Device and method for detecting serum quality
By taking blood sample images by rotating mechanisms and cameras, combined with neural network models, the problems of low efficiency and high cost of serum quality detection in the prior art are solved, and efficient and accurate quantitative detection of serum quality is achieved.
Patent Information
- Application Number
- CN202311837242.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-12-28
- Publication Date
- 2025-07-01
AI Technical Summary
The prior art is difficult to obtain quantitative test results of serum quality efficiently and at low cost, and relying on naked eye observation and subsequent reagent detection leads to high time cost and low efficiency.
The rotating mechanism is used to drive the container to rotate, multiple sample images are taken through the camera, and the processor is used to obtain the chylo index, hemolysis index and jaundice index based on image analysis, and combine it with the neural network model to improve detection accuracy.
It achieves low-cost and efficient quantitative detection results for serum quality, improves the accuracy and efficiency of detection, and reduces dependence on additional assay processes.
Smart Images
Figure CN120232883A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of in vitro diagnostics, and particularly to a device for detecting the quality of serum and a method for detecting the quality of serum. Background Art
[0002] The quality of serum in blood samples (such as hemolysis, lipemia, and jaundice, etc.) is an important part of pre-laboratory quality control. Abnormal serum quality will lead to doubts about the accuracy of the test results of blood samples. For example, abnormal quality of blood samples will interfere with the results of blood cell parameters obtained by blood cell assembly line testing (for example, chylous samples will cause false high hemoglobin measurement results).
[0003] Generally, laboratory technicians do not visually observe whether there is any abnormality in the serum quality of each blood sample before measurement. Usually, after the blood sample is tested by the assembly line to obtain the measurement result, when the clinician finds that the measurement result is abnormal, the clinician will confirm whether the blood sample is normal and mainly rely on the visual observation of the blood sample to determine whether there is any abnormality in the blood sample.
[0004] With the development of artificial intelligence, some methods for automatically detecting whether a blood sample is normal have emerged. These methods can qualitatively give whether the blood sample is abnormal, or give a conclusion that the blood sample belongs to one or more of chylous, hemolytic, or jaundiced abnormalities. However, different degrees of chylous, hemolytic, or jaundiced abnormalities have different effects on the results of different test items. Only obtaining the qualitative result of abnormal blood sample quality is not sufficient to meet the clinical requirements for the accuracy of test items.
[0005] In addition, although after obtaining the qualitative result of abnormal blood sample, relevant reagents can be further used to detect the serum index of the blood sample to obtain the quantitative result of each serum index, this detection method has a high time cost and low detection efficiency. Summary of the Invention
[0006] Therefore, the task of the present application is to provide a device for detecting the quality of serum and a method for detecting the quality of serum, which can obtain the quantitative detection result of serum quality at low cost and high efficiency.
[0007] To achieve the above task of the present application, the first aspect of the present application first proposes a device for detecting the quality of serum, including:
[0008] A sample base for placing a container containing a blood sample;
[0009] A rotating mechanism for rotating the sample base to drive the container placed in the sample base to rotate;
[0010] A camera, configured to capture a blood sample in the container during rotation of the container to obtain a plurality of sample images of the blood sample;
[0011] A processor, configured to:
[0012] Obtain one or more of a chylomicron index, a hemolysis index, and a jaundice index of the blood sample based on the plurality of sample images.
[0013] A second aspect of the present application provides a method for detecting the quality of serum, including:
[0014] Rotating a container containing a blood sample;
[0015] Capturing a blood sample in the container during rotation of the container to obtain a plurality of sample images of the blood sample;
[0016] Obtain one or more of a chylomicron index, a hemolysis index, and a jaundice index of the blood sample based on the plurality of sample images.
[0017] The method for detecting the quality of serum according to the second aspect of the present application is particularly applicable to the device for detecting the quality of serum according to the first aspect of the present application.
[0018] In the technical solutions proposed in various aspects of the present application, only a plurality of sample images of the blood sample during the rotation of the container need to be obtained, and one or more of the chylomicron index, the hemolysis index, and the jaundice index of the blood sample can be obtained without performing an additional determination process for the serum index, so that a quantitative detection result of the serum quality can be obtained at low cost and high efficiency. Description of the Drawings
[0019] Figure 1 FIG. is a schematic structural diagram of a device for detecting the quality of serum according to some embodiments of the present application.
[0020] Figure 2 FIG. is a schematic diagram showing the change of the color and turbidity of the serum region with the degree of chylomicron, hemolysis, or jaundice of the blood sample.
[0021] Figure 3 FIG. is a schematic diagram of a scheme for screening serum images according to some embodiments of the present application.
[0022] Figure 4 FIG. is a schematic diagram of a scheme for screening serum images according to some other embodiments of the present application.
[0023] Figures 5 to 7 FIG. is a schematic diagram of a scheme for obtaining a serum index using a second neural network model according to some different embodiments of the present application.
[0024] Figure 8Schematic diagram of the correlation between the hemolysis index obtained by the method proposed in the embodiments of the present application and the hemolysis index measured by an instrument.
[0025] Figure 9 Schematic diagram of the correlation between the chylomicron index obtained by the method proposed in the embodiments of the present application and the chylomicron index measured by an instrument.
[0026] Figure 10 Schematic diagram of the correlation between the jaundice index obtained by the method proposed in the embodiments of the present application and the jaundice index measured by an instrument.
[0027] Figure 11 Flow schematic diagram of the method for detecting the quality of serum according to some embodiments of the present application.
[0028] Figure 12 Flow schematic diagram of the method for detecting the quality of serum according to other embodiments of the present application. Detailed implementation manners
[0029] The embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the scope of protection of the present application.
[0030] It should be noted that the terms "first\second\third" involved in the embodiments of the present application are only used to distinguish similar objects, and do not represent a specific order for the objects. It can be understood that "first\second\third" can be interchanged with a specific order or sequence when permitted.
[0031] Figure 1 Structural schematic diagram of a device for detecting the quality of serum according to some embodiments of the present application. The device 100 for detecting the quality of serum includes a sample base 110, a rotating mechanism 120, a camera 130, and a processor 140.
[0032] The sample base 110 is used to place a container 10 containing a blood sample.
[0033] The rotating mechanism 120 is used to rotate the sample base 110 to drive the container placed in the sample base 110 to rotate, for example, rotate at least 180°, preferably rotate 360°.
[0034] The camera 130 is used to take pictures of the blood sample in the container during the rotation of the container to obtain a plurality of sample images of the blood sample.
[0035] The processor 140 is configured to obtain one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the plurality of sample images.
[0036] In the above device 100 for detecting the quality of serum, a rotating mechanism is used to drive the container placed in the sample base to rotate. During the rotation of the container, a camera is used to capture the blood sample in the container to obtain a plurality of sample images of the blood sample, and then one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample are obtained based on the plurality of sample images. In this way, without performing an additional measurement process for the serum index, one or more of the chylomicron index, hemolysis index, and jaundice index (i.e., the serum index) of the blood sample can be obtained.
[0037] The following further describes the device 100 for detecting the quality of serum provided by the present application in conjunction with some embodiments.
[0038] In some embodiments, the rotating mechanism 120 can be used to drive the container placed in the sample base 110 to rotate at a preset speed, and the camera 130 can be used to continuously capture the blood sample in the container at a preset frame rate during the rotation of the container to obtain the plurality of sample images.
[0039] In some embodiments, the processor 140 can perform visual processing on the intermediate processing result or the final processing result, and then display it through a display device ( Figure 1 not shown). For example, the display device can include a user interface, and the processor 140 can output and display one or more of the chylomicron index, hemolysis index, and jaundice index of the obtained blood sample on the user interface of the display device.
[0040] In some embodiments, the processor 140 includes, but is not limited to, a central processing unit (CPU), a microcontroller unit (MCU), a field-programmable gate array (FPGA), a digital signal processor (DSP), etc., which are devices for interpreting computer instructions and processing data in computer software. For example, the processor 140 is used to execute various computer application programs in a computer-readable storage medium, so that the device 100 for detecting the quality of serum performs corresponding detection processes and analyzes the plurality of sample images obtained through the camera 130 in real time. In the embodiments of the present application, the processor 140 is configured to implement the method steps described in further detail later.
[0041] In some embodiments, the device 100 for detecting the quality of serum may further include a light source for irradiating the container during the rotation of the container ( Figure 1 not shown). For example, a white light source may be used to irradiate the blood sample in the container to avoid the influence of the color of other colored light sources on the detection result of the serum index.
[0042] In some embodiments, as Figure 1 shown, the device 100 for detecting the quality of serum may further include a sample storage mechanism 150, a centrifugation mechanism 160, and a transportation mechanism 170.
[0043] The sample storage mechanism 150 includes a centrifugation area 151 for placing the container loaded with the centrifuged blood sample and a non - centrifugation area 152 for placing the container loaded with the non - centrifuged blood sample.
[0044] The centrifugation mechanism 160 is used to perform a centrifugation operation on the blood sample.
[0045] The transportation mechanism 170 is used to directly transport the container in the centrifugation area to the sample base 110 to detect the quality of the serum in the blood sample in the container, and to transport the container in the non - centrifugation area to the centrifugation mechanism 160 to centrifuge the blood sample in the container, and after centrifugation, transport the container to the sample base 110 to detect the quality of the serum in the blood sample in the container.
[0046] The following further describes the specific manner in which the processor 140 obtains one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample in conjunction with some embodiments.
[0047] In some embodiments, the processor 140 may be configured to, when obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the plurality of sample images:
[0048] Obtain one or more serum images including serum regions based on the plurality of sample images; and
[0049] Obtain one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the serum images.
[0050] For example, the chylomicron index, hemolysis index, and jaundice index of the blood sample can be obtained based on the serum images.
[0051] In some embodiments, the processor 140 may be configured to, when obtaining one or more serum images including serum regions based on the plurality of sample images:
[0052] Image feature extraction is performed on each of the multiple sample images to obtain a first image feature of each sample image; and
[0053] Based on the first image feature of each sample image, the serum image is screened out from the multiple sample images.
[0054] As Figure 2 shown, the larger the chylomicron index (i.e., the larger the value of L), the cloudier the serum region; the larger the hemolysis index (i.e., the larger the value of H) or the larger the jaundice index (i.e., the larger the value of l), the darker the color of the serum region. That is, the color and turbidity of the serum region vary with the degree of chylomicron, hemolysis or jaundice in the blood sample.
[0055] Thus, in some embodiments, the first image feature of each sample image may include one or more features of the color feature and the brightness feature of the sample image. For example, the extracted color feature of each sample image may include information such as the hue and saturation of the serum region in the sample image, and the brightness feature may include information such as the light and dark degree of the sample image.
[0056] As Figure 3 shown, 4 sample images, namely images A, B, C, and D, taken from different angles are obtained by the camera 130. Image feature extraction is performed on each of the 4 sample images to obtain the color feature and the brightness feature of each sample image. Based on the color feature and the brightness feature of each sample image, the serum image X is screened out from the 4 sample images.
[0057] In this way, since the multiple sample images are obtained during the rotation of the container, and some images may have problems such as low brightness and color distortion due to shooting angle or light, etc., by performing image feature extraction on each sample image and screening out the serum image from the multiple sample images based on the extracted first image feature, the accuracy of the serum image used for detecting the serum quality can be effectively improved, thereby improving the accuracy of the serum quality detection.
[0058] In some other embodiments, the processor 140 may be configured to, when obtaining one or more serum images including serum regions based on the multiple sample images: input the multiple sample images into a first neural network model to obtain the output of the first neural network model as the serum image.
[0059] In some embodiments, the first neural network model may include one of a classification neural network model, a segmentation neural network model, and a detection neural network model. That is, the multiple sample images may be input into any one of the classification neural network model, the segmentation neural network model, and the detection neural network model to obtain the output of any one of the models as the serum image.
[0060] As Figure 4 shown, four sample images (i.e., images A, B, C, and D) taken from different angles are obtained by the camera 130, and these four sample images are input into the first neural network model to obtain the output of the first neural network model as the serum image X.
[0061] In some embodiments, the processor 140 may be configured to, when obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the serum image:
[0062] extract image features from the serum image to obtain second image features of the serum image; and
[0063] obtain one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the second image features of the serum image.
[0064] In some embodiments, the second image features may include one or more of color features, grayscale histograms, and image gradients. For example, the color features of the serum image may include information such as the hue and saturation of the serum region in the serum image.
[0065] As some implementation manners, the processor 140 may be configured to, when obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the second image features of the serum image: input the second image features of the serum image into a machine learning model to predict one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample. For example, input the extracted second image features of the serum image into a machine learning model (such as any one of a logistic regression model, a nearest neighbor model, and a multi-layer perceptron model) to obtain the prediction result output by the machine learning model as one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.
[0066] In some other embodiments, the processor 140 may be configured to, when obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the serum image: input one or more serum images including serum regions into a second neural network model to obtain the output of the second neural network model as one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.
[0067] For example, as Figure 5 shown, input three serum images selected from the multiple sample images into the second neural network model to obtain the output of the second neural network model as one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.
[0068] In a specific example, as Figure 6 shown, the second neural network model may at least include an image input layer, one or more convolutional layers, a pooling layer, a fully connected layer, and an activation function ReLu. Input one or more serum images including serum regions into the second neural network model through the image input layer to obtain the output of the second neural network model as one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.
[0069] In still some other embodiments, the processor 140 may be configured to, when obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the serum image: input the serum region image cropped from the serum image into the second neural network model to obtain the output of the second neural network model as one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.
[0070] Since the support part of the sample base may block the serum region, thereby interfering with the detection of the serum quality, therefore, as some implementation manners, the support part region of the sample base in the serum image may be removed (for example, cropping the support part region in the serum image according to the image height corresponding to the support part, or separating the support part region in the serum image by segmentation and then removing it) to obtain a serum region image only including the serum region, and input the serum region image into the second neural network model to obtain the output of the second neural network model as one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.
[0071] For example, as Figure 7As shown, three serum images selected from the multiple sample images are respectively cropped to obtain three serum region images X1, X2, and X3 (i.e., images containing only the serum region). These three serum region images are input into the second neural network model, and the output of the second neural network model is used as one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.
[0072] In a specific example, the second neural network model can adopt Figure 6 the structure shown. The serum region images cropped from one or more serum images are input into the second neural network model through the image input layer, and the output of the second neural network model is used as one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.
[0073] In this way, by cropping the serum region images from the serum images and obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the cropped serum region images, the interference of other factors in non-serum regions on the detection of serum quality can be effectively reduced, thereby further improving the accuracy of serum quality detection.
[0074] Figures 8 to 10 shows the correlation between each serum index obtained by using the second neural network model and each serum index measured by Mindray BS-2800M instrument.
[0075] As Figure 8 shown, under the above method provided in the embodiment of the present application, the correlation between the hemolysis index (i.e., the predicted value) of the blood sample obtained by the second neural network model and the hemolysis index (i.e., the instrument value) of this blood sample measured by Mindray BS-2800M instrument reaches 0.9578, indicating that the obtained hemolysis index of the blood sample has good consistency with the hemolysis index measured by the instrument, which shows that the accuracy of the hemolysis index obtained by using the above method provided in the embodiment of the present application is relatively high.
[0076] As Figure 9 shown, under the above method provided in the embodiment of the present application, the correlation between the chylomicron index of the blood sample obtained by the second neural network model and the chylomicron index of this blood sample measured by Mindray BS-2800M instrument reaches 0.9366, indicating that the obtained chylomicron index of the blood sample has good consistency with the chylomicron index measured by the instrument, which shows that the accuracy of the chylomicron index obtained by using the above method provided in the embodiment of the present application is relatively high.
[0077] As Figure 10As shown, in the above manner provided by the embodiments of the present application, the correlation between the jaundice index of the blood sample obtained through the second neural network model and the jaundice index of the blood sample measured by the Mindray BS-2800M instrument reaches 0.8412, indicating that the jaundice index of the obtained blood sample and the jaundice index measured by the instrument have good consistency, which shows that the accuracy of the jaundice index obtained in the above manner provided by the embodiments of the present application is relatively high.
[0078] It can be seen that in the above manner provided by the embodiments of the present application, the accuracy of each serum index can be relatively high, so as to achieve accurate serum quality detection.
[0079] In some embodiments, the processor 140 may be configured to, when obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the plurality of sample images:
[0080] Perform normalization processing on each of the plurality of sample images to obtain a plurality of normalized images;
[0081] Based on the plurality of normalized images, obtain one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.
[0082] It should be noted that normalization processing refers to uniformly transforming variables in different dimensions in the image to the same dimension to eliminate the dimensional differences between variables. For example, the pixel value ranges of each of the plurality of sample images may be mapped to the same preset range.
[0083] In some embodiments, the processor 140 may be configured to, when obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the plurality of normalized images: screen out one or more serum images containing serum regions from the plurality of normalized images, and obtain one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the serum images.
[0084] Here, it can be similarly implemented with reference to the relevant methods of screening serum images and obtaining serum indexes based on the serum images in the foregoing related embodiments. For specific descriptions, reference can be made to the descriptions in the foregoing related embodiments, and details are not described herein again.
[0085] In some embodiments, the processor 140 may be further configured to: output the obtained indexes. For example, if the obtained indexes include multiple indexes among the chylomicron index, hemolysis index, and jaundice index of the blood sample, the processor 140 may output these multiple indexes to the display device for the doctor's clinical reference.
[0086] In some embodiments, the processor 140 may be further configured to: when the acquired index exceeds a preset threshold, output an alarm prompt indicating that the blood sample is abnormal. For example, if the acquired indexes include multiple indexes among the chylomicron index, hemolysis index, and jaundice index of the blood sample, the processor 140 may output the alarm prompt indicating that the blood sample is abnormal when any one of the indexes exceeds the corresponding preset threshold, or the processor 140 may output the alarm prompt indicating that the blood sample is abnormal when all of these multiple indexes respectively exceed the corresponding preset thresholds.
[0087] The present application also provides a method for detecting the quality of serum, as Figure 11 shown, the method 200 for detecting the quality of serum may include the following steps:
[0088] S210, rotating the container containing the blood sample;
[0089] S220, photographing the blood sample in the container during the rotation of the container to obtain a plurality of sample images of the blood sample;
[0090] S230, obtaining one or more indexes among the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the plurality of sample images.
[0091] In some embodiments, as Figure 12 shown, step 230 may include the following steps:
[0092] S231: obtaining one or more serum images including serum regions based on the plurality of sample images; and
[0093] S232: obtaining one or more indexes among the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the serum images.
[0094] In some embodiments, in step S231, image feature extraction may be performed on each of the plurality of sample images to obtain a first image feature of each sample image, and based on the first image feature of each sample image, the serum images may be screened out from the plurality of sample images.
[0095] In some alternative or additional embodiments, in step S231, the plurality of sample images may be input into a first neural network model to obtain the output of the first neural network model as the serum images.
[0096] In some embodiments, in step S232, image feature extraction may be performed on the serum image to obtain second image features of the serum image, and based on the second image features of the serum image, one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample may be obtained.
[0097] In some embodiments, the second image features include one or more of color features, grayscale histograms, and image gradients.
[0098] In some alternative or additional embodiments, in step S232, the serum image or a serum region image cropped from the serum image may be input into a second neural network model to obtain the output of the second neural network model as one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.
[0099] In some embodiments, the serum region image may be cropped from the serum image in the following manner:
[0100] The support portion region of the sample base in the serum image is removed to obtain the serum region image.
[0101] In some embodiments, in step 230, each of the multiple sample images may be normalized to obtain multiple normalized images, and based on the multiple normalized images, one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample may be obtained.
[0102] In some embodiments, the method 200 for detecting serum quality may further include one or more of the following steps:
[0103] Output the obtained index; and
[0104] When the obtained index exceeds a preset threshold, output an alarm prompt indicating that the blood sample is abnormal.
[0105] For more embodiments and advantages of the method 200 for detecting serum quality proposed in the embodiments of the present application, reference may be made to the above description of the device 100 for detecting serum quality, which will not be elaborated here.
[0106] The features or combinations of features mentioned in the above specification, drawings, and claims, as long as they are meaningful within the scope of the present application and do not conflict with each other, may be arbitrarily combined with each other or used alone. The advantages and features described for the device for detecting serum quality provided in the present application are applicable to the method for detecting serum quality provided in the present application in a corresponding manner, and vice versa.
[0107] The above are only the preferred embodiments of the present application, and do not limit the patent scope of the present application accordingly. Any equivalent transformation made under the inventive concept of the present application by using the content of the specification and drawings of the present application, or any direct / indirect application in other related technical fields is included in the patent protection scope of the present application.
Claims
1. A device for detecting the quality of serum, comprising: A sample base for placing a container containing a blood sample; A rotating mechanism for rotating the sample base to drive the container placed in the sample base to rotate; A camera for photographing the blood sample in the container during the rotation of the container to obtain a plurality of sample images of the blood sample; A processor configured to: Obtain one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the plurality of sample images.
2. The device according to claim 1, characterized in that The device further includes a sample storage mechanism, a centrifugation mechanism, and a transportation mechanism. Among them, the sample storage mechanism includes a centrifugation area for placing a container loaded with a centrifuged blood sample and a non-centrifugation area for placing a container loaded with an uncentrifuged blood sample. The centrifugation mechanism is used to perform a centrifugation operation on the blood sample. The transportation mechanism is used to directly transport the container in the centrifugation area to the sample base to detect the quality of the serum in the container, and the transportation mechanism is also used to transport the container in the non-centrifugation area to the centrifugation mechanism to centrifuge the blood sample in the container, and after centrifugation, transport the container to the sample base to detect the quality of the serum in the container.
3. The device according to claim 1 or 2, characterized in that, The processor obtains one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the plurality of sample images, including: the processor Obtains one or more serum images containing serum regions based on the plurality of sample images; Obtains one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the serum images.
4. The device according to claim 3, characterized in that, The processor obtains one or more serum images containing serum regions based on the plurality of sample images, including: the processor Performs image feature extraction on each of the plurality of sample images to obtain a first image feature of each sample image; Based on the first image feature of each sample image, filters out the serum images from the plurality of sample images.
5. The device according to claim 3, characterized in that, The processor obtains one or more serum images containing serum regions based on the plurality of sample images, including: the processor Inputs the plurality of sample images into a first neural network model to obtain the output of the first neural network model as the serum image.
6. The device according to any one of claims 3 to 5, characterized in that The processor obtains one or more of the chylomicron index, hemolysis index, or jaundice index of the blood sample based on the serum images, including: the processor Performs image feature extraction on the serum image to obtain a second image feature of the serum image; Based on the second image feature of the serum image, obtains one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.
7. The device according to claim 6, characterized in that, The second image feature includes one or more features among color features, grayscale histograms, and image gradients.
8. The device according to any one of claims 3-5, characterized in that, The processor obtains one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the serum images, including: the processor Input the serum image or the serum region image cropped from the serum image into a second neural network model to obtain the output of the second neural network model as one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.
9. The device according to claim 8, characterized in that, The processor is further configured to crop the serum region image from the serum image in the following manner: Remove the support portion region of the sample base in the serum image to obtain the serum region image.
10. The device according to any one of claims 1-9, characterized in that, The processor is further configured to: Output the obtained index; and / or When the obtained index exceeds a preset threshold, output an alarm prompt indicating that the blood sample is abnormal.
11. The device according to any one of claims 1-10, characterized in that, The device further includes a light source for irradiating the container during the rotation of the container.
12. The device according to any one of claims 1-11, characterized in that, The processor obtains one or more of the chylomicron index, hemolysis index, or jaundice index of the blood sample based on the plurality of sample images, including: the processor Performs normalization processing on each of the plurality of sample images to obtain a plurality of normalized images; Based on the plurality of normalized images, obtains one or more of the chylomicron index, hemolysis index, or jaundice index of the blood sample.
13. A method for detecting the quality of serum, including: Rotating a container containing a blood sample; Taking pictures of the blood sample in the container during the rotation of the container to obtain a plurality of sample images of the blood sample; Obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the plurality of sample images.
14. The method according to claim 13, wherein Obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the plurality of sample images, including: Obtaining one or more serum images including a serum region based on the plurality of sample images; Obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the serum image.
15. The method according to claim 14, characterized in that, Obtaining one or more serum images including a serum region based on the plurality of sample images, including: Performing image feature extraction on each of the plurality of sample images to obtain a first image feature of each sample image, and based on the first image feature of each sample image, screening out the serum image from the plurality of sample images; and / or Inputting the plurality of sample images into a first neural network model to obtain the output of the first neural network model as the serum image.
16. The method according to claim 14 or 15, characterized in that Obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the serum image, including: Performing image feature extraction on the serum image to obtain a second image feature of the serum image, and based on the second image feature of the serum image, obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample; and / or Input the serum image or the serum region image cropped from the serum image into a second neural network model, and use the output of the second neural network model as one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.
17. The method according to claim 16, wherein The second image feature includes one or more of color feature, grayscale histogram, and image gradient.
18. The method according to claim 16, wherein The serum region image is cropped from the serum image in the following manner: Remove the support part region of the sample base in the serum image to obtain the serum region image.
19. The method according to any one of claims 13-18 further includes: Output the obtained index; and / or When the obtained index exceeds a preset threshold, output an alarm prompt indicating that the blood sample is abnormal.
20. The method according to any one of claims 13-19, characterized in that, Obtaining one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample based on the plurality of sample images includes: Normalize each of the plurality of sample images to obtain a plurality of normalized images; Based on the plurality of normalized images, obtain one or more of the chylomicron index, hemolysis index, and jaundice index of the blood sample.