CN116609330A - Method and device for identifying interferents in serum and plasma - Google Patents
Method and device for identifying interferents in serum and plasma Download PDFInfo
- Publication number
- CN116609330A CN116609330A CN202310633518.8A CN202310633518A CN116609330A CN 116609330 A CN116609330 A CN 116609330A CN 202310633518 A CN202310633518 A CN 202310633518A CN 116609330 A CN116609330 A CN 116609330A
- Authority
- CN
- China
- Prior art keywords
- serum
- plasma
- blood
- image
- collection tube
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 210000002966 serum Anatomy 0.000 title claims abstract description 158
- 238000000034 method Methods 0.000 title claims abstract description 43
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 24
- 238000001514 detection method Methods 0.000 claims abstract description 16
- 239000007788 liquid Substances 0.000 claims abstract description 16
- 238000013135 deep learning Methods 0.000 claims abstract description 9
- 210000002381 plasma Anatomy 0.000 claims description 131
- 210000004369 blood Anatomy 0.000 claims description 89
- 239000008280 blood Substances 0.000 claims description 89
- 230000011218 segmentation Effects 0.000 claims description 49
- 230000002452 interceptive effect Effects 0.000 claims description 10
- 102000009123 Fibrin Human genes 0.000 abstract description 13
- 108010073385 Fibrin Proteins 0.000 abstract description 13
- BWGVNKXGVNDBDI-UHFFFAOYSA-N Fibrin monomer Chemical compound CNC(=O)CNC(=O)CN BWGVNKXGVNDBDI-UHFFFAOYSA-N 0.000 abstract description 13
- 229950003499 fibrin Drugs 0.000 abstract description 13
- 238000005070 sampling Methods 0.000 abstract description 7
- 230000002159 abnormal effect Effects 0.000 abstract description 3
- 238000005516 engineering process Methods 0.000 abstract description 3
- 238000000926 separation method Methods 0.000 description 13
- 208000007536 Thrombosis Diseases 0.000 description 12
- 230000000903 blocking effect Effects 0.000 description 4
- 239000012530 fluid Substances 0.000 description 3
- 239000000126 substance Substances 0.000 description 3
- 102100037242 Amiloride-sensitive sodium channel subunit alpha Human genes 0.000 description 2
- 101000740448 Homo sapiens Amiloride-sensitive sodium channel subunit alpha Proteins 0.000 description 2
- 230000001960 triggered effect Effects 0.000 description 2
- 238000010241 blood sampling Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000005119 centrifugation Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000001900 immune effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/01—Arrangements or apparatus for facilitating the optical investigation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/85—Investigating moving fluids or granular solids
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Biochemistry (AREA)
- Chemical & Material Sciences (AREA)
- Software Systems (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Analytical Chemistry (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Molecular Biology (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Multimedia (AREA)
- Quality & Reliability (AREA)
- Databases & Information Systems (AREA)
- Radiology & Medical Imaging (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention discloses a method and a device for identifying interferents in serum and plasma, which construct a partitioned convolutional neural network by applying a deep learning technology, automatically realize the detection and accurate positioning of the existence of interferents such as fibrin wires or clots in the serum/plasma, provide effective serum/plasma height and liquid amount which are not influenced by the interferents, effectively filter abnormal serum state samples, effectively guide the movement of a sampling needle according to the serum/plasma height and liquid amount which are not influenced by the interferents, avoid the needle blockage of the sampling needle, replace the original method for manually observing and judging whether the interferents such as fibrin wires, clots and the like exist in the serum by a doctor, reduce the workload of the doctor, and improve the sample detection efficiency.
Description
Technical Field
The invention relates to the field of blood sample detection, in particular to a method and a device for identifying interferents in serum and plasma.
Background
Blood sample testing is commonly used in laboratory biochemical and immunological testing projects, where whole blood is typically first centrifuged to rapidly obtain serum/plasma for testing. However, in practice, there may be interference such as fibrin threads or clots in the serum/plasma due to blood sampling, centrifugation, or improper storage. These interferents may not only interfere with detection, but may also cause needle blocking or partial needle blocking. The complete blocking needle can directly lead to the incapability of running of detection equipment, and professional personnel are required to process the complete blocking needle, so that the detection period is influenced, and the detection cost is increased. Partial needle blockage can cause insufficient amount of blood serum to be sucked, and the detection quality is affected.
At present, before a sampling needle sucks serum/plasma, a doctor needs to manually observe and judge whether the serum contains interferences such as fibrin wires, clots and the like, and judge whether the position of the interferences influences detection, so that the method has the defects of strong subjectivity, high requirements on the operation experience of the doctor and high workload. If the doctor in the early stage does not judge that the interference objects such as fibrin wires, clots and the like exist in the serum, the method can only rely on MTS (Mahalanobis-Taguchi System) to detect blockage, but the method can only send an alarm when the blockage occurs, and cannot avoid the occurrence of the blockage.
Disclosure of Invention
The invention aims to provide a method for identifying interferents in serum and plasma, which is used for solving the problem that whether the interferents such as fibrin wires, clots and the like exist in the serum/plasma or not and the exact positions of the interferents such as the fibrin wires, the clots and the like cannot be effectively judged in advance by adopting an automatic technology, and effectively preventing needles from being blocked. Another object of the present invention is to provide a device for recognizing an interfering substance in serum or plasma.
In order to achieve the above purpose, the invention adopts the following technical scheme:
the invention relates to a method for identifying interferents in serum and plasma, which comprises the following steps:
s1, acquiring a plurality of blood collection tube images with different angles;
s2, dividing each blood collection tube image according to categories by using a division convolutional neural network model 1;
s3, calculating the area of the blood serum/plasma area separated from each blood collection tube image, and determining the blood collection tube image with the largest area of the blood serum/plasma area;
s4, detecting an interfering object by using the segmentation convolutional neural network model 2 on the blood serum/plasma region segmentation image of the blood collection tube image selected in the S3;
s5, determining the effective serum/plasma height and the liquid amount without interference according to the detection result of the step S4.
Further, in step S1, the blood collection tube is rotated by a preset rotation angle to capture a plurality of blood collection tube images of different angles including the complete blood collection tube, i.e., the background plate.
Further, in step S2, the segmentation convolutional neural network model 1 is obtained through deep learning training, and segmentation is performed on the blood collection tube image according to N categories; the N categories include at least serum/plasma categories.
Further, in step S3, the blood serum/plasma region segmentation image refers to a minimum circumscribed rectangle of a blood collection tube image corresponding to the blood serum/plasma segmentation result; the serum/plasma area is the total number of pixels of the serum/plasma area segmented image.
Further, in step S4, the blood serum/plasma region segmented image of the selected blood collection tube image is segmented again by category using the segmented convolutional neural network model 2 obtained by the deep learning training, and an interfering object in the blood serum/plasma region segmented image is identified.
Further, in step S5, the determination of the effective serum/plasma height and fluid volume without interference includes: for a blood serum/blood plasma area segmentation image in which no interference is detected, taking the bottom of blood serum/blood plasma as the lowest position from which blood serum can be extracted and the top of blood serum/blood plasma as the highest position according to the segmentation result in the step S2, and acquiring the effective blood serum/blood plasma height and liquid amount without interference; for the blood serum/blood plasma area segmentation image of the detected interference object, the top position of the detected interference object is taken as the lowest position of the extractable blood serum, and the top of the blood serum/blood plasma is taken as the highest position, so that the effective blood serum/blood plasma height and liquid amount without interference are obtained.
The invention relates to a device for identifying interferents in serum and plasma, which comprises image acquisition equipment, a light source, a gripper, a background plate, a photoelectric sensor and computing equipment, wherein the computing equipment stores and runs the method for identifying the interferents in serum and plasma according to claim 1;
the photoelectric sensor is used for judging whether the blood collection tube reaches a specified shooting position or not;
the image acquisition equipment is used for acquiring image data of different angles of the blood collection tube and transmitting the image data to the computing equipment;
the gripper is used for gripping the blood collection tube and rotating the blood collection tube according to a preset rotation angle, and assisting the image acquisition equipment to complete image data acquisition of different angles of the blood collection tube;
the background plate is used for reducing the influence of shooting background on the identification accuracy of the identification method of the interferents in the serum and the plasma according to the claim 1;
the light sources are symmetrically arranged on two sides of the image acquisition equipment and form a certain included angle with the image acquisition equipment, so that a proper acquisition environment is provided for the image acquisition equipment.
Further, the background plate is a white baffle with rough surface.
Further, the light source includes a point light source or a bar light source.
The invention has the advantages that the deep learning technology is used for constructing the partitioned convolutional neural network, the detection and accurate positioning of the interferences such as fibrin wires or clots in serum/plasma are automatically realized, the effective serum/plasma height and liquid amount which are not influenced by the interferences are provided, abnormal serum state samples can be effectively filtered, the movement of a sampling needle can be effectively guided according to the serum/plasma height and liquid amount which are not influenced by the interferences, the needle blockage of the sampling needle is avoided, the original method for judging whether the interferences such as fibrin wires, clots and the like exist in the serum by means of manual observation of doctors is replaced, the workload of doctors is reduced, and the sample detection efficiency is improved.
Drawings
Fig. 1 is a flow chart of the method of the present invention.
Fig. 2 is a top view of a photograph of a label being captured by a camera in the method of the present invention.
Fig. 3 is a front view of a camera facing a tag in the method of the present invention.
Fig. 4 is a top view of a photograph of a camera facing away from a tag in the method of the present invention.
Fig. 5 is a front view of a camera facing away from a tag in the method of the present invention.
FIG. 6 is a schematic view showing the segmentation of an image of a blood collection tube without a separator in the method of the present invention.
FIG. 7 is a schematic view showing the segmentation of an image of a blood collection tube containing a separator gel in the method of the present invention.
Fig. 8 is a schematic view of the apparatus of the present invention.
Detailed Description
The following description of the technical solutions in the embodiments of the present invention will be clear and complete, and it is obvious that the described embodiments are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
EXAMPLE 1 detailed description of the method for identifying an interfering substance in serum or plasma according to the present invention
As shown in fig. 1, the method for identifying the interferents in the serum and the plasma comprises the following steps:
s1, acquiring a plurality of blood collection tube images with different angles;
since the label for recording personal information or the like is attached to the surface of the blood collection tube, it is necessary to obtain a region having the minimum shielding position of the label in consideration of the position of the label. Therefore, a plurality of blood collection tube images of different angles are photographed by rotating the blood collection tube at a preset rotation angle. Each blood collection tube image should contain the complete blood collection tube, the tube cap at the bottom and top of the blood collection tube, and the tube diameter of the whole blood collection tube. While each tube image should contain a partial background plate image. The image size may be 2048X700 pixels. In other embodiments, 1500X450 pixels are also possible. That is, the pixel condition can be adjusted according to the image capturing apparatus and the photographing condition that are actually adopted. The purpose of rotating the blood collection tube to collect multiple images is to be able to collect the image of the blood collection tube with the least shielding of the tag, i.e. the image of the largest area of the effective serum area.
As shown in fig. 2, the label 2 of the blood collection tube 1 faces the camera 3 of the image collection device, and due to the shielding of the label 2, the collected image is shown in fig. 3, only the cap 4 of the blood collection tube 1, the label 2 and the blood clot area 5 at the lower part can be seen, and the effective serum/plasma area 6 is shielded by the label.
As shown in fig. 4, the label 2 of the blood collection tube 1 faces away from the camera 3 of the image collection device, and at this time, the collected image is shown in fig. 5, so that the cap 4, the label 2, the serum/plasma region 6, the separation gel 7 and the blood clot 5 of the blood collection tube 1 can be seen, and the effective serum/plasma region 6 can be clearly seen.
S2, dividing each blood collection tube image according to categories by using a division convolutional neural network model 1;
the obtained blood collection tube images of the blood collection tubes are respectively introduced into the divided convolutional neural network model 1, and each blood collection tube image is divided by category by the convolutional neural network model 1. The split convolutional neural network model 1 is obtained through deep learning training, and any split convolutional neural network algorithm, such as a split algorithm Unet, can be adopted. Any computational framework, program or tool box may also be used for encoding, for example, frameworks such as Tensorflow, pytorch, etc.
The segmentation convolutional neural network model 1 obtained by the deep learning training can segment the blood collection tube image according to N segmentation classes. The N categories must include serum/plasma categories. Any of blood clots, separation gel, labels, air, caps, etc. may also be included depending on the actual situation.
S3, calculating the area of the blood serum/plasma area divided in each blood collection tube image, and determining the blood collection tube image with the largest area of the blood serum/plasma area and the blood serum/plasma area image;
and calculating the area of the blood serum/plasma area divided in each blood collection tube image, namely counting the number of pixels corresponding to the blood serum/plasma division result, and obtaining a plurality of areas of the blood serum/plasma area. Comparing the areas of the serum/plasma areas to find a blood collection tube image with the largest area of the serum/plasma areas and a serum/plasma area segmentation image corresponding to the blood collection tube image;
the serum/plasma area refers to the total number of pixels of the serum/plasma area divided image.
The blood serum/blood plasma region segmentation image refers to the minimum circumscribed rectangle of the blood collection tube image corresponding to the blood serum/blood plasma segmentation result.
The image of the blood collection tube with the largest area of the serum/plasma area is found out, so that the influence of shielding of the tag is reduced as much as possible. Because if the interfering object in the serum/plasma area is located behind the tag, it will be impossible to recognize it by image because the camera cannot capture the interfering object. Therefore, the condition that the interference objects are blocked by the labels and cannot be detected can be effectively reduced or avoided by adopting the blood collection tube image with the largest serum/plasma area.
S4, detecting an interfering object by using the segmentation convolutional neural network model 2 on the blood serum/plasma region segmentation image of the blood collection tube image selected in the S3;
and (3) introducing the blood serum/blood plasma region segmentation image of the blood collection tube image selected in the step (S3) into a fine tuning segmentation convolutional neural network model 2, and segmenting again according to categories to identify the interferents in the blood serum/blood plasma region segmentation image.
The partitioned convolutional neural network model 2 may be any partitioned convolutional neural network model, and may be trained with the partitioned convolutional neural network model 1 to obtain a plurality of partitioned result categories.
In one embodiment, the two types of interferents, namely fibrin threads and clots, may be present, and in other embodiments, more types of interferents may be present, or may be considered as one type, without distinguishing between the detailed types of interferents.
And re-segmenting the blood serum/blood plasma region segmentation image of the selected blood collection tube image by category by using the segmentation convolutional neural network model 2 obtained through deep learning training, and identifying an interference object in the blood serum/blood plasma region segmentation image.
The serum/serum region is imported into a refined fine tuning segmentation network model (SCNN 2), which may be any segmentation convolutional neural network model that may be trained with SCNN1, may obtain multiple segmentation result categories, in one embodiment, may be two of fibrin filaments, clots in interferents, in other embodiments, may be more interferent categories, or may be considered a class, without distinguishing the interferent detailed categories. Here, the purpose of further using the segmentation model is to obtain the interferents more accurately, because the interferents are very fine for the whole blood collection tube image, for example, for a 2048X700 blood collection tube image, the interferents may be only in the range of 10X10 pixels. Therefore, the SCNN1 is adopted to determine the serum/plasma area, the segmentation range is reduced, and the detection of the interference objects can be facilitated.
S5, determining the effective serum/plasma height and the liquid amount without interference according to the detection result of the step S4.
The determination of the effective serum/plasma height and fluid volume without interference includes: for a blood serum/blood plasma area segmentation image in which no interference is detected, taking the bottom of blood serum/blood plasma as the lowest position from which blood serum can be extracted and the top of blood serum/blood plasma as the highest position according to the segmentation result in the step S2, and acquiring the effective blood serum/blood plasma height and liquid amount without interference; for the blood serum/blood plasma area segmentation image of the detected interference object, the top position of the detected interference object is taken as the lowest position of the extractable blood serum, and the top of the blood serum/blood plasma is taken as the highest position, so that the effective blood serum/blood plasma height and liquid amount without interference are obtained.
The blood collection tube according to the present invention includes all types of blood collection tubes, and in the blood collection tube without separation gel, the cap 4, serum/plasma 6, and blood clot 5 in the blood collection tube 1 without separation gel can be separated in step S2, as shown in fig. 6, in which the blood collection tube 1 is divided into the serum/plasma region 6 and the blood clot region 5.
In fig. 6, it is also shown that the interference 8 exists in the serum/plasma region 6, and then the top of the detected interference 8, i.e. the position of the serial number 15, is taken as the lowest position from which serum can be extracted, and the top of the serum/plasma 6, i.e. the position of the serial number 17, is taken as the highest position, and the effective serum/plasma height, i.e. the height represented by the serial number 19, is obtained without interference, and the amount of the effective serum/plasma is obtained. In fig. 6, reference numeral 18 denotes a distance from the top of the blood clot 5, i.e., reference numeral 16, to the top of the serum/plasma 6, i.e., reference numeral 17, which is the serum plasma height. If no interfering substance 8 is present, the effective serum/plasma height is defined by the serum/plasma height (number 18) and the fluid volume determined by the top position of the blood clot 5 (number 16) and the top position of the serum/plasma 5 (number 17).
In the blood collection tube containing the separation gel, in step S2, the cap 4, the serum/plasma 6, the separation gel 7, and the blood clot 5 in the blood collection tube 1 containing the separation gel can be separated, and as shown in fig. 7, the blood collection tube 1 is divided into the serum/plasma region 6, the blood clot region 5, and the separation gel region 7. In the serum/plasma zone 6 shown in fig. 7, the height of the serum/plasma, i.e. number 18, and the amount of serum/plasma is determined by the top position of the separation gel 7, i.e. number 20, and the top position of the serum/plasma 6, i.e. number 17.
In addition, according to the segmentation results of blood clot and separation gel of the blood collection tube in the step S2, the heights and volumes of the blood clot and the separation gel can be obtained easily.
The amount of serum/plasma in the present invention is calculated based on the serum/plasma height. Based on the serum/serum height and the known tube diameter information, a serum/plasma volume calculation is performed.
The invention provides a method for detecting interferents (fibrin wires and clots) in serum/plasma by using a segmented convolutional neural network end-to-end. Firstly, a segmentation network is adopted to obtain a serum/plasma region, and based on the area comparison of the serum/plasma region in a segmentation result, a blood collection tube image with least label shielding is obtained; detecting the interferents in serum/plasma by adopting a fine tuning segmentation network; the blood serum/plasma height and the liquid amount can be obtained based on the segmentation model, the separation gel height and the separation gel volume and the blood clot height can be used for effectively filtering abnormal blood serum samples, the movement of a sampling needle can be effectively guided according to the blood serum/plasma height and the liquid amount which are not influenced by interference objects, the needle blockage of the sampling needle is avoided, the original method for judging whether the interference objects such as fibrin wires, clots and the like exist in blood serum by relying on manual observation of doctors is replaced, the workload of doctors is reduced, and the detection efficiency of the samples is improved.
The invention relates to a device for identifying interferents in serum and plasma, which is shown in fig. 8, and comprises an image acquisition device camera 3, a light source 11, a grip 9, a background plate 10, a photoelectric sensor 12 and a computing device 13 for storing and operating the method for identifying interferents in serum and plasma according to claim 1;
the photoelectric sensor 12 is used for judging whether the blood collection tube reaches a specified shooting position; when the blood collection tube 1 reaches a designated position, the photoelectric sensor 12 can be triggered, the built-in control system is triggered, the top gripper 9 grips the blood collection tube 1 to rotate 360 degrees for image acquisition, and a plurality of images with different angles are shot.
The image acquisition device camera 3 is used for acquiring image data of blood collection tubes at different angles and transmitting the image data to the computing device 13;
the gripper 9 is used for gripping the blood collection tube 1 from the tube support 14 and rotating the blood collection tube according to a preset rotation angle until 360-degree rotation of the blood collection tube 1 is completed. The auxiliary image acquisition equipment camera 3 is used for completing image data acquisition of blood collection tubes at different angles;
the background plate 10 is used for reducing the influence of the shooting background on the recognition accuracy of the method for recognizing the interferents in the serum and the plasma according to claim 1, and is a white baffle with rough surface. Thus, the influence of the background color problem on the serum area of the blood collection tube can be reduced as much as possible.
The light sources 11 are symmetrically arranged on two sides of the image acquisition equipment and form a certain included angle with the image acquisition equipment camera 3, so that a proper acquisition environment is provided for the image acquisition equipment camera 3, and the image acquisition is performed under the environment with sufficient illumination. The light source 11 may be a Light Emitting Diode (LED) including a point light source or a bar light source. The angle between the light source 11 and the camera 3 of the image acquisition device can be adjusted according to the actual situation.
Claims (9)
1. A method for identifying interferents in serum and plasma is characterized by comprising the following steps of: the method comprises the following steps:
s1, acquiring a plurality of blood collection tube images with different angles;
s2, dividing each blood collection tube image according to categories by using a division convolutional neural network model 1;
s3, calculating the area of the blood serum/plasma area divided in each blood collection tube image, and determining the blood collection tube image with the largest area of the blood serum/plasma area and the blood serum/plasma area division image;
s4, detecting an interfering object by using the segmentation convolutional neural network model 2 on the blood serum/plasma region segmentation image of the blood collection tube image selected in the S3;
s5, determining the effective serum/plasma height and the liquid amount without interference according to the detection result of the step S4.
2. The method and device for identifying the interferents in the serum and the plasma according to claim 1, wherein the method comprises the following steps: and S1, rotating the blood collection tube at a preset rotation angle to shoot a plurality of blood collection tube images with different angles, wherein the blood collection tube images comprise a complete blood collection tube, namely a background plate.
3. The method and device for identifying the interferents in the serum and the plasma according to claim 1, wherein the method comprises the following steps: s2, the segmentation convolutional neural network model 1 is obtained through deep learning training, and segmentation is carried out on blood collection tube images according to N categories; the N categories include at least serum/plasma categories.
4. The method and device for identifying the interferents in the serum and the plasma according to claim 1, wherein the method comprises the following steps: s3, the blood serum/blood plasma region segmentation image refers to the minimum circumscribed rectangle of the blood collection tube image corresponding to the blood serum/blood plasma segmentation result; the serum/plasma area is the total number of pixels of the serum/plasma area segmented image.
5. The method and device for identifying the interferents in the serum and the plasma according to claim 1, wherein the method comprises the following steps: and S4, re-segmenting the blood serum/blood plasma region segmentation image of the selected blood collection tube image by category by using the segmentation convolutional neural network model 2 obtained through deep learning training, and identifying an interfering object in the blood serum/blood plasma region segmentation image.
6. The method and device for identifying the interferents in the serum and the plasma according to claim 1, wherein the method comprises the following steps: in step S5, the determination process of the effective serum/plasma height and liquid amount without interference comprises the following steps: for a blood serum/blood plasma area segmentation image in which no interference is detected, taking the bottom of blood serum/blood plasma as the lowest position from which blood serum can be extracted and the top of blood serum/blood plasma as the highest position according to the segmentation result in the step S2, and acquiring the effective blood serum/blood plasma height and liquid amount without interference; for the blood serum/blood plasma area segmentation image of the detected interference object, the top position of the detected interference object is taken as the lowest position of the extractable blood serum, and the top of the blood serum/blood plasma is taken as the highest position, so that the effective blood serum/blood plasma height and liquid amount without interference are obtained.
7. An identification device for interferents in serum and plasma, which is characterized in that: a computing device comprising an image acquisition device, a light source, a gripper, a background plate, a photosensor, and a method of storing and operating the method of identifying interferents in serum and plasma of claim 1;
the photoelectric sensor is used for judging whether the blood collection tube reaches a specified shooting position or not;
the image acquisition equipment is used for acquiring image data of different angles of the blood collection tube and transmitting the image data to the computing equipment;
the gripper is used for gripping the blood collection tube and rotating the blood collection tube according to a preset rotation angle, and assisting the image acquisition equipment to complete image data acquisition of different angles of the blood collection tube;
the background plate is used for reducing the influence of shooting background on the identification accuracy of the identification method of the interferents in the serum and the plasma according to the claim 1;
the light sources are symmetrically arranged on two sides of the image acquisition equipment and form a certain included angle with the image acquisition equipment, so that a proper acquisition environment is provided for the image acquisition equipment.
8. The device for identifying interferents in serum or plasma according to claim 7, wherein: the background plate is a white baffle with rough surface.
9. The device for identifying interferents in serum or plasma according to claim 7, wherein: the light source comprises a point light source or a bar light source.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310633518.8A CN116609330A (en) | 2023-05-31 | 2023-05-31 | Method and device for identifying interferents in serum and plasma |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310633518.8A CN116609330A (en) | 2023-05-31 | 2023-05-31 | Method and device for identifying interferents in serum and plasma |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116609330A true CN116609330A (en) | 2023-08-18 |
Family
ID=87685215
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310633518.8A Pending CN116609330A (en) | 2023-05-31 | 2023-05-31 | Method and device for identifying interferents in serum and plasma |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116609330A (en) |
-
2023
- 2023-05-31 CN CN202310633518.8A patent/CN116609330A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109477848B (en) | Systems, methods, and apparatus for identifying sample container lids | |
| US5647025A (en) | Automatic focusing of biomedical specimens apparatus | |
| JP2019504997A (en) | Method and apparatus configured to quantify a sample from a multilateral perspective | |
| CN107977682B (en) | Lymphocyte classification method and device based on polar coordinate transformation data enhancement | |
| CN103487104B (en) | A kind of image acquisition identification system and method for capillary pipe liquid level type data | |
| US20050163354A1 (en) | Method and device for the analysis of body fluids | |
| WO2011055791A1 (en) | Device for harvesting bacterial colony and method therefor | |
| JPH0652263B2 (en) | Cell analyzer | |
| US9589174B2 (en) | Imaging mass spectrometer and mass spectrometry data processing method | |
| CN111340798A (en) | Application of deep learning in product appearance flaw detection | |
| CN105139384B (en) | The method and apparatus of defect capsule detection | |
| JP2019158871A (en) | Automatic urine tracking method and apparatus | |
| CN107389681A (en) | A kind of image interpretation method of full-automatic blood type analytical instrument | |
| AU2022283636B2 (en) | A microfluidic image analysis system | |
| CN114820475B (en) | Edge identification method and system, wafer processing device and method for determining concentric state of wafer and processing table | |
| JP2022507121A (en) | Methods and equipment for pia mater imaging | |
| US20220358647A1 (en) | Device and method for attesting the operations of an in-vitro diagnostic device | |
| CN116609330A (en) | Method and device for identifying interferents in serum and plasma | |
| CN111398138A (en) | Optical detection system and method of dry type blood cell analysis device | |
| CN114324382A (en) | Panel terminal cleanliness detection method and panel terminal cleanliness detection device | |
| TW201404878A (en) | Device for automatically rapidly analyzing biological cells and related method thereof | |
| JP2010151523A (en) | Method and device for analyzing particle image | |
| WO2023031622A1 (en) | System and method for identifying and counting biological species | |
| CN114878844A (en) | Method, system and equipment for automatically detecting quality of centrifuged blood sample | |
| CN113302472B (en) | Biological sample detection device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |