and identifying suspected parasites in the image; the suspected parasite feature data is obtained by training the suspected parasite image. It also includes obtaining the number of suspected parasites in the microscopic sample image, obtaining the volume corresponding to the microscopic sample image, and calculating the content of suspected parasites per unit volume.

[0039] like Fig. 9 , a suspected blood parasite detection method also includes obtaining the number of suspected parasites in the microscopic sample image, obtaining the number of red blood cells corresponding to the microscopic sample image, and calculating the ratio of the number of suspected parasites to the number of red blood cells.

[0040] like Fig. 9, a suspected blood parasite detection method is used to evaluate the possibility of blood infection with parasites, obtain the number of suspected parasites in a microscopic sample image, obtain the number of red blood cells corresponding to the microscopic sample image, and calculate the ratio of the number of suspected parasites to red blood cells = the number of suspected parasites / the number of red blood cells.

[0041] like Figure 8 A method for detecting suspected blood parasites is used to evaluate the possibility of blood infection with parasites, obtain the number of suspected parasites in the microscopic sample image, obtain the volume corresponding to the microscopic sample image, and calculate the content of suspected parasites in unit volume = (number of suspected parasites / microscopic sample volume) × dilution factor of the microscopic sample.

[0042] Essentially, the above process, including the number of suspected parasites identified by AI, the number of cells and their confidence calculation, is used to comprehensively obtain the probability of parasites. When the probability reaches a certain threshold, it is judged as parasite infection. The threshold can be set according to different target scenarios.

[0043] A computing and processing device or a detection device, including all or part of the devices for running the above method; the memory of the computing and processing device includes the above suspected blood parasite characteristic data set.

[0044] A data storage device stores all or part of the program code for executing the above method; and stores the above suspected blood parasite feature data set.

[0045] Although the present invention is illustrated and described according to the preferred embodiment and several alternatives, the invention is not limited by the specific description in this specification. Other additional replacement or equivalent components can also be used to practice the present invention.

Claims

1. A suspected blood parasite detection AI training method, characterized in that: include: Collect infected blood samples; The infected blood sample is infected with a parasite; Pre-process blood samples to obtain samples for microscopic examination; Set the magnification of the microscope, take pictures of the samples under the microscope, and obtain images; For images suspected of being parasites in the picture, mark them as suspected parasites to obtain a marked image; Use labeled images to train AI and obtain data on suspected parasite characteristics; The suspected parasite feature data is combined with the corresponding AI algorithm to have the ability to identify parasites.

2. The method according to claim 1, characterized in that: The parasites include any one or more of the following parasites: TB10: Babesia, the suspected parasite characteristic data is suspected Babesia characteristic data; TB20: Babesia major, the suspected parasite characteristic data is suspected Babesia major characteristic data; TB30: Babesia microtyphi, the suspected parasite characteristic data is suspected Babesia microtyphi; TB40: Plasmodium, the suspected parasite characteristic data is suspected Plasmodium characteristic data.

3. The method according to claim 2, characterized in that The parasite is Babesia major, and the magnification is greater than 10 times and less than 60 times; The parasite is Babesia microti, and the magnification is greater than 10 times and less than 60 times; The parasite is Plasmodium, and the magnification is greater than 10 times and less than 60 times.

4. The method according to claim 2, characterized in that: include: The parasite is Babesia major, and the magnification is set to be greater than 80 times, and an image is taken to confirm that the suspected parasite is Babesia major; The parasite is Babesia microti, and the magnification is set to be greater than 90 times, and an image is taken to confirm that the suspected parasite is Babesia microti; The parasite is Plasmodium, and the magnification is set to be greater than 100 times, and an image is taken to confirm that the suspected parasite is Babesia microti.

5. A method for detecting suspected blood parasites, characterized in that: Pre-process blood samples to obtain samples for microscopic examination; Setting the magnification of the microscope, photographing the microscopic sample, and obtaining an image of the microscopic sample; The microscopic sample image is analyzed by using an AI recognition algorithm in combination with the suspected parasite characteristic data to identify the suspected parasites in the image; the suspected parasite characteristic data is obtained by training on the suspected parasite images.

6. The method according to claim 5, characterized in that The parasites include any one or more of the following parasites: TB10: Babesia; TB20: Babesia major; TB30: Babesia microti; TB40: Plasmodium.

7. The method according to claim 5, characterized in that The method also includes obtaining the number of suspected parasites in the microscopic sample image, obtaining the volume corresponding to the microscopic sample image, and calculating the content of the suspected parasites per unit volume.

8. The method according to claim 5, characterized in that The method also includes obtaining the number of suspected parasites in the microscopic sample image, obtaining the number of red blood cells corresponding to the microscopic sample image, and calculating the ratio of the number of suspected parasites to the number of red blood cells.

9. A computing device or a detection device, characterized in that: Include any of the following technical features: TEA1: used to perform all or part of the method according to any one of claims 1 to 8; TEA2: The memory of the computing and processing device includes the suspected blood parasite feature data set described in any one of claims 1 to 8.

10. A data storage device, characterized in that: Include any of the following technical features: TEB1: storing program codes for executing all or part of the method according to any one of claims 1 to 8; TEB2: stores a suspected blood parasite feature dataset as described in any one of claims 1 to 8.

Images