A sample feeder

Abstract

The application provides a sample feeder. The sample feeder comprises a walking chassis configured to be able to walk on a predetermined track, a sample rack moving mechanism for moving a target sample rack from a sample bin to the walking chassis and adjusting the position of the target sample rack on the walking chassis so that the target test tube is in an identification position, a test tube rotating mechanism for rotating the target test tube in the identification position, and an image acquisition and processing module for acquiring images of the rotating target test tube, obtaining corresponding barcode information based on the acquired images containing the barcode, obtaining the actual height of the sample liquid surface based on the acquired images containing the sample liquid surface height information, and obtaining the serum quality identification result of the sample based on the acquired images containing the serum quality information. According to the application, the problems of the existing full-automatic blood sample analyzer, such as the influence of the sample detection progress by the manual test tube placing link, the long overall time consumption of the sampling link and the inaccurate detection result, can be solved.

Description

Technical Field

[0001] This invention belongs to the field of medical devices and relates to a sample feeder, and more specifically, to a sample feeder for use in a fully automated analytical instrument that uses blood samples as the testing object. The fully automated analytical instrument includes, but is not limited to, a fully automated biochemical analyzer and a fully automated immunoassay analyzer. Background Technology

[0002] In recent years, with the continuous deepening of population aging in my country, the number of patients seeking medical treatment in hospitals has been increasing. Against this backdrop, secondary and higher-level hospitals have generally adopted Laboratory Information Management Systems (LIS) to achieve barcode management of patients' blood samples, and have also adopted fully automated blood sample analyzers to achieve high-speed, high-throughput testing of blood samples. However, existing fully automated blood sample analyzers mainly suffer from the following three problems in use: 1. For fully automated blood sample analyzers without a sample pretreatment module or Laboratory Automation System (LAS), operators must place the test tubes containing the sample onto the sample rack, ensuring the barcode on the tube is facing the predetermined direction. This allows the barcode to be scanned by the analyzer's barcode scanner after the sample rack is transported to the sampling position, enabling subsequent sample testing. However, because the barcode on the test tubes must be facing the predetermined direction, the manual sample placement method is not only time-consuming and labor-intensive, affecting the sample testing progress, but also prone to cases where the barcode is not facing the predetermined direction. In such cases, the barcode scanner will be unable to recognize the barcode on the corresponding test tube, thus affecting subsequent sample testing, which will undoubtedly further impact the sample testing progress.

[0003] 2. Before the fully automated blood sample analyzer tests the sample in the test tube, the sample level in the tube needs to be measured. Only after obtaining the sample level information will the sampling arm control the sampling needle mounted on it to draw up the sample. This is because there are corresponding standards for the depth to which the sampling needle is inserted into the sample liquid. If the sampling needle is inserted too deeply, the outer wall of the sampling needle will be contaminated, thus affecting the test results of the next sample; if the sampling needle is inserted too shallowly, insufficient sample volume may occur, thus affecting the test results of the current sample.

[0004] Current fully automated blood sample analyzers typically measure the sample liquid level using the following methods: A capacitive liquid level sensor is installed on the sample dispensing arm. Driven by the drive mechanism, the sample dispensing arm moves the capacitive liquid level sensor closer and closer to the sample liquid surface until the capacitive liquid level sensor detects the sample liquid surface, thereby realizing the measurement of the sample liquid level height.

[0005] However, in order to accurately measure the height of the sample liquid level, instrument manufacturers usually adjust the capacitive liquid level sensor to be more sensitive. Correspondingly, the driving speed of the drive mechanism is set to be relatively low. This means that the process of the capacitive liquid level sensor continuously approaching the sample liquid level takes a long time, resulting in a longer overall sampling time.

[0006] Third, existing fully automated blood sample analyzers cannot identify jaundice, lipemia, and hemolysis-type unqualified samples. The accuracy of test results obtained based on these unqualified and meaningless samples is poor, which may affect the patient's subsequent diagnosis and treatment, and also waste test reagents. Summary of the Invention

[0007] The purpose of this invention is to solve the problems of existing fully automated blood sample analyzers, which affect the sample testing progress due to special requirements for the placement of test tubes in the sample rack, have a long overall sampling time due to the time-consuming sample liquid level measurement step, and have inaccurate test results due to the inability to identify unqualified samples in advance.

[0008] To achieve the above objectives, the present invention provides a sample feeder comprising: The chassis is configured to travel on a predetermined track; The sample rack moving mechanism is used to move the target sample rack from the sample compartment to the walking chassis by supporting it, and to adjust the position of the target sample rack on the walking chassis so that the target test tubes therein are in a predetermined identification position. A test tube rotation mechanism is used to rotate the target test tube located at the identification position; The image acquisition and processing module is used to acquire images of the rotating target test tube to obtain images containing barcodes, images containing sample liquid level information, and images containing serum quality information. A barcode localization algorithm based on orientation consistency is used to locate the barcode in the image containing the barcode, and the barcode is decoded to obtain the barcode information. The pixel height of the sample liquid surface is obtained from the image containing the sample liquid surface height information, and the actual height of the sample liquid surface is obtained based on the pre-acquired mapping relationship between pixel size and actual size. Based on the first acquired image containing serum quality information, the qualitative identification result of the sample's serum quality is obtained based on a pre-built serum quality identification model. When the serum quality of the sample is unqualified, the quantitative identification result of the sample's serum quality is obtained based on multiple subsequently acquired images containing serum quality information and a pre-acquired serum quality benchmark color image.

[0009] Optionally, the image acquisition and processing module is further configured to, after acquiring the barcode information, output a test tube alignment command based on the position information of the barcode in the image containing the barcode. The test tube alignment command is used to control the test tube rotation mechanism to rotate the target test tube by a corresponding angle so that the orientation of the barcode is consistent with the barcode orientation required by the barcode scanner of the sample analyzer.

[0010] Optionally, the image acquisition and processing module is also used to send the acquired barcode information, the actual height of the sample liquid level, and the serum quality identification results of the sample to the sample analyzer.

[0011] Optionally, the sampler further includes: The supplementary lighting module is used to provide supplementary lighting during the image acquisition and processing process of the image acquisition and processing module. A light-shielding reflector is arranged in the direction of the light output of the supplementary lighting module to form a semi-open image acquisition chamber.

[0012] Optionally, the method of locating the barcode in the image containing the barcode using a barcode localization algorithm based on orientation consistency includes: The image containing the barcode is divided into blocks to obtain multiple image blocks; Background image blocks are filtered out from the multiple image blocks to obtain multiple image blocks to be consistent with the calculation. Obtain the gradient direction consistency of each image block to be consistent, and take the image block whose gradient direction consistency reaches the predetermined gradient direction consistency threshold as the valid image block; Perform a connectivity operation on valid image blocks with the same average gradient direction to obtain multiple connected regions; Obtain the minimum bounding rectangle region corresponding to the multiple connected regions, and perform background filtering on the minimum bounding rectangle region to obtain the barcode region image.

[0013] Optionally, obtaining the pixel height of the sample liquid surface from the image containing the sample liquid surface height information includes: The image containing the sample liquid level information is subjected to color enhancement processing; The color-enhanced image containing sample liquid level information is converted to the HSV color space; Binarize the image containing sample liquid level information in the HSV color space to highlight the uppermost serum region; Image processing is performed on prominent serum regions, including erosion and dilation. Obtain the minimum bounding rectangle region corresponding to the serum region after image processing, and use the pixel height of the upper edge of the minimum bounding rectangle region as the pixel height of the sample liquid surface.

[0014] Optionally, the method for obtaining the mapping relationship between the pixel size and the actual size includes: The camera calibration board is pre-set to acquire images of the camera calibration board. Obtain the pixel dimensions of the marker blocks in the camera calibration board image and the pixel dimensions of the distance between two adjacent marker blocks; The mapping relationship between the pixel size and the actual size is obtained based on the pixel size of the obtained identifier block and the pixel size of the distance between two adjacent identifier blocks, and based on the actual size of the pre-obtained identifier block and the actual size of the distance between two adjacent identifier blocks.

[0015] Optionally, the step of obtaining the qualitative identification result of the sample's serum quality based on the initially acquired image containing serum quality information and a pre-built serum quality identification model specifically involves: The image containing serum quality information is input into the serum quality recognition model to obtain the qualitative recognition result of the serum quality of the sample; The qualitative identification results of serum quality include two categories: qualified and unqualified. The unqualified category includes any one of the following three types, any two of them combined, or all of them combined: Jaundice, lipemia, hemolysis.

[0016] Optionally, when the serum quality of a sample is substandard, obtaining the quantitative identification result of the sample's serum quality based on multiple subsequently acquired images containing serum quality information and a pre-acquired serum quality reference color image includes: Clustering operations are performed on multiple images containing serum quality information to obtain multiple image sets; Determine the optimal image set among the plurality of image sets; A fusion operation is performed on the images containing serum quality information within the optimal image set to obtain a serum color image; The Euclidean distance for color space conversion is determined based on the case of substandard serum quality, and the serum color image is converted to the HSV color space based on the Euclidean distance. Semi-quantitative values ​​corresponding to serum quality deficiencies are obtained based on the serum color image in the HSV color space and the corresponding serum quality benchmark color image.

[0017] Optionally, when the serum quality non-compliance is of a single type, the method for obtaining the serum quality reference color image includes: Multiple image acquisitions are performed on the reference color blocks corresponding to the pre-set corresponding non-conforming types to obtain multiple reference color block images; Clustering operations are performed on the multiple reference color block images to obtain multiple image sets; Determine the optimal image set among the plurality of image sets; A fusion operation is performed on the reference color patch images within the optimal image set to obtain a reference color image; The reference color image is converted to the HSV color space to obtain the serum quality reference color image; When the serum quality failure is a combination of factors, the method for obtaining the serum quality reference color image includes: Obtain the serum quality baseline color image corresponding to each non-compliant type; The serum quality baseline color images corresponding to each non-compliant type are fused to obtain the serum quality baseline color image under the combined type.

[0018] The beneficial effects of this invention are as follows: Firstly, the sample feeder of the present invention employs an image acquisition and processing module that uses a barcode positioning algorithm based on orientation consistency to locate the barcode in a pre-acquired image containing barcodes, and then decodes the barcode to obtain barcode information. After acquiring the barcode information of each test tube in the target sample rack, the image acquisition and processing module sends all barcode information to the sample analyzer. Based on the acquired barcode information, the sample analyzer interacts with the hospital's laboratory information management system to obtain the corresponding testing items for each sample in each test tube. Therefore, after configuring the sample feeder of the present invention, the sample analyzer no longer needs to have a barcode scanning step for the test tubes; the barcode information of test tubes with random barcode orientation can be acquired directly at the sample feeder, eliminating the need for operators to adjust the barcode orientation after placing the test tubes on the sample rack, as is typically required in existing technologies.

[0019] Secondly, the sample feeder of the present invention has an image acquisition and processing module that acquires the pixel height of the sample liquid level based on a pre-acquired image containing sample liquid level information, and acquires the actual height of the sample liquid level based on the mapping relationship between the pre-acquired pixel size and the actual size, thereby realizing the measurement of the sample liquid level height in the test tube. Therefore, for a sample analyzer equipped with the sample feeder of the present invention, the measurement of the sample liquid level height in the test tube can be achieved at the sample feeder based on image processing technology. Compared with the sample liquid level measurement method used in existing fully automated blood sample analyzers, this sample liquid level measurement method based on image processing technology has a faster measurement rate, resulting in a shorter time for the corresponding sample analyzer in the sample liquid level measurement stage.

[0020] Thirdly, the sample feeder of the present invention includes an image acquisition and processing module that acquires a qualitative identification result of the sample's serum quality based on an initially acquired image containing serum quality information and a pre-built serum quality identification model. Furthermore, when the serum quality of a sample is unqualified, it acquires a quantitative identification result of the sample's serum quality based on multiple subsequently acquired images containing serum quality information and a pre-acquired serum quality reference color image. After acquiring the serum quality identification result, the image acquisition and processing module sends the serum quality identification result to the sample analyzer. For samples with qualified serum quality identification results, the sample analyzer performs normal testing; for samples with unqualified serum quality identification results, the sample analyzer will not perform testing and will execute predetermined measures corresponding to samples with unqualified serum quality identification results.

[0021] Based on the above three aspects, it can be seen that after adopting the sample feeder of the present invention, the corresponding sample analyzer can effectively solve the problems of existing fully automatic blood sample analyzers, which are affected by the special requirements for the placement of test tubes in the sample rack, resulting in a long overall sampling time due to the long time required for measuring the sample liquid level, and inaccurate test results due to the inability to identify unqualified samples in advance.

[0022] Other features and advantages of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0023] The present invention can be better understood by referring to the following description taken in conjunction with the accompanying drawings, in which the same or similar reference numerals are used throughout the drawings to denote the same or similar parts.

[0024] Figure 1 A schematic diagram of the sample feeder according to an embodiment of the present invention is shown; Figure 2 A schematic diagram of the layout of the sample chamber, the sampler, and the external track according to an embodiment of the present invention is shown, wherein 001 is the sampler and 002 is the sampling position; Figure 3 A schematic diagram showing the relative positional relationship between the sample feeder and the sample rack when the first test tube on the sample rack is in the identification position according to an embodiment of the present invention is shown, wherein 003 is the test tube in the identification position, 004 is the sample liquid level, 005 is the test tube barcode, 006 is the lens position, and 007 is the test tube position barcode on the sample rack. Figure 4 A schematic diagram showing the relative positional relationship between the image acquisition and processing module and the supplementary lighting module according to an embodiment of the present invention is provided. Figure 5A schematic diagram showing the relative positional relationship between the image acquisition and processing module, the supplementary lighting module, and the light-shielding reflector according to an embodiment of the present invention is provided. Figure 6 A schematic diagram of an image containing sample liquid level information is shown according to an embodiment of the present invention; Figure 7 A schematic diagram of an image containing sample liquid level height information in the HSV color space according to an embodiment of the present invention is shown; Figure 8 A schematic diagram of a binarized image containing sample liquid level information is shown according to an embodiment of the present invention; Figure 9 A schematic diagram of the minimum bounding rectangle corresponding to the serum region according to an embodiment of the present invention is shown, wherein 008 is the minimum bounding rectangle; Figure 10 A schematic diagram of serum samples with different levels of lipemia according to an embodiment of the present invention is shown; Figure 11 A schematic diagram of serum samples with different degrees of jaundice is shown according to an embodiment of the present invention; Figure 12 A schematic diagram of serum samples with different degrees of hemolysis according to an embodiment of the present invention is shown; Figure 13 A schematic diagram of the test tube rotating mechanism from a first perspective according to an embodiment of the present invention is shown; Figure 14 A schematic diagram of the test tube rotation mechanism from a second perspective, according to an embodiment of the present invention, is shown. Detailed Implementation

[0025] To enable those skilled in the art to more fully understand the technical solutions of the present invention, exemplary embodiments of the present invention will be described more comprehensively and in detail below with reference to the accompanying drawings. Obviously, the one or more embodiments of the present invention described below are merely one or more specific ways to implement the technical solutions of the present invention, and are not exhaustive. It should be understood that other ways belonging to a general inventive concept can be used to implement the technical solutions of the present invention, and should not be limited to the embodiments described exemplary. Based on one or more embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0026] Example: Figure 1 A schematic diagram of the sample feeder according to an embodiment of the present invention is shown. (Refer to...) Figure 1 The sampler in this embodiment of the invention includes: The chassis 1000 is configured to travel on a predetermined track; The sample rack moving mechanism 100 is used to move the target sample rack from the sample compartment to the walking chassis 1000 by supporting it, and to adjust the position of the target sample rack on the walking chassis 1000 so that the target test tubes therein are in a predetermined identification position. The test tube rotation mechanism 200 is used to rotate the target test tube located in the identification position. The image acquisition and processing module 300 is used to acquire images of the rotating target test tube to obtain images containing barcodes, images containing sample liquid level information, and images containing serum quality information. A barcode localization algorithm based on orientation consistency is used to locate barcodes in an image containing barcodes, and the located barcodes are decoded to obtain barcode information. The pixel height of the sample liquid surface is obtained from the image containing the sample liquid surface height information, and the actual height of the sample liquid surface is obtained based on the mapping relationship between the pre-acquired pixel size and the actual size. Based on the initially acquired image containing serum quality information and a pre-built serum quality identification model, the qualitative identification result of the sample's serum quality is obtained. Furthermore, when the sample's serum quality is substandard, the quantitative identification result of the sample's serum quality is obtained based on multiple subsequently acquired images containing serum quality information and a pre-acquired serum quality benchmark color image. The acquired barcode information, the actual height of the sample liquid level, and the serum quality identification results of the sample are sent to the sample analyzer.

[0027] Specifically, in this embodiment of the invention, the sample rack moving mechanism 100 includes an up-and-down motor actuator 110, a left-and-right motor actuator 120, and a test tube rack support 130. The up-and-down motor actuator 110 and the left-and-right motor actuator 120 work together to make the test tube rack support 130 support the target sample rack located in the sample compartment. Then, the left-and-right motor actuator 120 drives the test tube rack support 130 to move to a predetermined position on the walking chassis 1000, thereby making the target test tube in the target sample rack in a predetermined identification position.

[0028] Specifically, Figure 13 This diagram illustrates the structure of the test tube rotating mechanism from a first perspective according to an embodiment of the present invention. Figure 14 A schematic diagram of the test tube rotating mechanism from a second perspective, according to an embodiment of the present invention, is shown. (Refer to...) Figure 1 , Figure 13 and Figure 14In this embodiment of the invention, the test tube rotation mechanism 200 includes a roller structure 210 and a double bearing auxiliary structure 220. The roller structure 210 includes a roller 211 and a first motor 212 for driving the roller 211. The double bearing auxiliary structure 220 includes a mechanism 222 with two bearings 221 at its front end, a second motor 223, and a gear 224. When the target test tube moves horizontally to the identification position along with the test tube rack support 130, the corresponding part of the target test tube contacts the side of the roller 211. At the same time, the second motor 223 controls the gear 224 to rotate. Since the bottom of the mechanism 222 is provided with a bearing 224, the gear 224 rotates. The gear 224 meshes with the rack, and driven by the gear 224, it translates towards the target test tube along a path perpendicular to the extension direction of the sample holder until the two bearings 221 at the front end of the mechanism 222 contact the target test tube. When the roller 211 and the two bearings 221 are in contact with the target test tube, the roller 211 and the two bearings 221 provide three-point restraint for the target test tube. The two bearings 221 are configured to perform in-situ fixed-axis rotation, and their outer contact surfaces are both smooth. The roller 211 is made of polyurethane to provide sufficient friction when there is a relative motion tendency between the roller 211 and the target test tube. Then, the first motor 212 controls the roller 211 to rotate, and the rotating roller 211 drives the target test tube to perform in-situ fixed-axis rotation, while the two bearings 221 play a supporting and restraining role during the rotation of the target test tube.

[0029] Specifically, in this embodiment of the invention, the image acquisition and processing module 300 has a predetermined processing sequence, namely, sequentially executing the processing steps of acquiring the barcode information of the target test tube, acquiring the actual height of the sample liquid level in the target test tube, and acquiring the serum quality identification result of the sample. If the result corresponding to a certain processing step cannot be acquired at a certain processing step, the processing step cannot be skipped to the next processing step. Instead, a corresponding loop operation needs to be performed in order to acquire the result corresponding to that processing step. If the result corresponding to that processing step is still not acquired when the loop operation reaches a predetermined upper limit, the sample feeder reports an error through the sample analyzer.

[0030] Specifically, Figure 2 A schematic diagram showing the layout of the sample chamber, sample feeder, and external track according to an embodiment of the present invention is provided. (Refer to...) Figure 2In this embodiment of the invention, the sample chamber 400 has multiple test tube rack placement areas 410 and a sample chamber track 420. Each test tube rack placement area 410 can hold multiple sample racks, and each sample rack can hold multiple test tubes, each containing a blood sample to be tested. After the image acquisition and processing module 300 acquires the barcode information corresponding to each test tube on the target sample rack, the actual height of the sample liquid level, and the quantitative identification result of the serum mass of the sample, the image acquisition and processing module 300 sends this information to the sample analyzer. At the same time, the walking chassis 1000 travels along the sample chamber track 420 to the sample rack inflow end of the outer track 500. The sample rack moving mechanism 100 causes the target sample rack to flow into the outer track 500, and the outer track 500 transports the inflowing target sample rack to the sampling position of the sample analyzer.

[0031] Furthermore, in this embodiment of the invention, the image acquisition and processing module 300 is also used to output a test tube alignment command based on the position information of the barcode in the image containing the barcode after acquiring the barcode information. The test tube alignment command is used to control the test tube rotation mechanism 200 to rotate the target test tube by a corresponding angle so that the orientation of the barcode is consistent with the barcode orientation required by the barcode scanner of the sample analyzer.

[0032] Specifically, in addition to recognizing the barcode on the target test tube, the sample feeder in this embodiment of the invention can also orient the target test tube so that the barcode scanner of the sample analyzer can successfully scan the barcode on the target test tube. This is because, to ensure that the sampling needle of the sample analyzer draws the sample from the correct test tube, the sample analyzer needs to perform a secondary scan of the barcode on the test tube containing the sample before sampling to reconfirm the sample.

[0033] Furthermore, Figure 3 This diagram illustrates the relative positional relationship between the sample feeder and the sample rack when the first test tube on the sample rack is in the identification position, according to an embodiment of the present invention. Figure 4 This diagram illustrates the relative positional relationship between the image acquisition and processing module and the supplementary lighting module according to an embodiment of the present invention. Figure 5 The diagram illustrates the relative positions of the image acquisition and processing module, the supplementary lighting module, and the light-shielding reflector in an embodiment of the present invention.

[0034] Reference Figures 3-5 The sampler in this embodiment of the invention further includes: The supplementary lighting module 600 is used to provide supplementary lighting during the image acquisition process of the image acquisition and processing module 300; A light-shielding reflector 700 is positioned in the direction of the light output of the supplementary lighting module 600 to form a semi-open image acquisition chamber.

[0035] Specifically, in this embodiment of the invention, the supplementary lighting module 600 is implemented using two LED light strip boards. Both LED light strip boards and the image acquisition and processing module 300 are located on the first side of the target sample holder. The two LED light strip boards are arranged opposite each other, respectively located on both sides of the lens of the image acquisition and processing module 300 and both in front of the lens of the image acquisition and processing module 300. A light-shielding reflector 700 is disposed on the second side of the target sample holder.

[0036] Specifically, the sample feeder in this embodiment of the invention further includes a camera calibration plate 800 and a reference color block plate 900. The camera calibration plate 800 and the reference color block plate 900 are disposed on the second side of the target sample holder and are covered by a light-shielding reflector 700. The camera calibration plate 800 and the reference color block plate 900 are respectively located on both sides of the recognition position. The camera calibration plate 800 is used for the calibration and positioning of the camera part of the image acquisition and processing module 300. Three reference color blocks are distributed on the reference color block plate 900, which respectively provide initial reference colors for identifying jaundice, lipemia, and hemolysis samples.

[0037] Furthermore, in this embodiment of the invention, the step of locating the barcode in the image containing the barcode using a barcode positioning algorithm based on orientation consistency includes: The image containing the barcode is divided into blocks to obtain multiple image blocks; Background image blocks are filtered out from the multiple image blocks to obtain multiple image blocks to be consistent with the calculation. Obtain the gradient direction consistency of each image block to be consistent, and take the image block whose gradient direction consistency reaches the predetermined gradient direction consistency threshold as the valid image block; Perform a connectivity operation on valid image blocks with the same average gradient direction to obtain multiple connected regions; Obtain the minimum bounding rectangle region corresponding to the multiple connected regions, and perform background filtering on the minimum bounding rectangle region to obtain the barcode region image.

[0038] Furthermore, in this embodiment of the invention, obtaining the pixel height of the sample liquid surface based on the image containing sample liquid surface height information includes: The image containing the sample liquid level information is subjected to color enhancement processing; The color-enhanced image containing sample liquid level information is converted to the HSV color space; Binarize the image containing sample liquid level information in the HSV color space to highlight the uppermost serum region; Image processing is performed on prominent serum regions, including erosion and dilation. Obtain the minimum bounding rectangle region corresponding to the serum region after image processing, and use the pixel height of the upper edge of the minimum bounding rectangle region as the pixel height of the sample liquid surface.

[0039] Specifically, in this embodiment of the invention, a color lookup table is used to perform color enhancement processing on the image containing the sample liquid level information. The image containing the sample liquid level information is as follows: Figure 6 As shown, Figure 6 The blood sample is divided into three layers: upper, middle, and lower. The upper layer is serum, the middle layer is separating gel, and the lower layer is red blood cells. The serum in the upper layer is the sample sample taken and detected by the analyzer. An image containing sample liquid level information in the HSV color space is shown below. Figure 7 As shown. The binarized image containing the sample liquid level height information is as follows. Figure 8 As shown, Figure 8 The white area in the diagram represents the serum region. The minimum bounding rectangle corresponding to the serum region is shown below. Figure 9 As shown.

[0040] Furthermore, in this embodiment of the invention, the method for obtaining the mapping relationship between pixel size and actual size includes: The camera calibration board is pre-set to acquire images of the camera calibration board. Obtain the pixel dimensions of the marker blocks in the camera calibration board image and the pixel dimensions of the distance between two adjacent marker blocks; The mapping relationship between the pixel size and the actual size is obtained based on the pixel size of the obtained identifier block and the pixel size of the distance between two adjacent identifier blocks, and based on the actual size of the pre-obtained identifier block and the actual size of the distance between two adjacent identifier blocks.

[0041] Specifically, in this embodiment of the invention, the mapping relationship between the pixel size and the actual size is expressed by the following formula: X = ax + b In the above formula, X is the actual size, x is the pixel size, a is the calibration factor, and b is the compensation amount.

[0042] Furthermore, in this embodiment of the invention, the step of obtaining the qualitative identification result of the serum quality of the sample based on the first acquired image containing serum quality information and a pre-built serum quality identification model specifically involves: The image containing serum quality information is input into the serum quality recognition model to obtain the qualitative recognition result of the serum quality of the sample; The qualitative identification results of serum quality include two categories: qualified and unqualified. The unqualified category includes any one of the following three types, any two of them combined, or all of them combined: Jaundice, lipemia, hemolysis.

[0043] Specifically, in this embodiment of the invention, the image containing serum quality information is a serum region image.

[0044] Specifically, in the embodiments of the present invention, Figure 10 The diagram illustrates serum samples with different levels of lipemia according to embodiments of the present invention, wherein the degree of lipemia increases sequentially from left to right. Figure 11 A schematic diagram of serum samples with different degrees of jaundice, according to an embodiment of the present invention, is shown, wherein the degree of jaundice increases sequentially from left to right. Figure 12 The illustration shows schematic diagrams of serum samples with different degrees of hemolysis according to embodiments of the present invention, wherein the degree of hemolysis increases sequentially from left to right. The serum quality identification model in this embodiment is implemented based on a convolutional neural network. Specifically, the convolutional neural network is trained on images of serum samples with different degrees of lipemia, jaundice, hemolysis, and mixed-type serum samples to obtain the serum quality identification model. The mixed-type serum sample images with different degrees of mixing include a mixture of lipemia and jaundice, a mixture of lipemia and hemolysis, a mixture of jaundice and hemolysis, and a mixture of lipemia, jaundice, and hemolysis.

[0045] Specifically, in this embodiment of the invention, the serum quality identification model identifies serum samples as unqualified based on the following criteria: taking lipemia type as an example, if the color depth of the target sample is greater than that of a moderate lipemia serum sample and the magnitude of the difference reaches a predetermined threshold, then the target sample is determined to be an unqualified lipemia serum sample.

[0046] Specifically, in this embodiment of the invention, the three types of reference color blocks distributed on the reference color block plate 900 are reference color blocks corresponding to lipemia type, reference color blocks corresponding to jaundice type, and reference color blocks corresponding to hemolysis type. These three types of reference color blocks are respectively based on Figure 10 The color of a serum sample with moderate lipemia Figure 11 The color of serum samples with moderate jaundice and Figure 12 The color was created from a moderately hemolyzed serum sample.

[0047] Furthermore, in this embodiment of the invention, when the serum quality of the sample is substandard, obtaining the quantitative identification result of the serum quality of the sample based on multiple subsequently acquired images containing serum quality information and a pre-acquired serum quality reference color image includes: Clustering operations are performed on multiple images containing serum quality information to obtain multiple image sets; Determine the optimal image set among the plurality of image sets; A fusion operation is performed on the images containing serum quality information within the optimal image set to obtain a serum color image; The Euclidean distance for color space conversion is determined based on the case of substandard serum quality, and the serum color image is converted to the HSV color space based on the Euclidean distance. Semi-quantitative values ​​corresponding to serum quality deficiencies are obtained based on the serum color image in the HSV color space and the corresponding serum quality benchmark color image.

[0048] Furthermore, in this embodiment of the invention, when the serum quality failure is of a single type, the method for obtaining the serum quality reference color image includes: Multiple image acquisitions are performed on the reference color blocks corresponding to the pre-set corresponding non-conforming types to obtain multiple reference color block images; Clustering operations are performed on the multiple reference color block images to obtain multiple image sets; Determine the optimal image set among the plurality of image sets; A fusion operation is performed on the reference color patch images within the optimal image set to obtain a reference color image; The reference color image is converted to the HSV color space to obtain the serum quality reference color image; When the serum quality failure is a combination of factors, the method for obtaining the serum quality reference color image includes: Obtain the serum quality baseline color image corresponding to each non-compliant type; The serum quality baseline color images corresponding to each non-compliant type are fused to obtain the serum quality baseline color image under the combined type.

[0049] The sample feeder of this invention eliminates the need for operators to place test tubes on the sample rack according to a specific barcode orientation, greatly reducing the workload of operators and avoiding problems in subsequent sample testing due to improper barcode orientation. It also eliminates the need to detect the sample liquid level at the sampling position of the sample analyzer, saving time that can be used for sample aspiration by the sample loading arm and rinsing of the sampling needle. This not only improves the quality of test results but also increases the testing throughput of the sample analyzer. Furthermore, it can detect samples with severe jaundice, lipemia, or hemolysis in advance, reminding operators to handle them properly and improving the accuracy of test results.

[0050] While one or more embodiments of the present invention have been described above, those skilled in the art will recognize that the present invention can be implemented in any other form without departing from its spirit and scope. Therefore, the embodiments described above are illustrative and not restrictive, and many modifications and substitutions will be apparent to those skilled in the art without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. A sample feeder, characterized by, include: The chassis is configured to travel on a predetermined track; The sample rack moving mechanism is used to move the target sample rack from the sample compartment to the walking chassis by supporting it, and to adjust the position of the target sample rack on the walking chassis so that the target test tubes therein are in a predetermined identification position. A test tube rotation mechanism is used to rotate the target test tube located at the identification position; The image acquisition and processing module is used to acquire images of the rotating target test tube to obtain images containing barcodes, images containing sample liquid level information, and images containing serum quality information. A barcode localization algorithm based on orientation consistency is used to locate the barcode in the image containing the barcode, and the barcode is decoded to obtain the barcode information. The pixel height of the sample liquid surface is obtained from the image containing the sample liquid surface height information, and the actual height of the sample liquid surface is obtained based on the pre-acquired mapping relationship between pixel size and actual size. Based on the first image containing serum quality information, the qualitative identification result of the serum quality of the sample is obtained based on the pre-built serum quality identification model. When the serum quality of the sample is unqualified, the quantitative identification result of the serum quality of the sample is obtained based on multiple images containing serum quality information and the pre-acquired serum quality reference color image. The step of locating the barcode in the image containing the barcode using a barcode localization algorithm based on orientation consistency includes: The image containing the barcode is divided into blocks to obtain multiple image blocks; Background image blocks are filtered out from the multiple image blocks to obtain multiple image blocks to be consistent with the calculation. Obtain the gradient direction consistency of each image block to be consistent, and take the image block whose gradient direction consistency reaches the predetermined gradient direction consistency threshold as the valid image block; Perform a connectivity operation on valid image blocks with the same average gradient direction to obtain multiple connected regions; Obtain the minimum bounding rectangle region corresponding to the multiple connected regions, and perform background filtering on the minimum bounding rectangle region to obtain the barcode region image; The specific steps for obtaining the qualitative identification result of serum quality of the sample based on the first acquired image containing serum quality information and a pre-built serum quality identification model are as follows: The image containing serum quality information is input into the serum quality recognition model to obtain the qualitative recognition result of the serum quality of the sample; The qualitative identification results of serum quality include two categories: qualified and unqualified. The unqualified category includes any one of the following three types, any two of them combined, or all of them combined: Jaundice, lipemia, hemolysis; When the serum quality of a sample is substandard, the method for obtaining a quantitative identification result of the sample's serum quality based on multiple subsequently acquired images containing serum quality information and a pre-acquired serum quality reference color image includes: Clustering operations are performed on multiple images containing serum quality information to obtain multiple image sets; Determine the optimal image set among the plurality of image sets; A fusion operation is performed on the images containing serum quality information within the optimal image set to obtain a serum color image; The Euclidean distance for color space conversion is determined based on the case of substandard serum quality, and the serum color image is converted to the HSV color space based on the Euclidean distance. Semi-quantitative values ​​corresponding to serum quality deficiencies are obtained based on the serum color image in the HSV color space and the corresponding serum quality benchmark color image.

2. The sample presenter of claim 1, wherein, The image acquisition and processing module is also used to output a test tube alignment command based on the position information of the barcode in the image containing the barcode after acquiring the barcode information. The test tube alignment command is used to control the test tube rotation mechanism to rotate the target test tube by a corresponding angle so that the orientation of the barcode is consistent with the barcode orientation required by the barcode scanner of the sample analyzer.

3. The sample presenter of claim 1, wherein, The image acquisition and processing module is also used to send the acquired barcode information, the actual height of the sample liquid level, and the serum quality identification results of the sample to the sample analyzer.

4. The sample presenter of claim 1, wherein, Also includes: The supplementary lighting module is used to provide supplementary lighting during the image acquisition and processing process of the image acquisition and processing module. A light-shielding reflector is arranged in the direction of the light output of the supplementary lighting module to form a semi-open image acquisition chamber.

5. The sample presenter of claim 1, wherein, The step of obtaining the pixel height of the sample liquid surface based on the image containing the sample liquid surface height information includes: The image containing the sample liquid level information is subjected to color enhancement processing; The color-enhanced image containing sample liquid level information is converted to the HSV color space; Binarize the image containing sample liquid level information in the HSV color space to highlight the uppermost serum region; Image processing is performed on prominent serum regions, including erosion and dilation. Obtain the minimum bounding rectangle region corresponding to the serum region after image processing, and use the pixel height of the upper edge of the minimum bounding rectangle region as the pixel height of the sample liquid surface.

6. The sample feeder of claim 5, wherein, The method for obtaining the mapping relationship between pixel size and actual size includes: The camera calibration board is pre-set to acquire images of the camera calibration board. Obtain the pixel dimensions of the marker blocks in the camera calibration board image and the pixel dimensions of the distance between two adjacent marker blocks; The mapping relationship between the pixel size and the actual size is obtained based on the pixel size of the obtained identifier block and the pixel size of the distance between two adjacent identifier blocks, and based on the actual size of the pre-obtained identifier block and the actual size of the distance between two adjacent identifier blocks.

7. The sample feeder according to claim 1, characterized in that, When the serum quality failure is of a single type, the method for obtaining the serum quality reference color image includes: Multiple image acquisitions are performed on the reference color blocks corresponding to the pre-set corresponding non-conforming types to obtain multiple reference color block images; Clustering operations are performed on the multiple reference color patch images to obtain multiple image sets; Determine the optimal image set among the plurality of image sets; A fusion operation is performed on the reference color patch images within the optimal image set to obtain a reference color image; The reference color image is converted to the HSV color space to obtain the serum quality reference color image; When the serum quality failure is a combination of factors, the method for obtaining the serum quality reference color image includes: Obtain the serum quality baseline color image corresponding to each non-compliant type; The serum quality baseline color images corresponding to each non-compliant type are fused to obtain the serum quality baseline color image under the combined type.

Images