Search

CN-122017237-A - Method and system for measuring blood coagulation and blood coagulation inhibition experimental result

CN122017237ACN 122017237 ACN122017237 ACN 122017237ACN-122017237-A

Abstract

The invention discloses a method and a system for measuring blood coagulation and blood coagulation inhibition experimental results, wherein the method comprises the steps of placing a 96-well plate for blood coagulation or blood coagulation inhibition experiments to be measured in an opaque closed cavity, embedding the 96-well plate in a clamping groove on an LED backlight plate, starting the LED backlight plate and inclining the LED backlight plate to a preset angle to enable red blood cells to generate corresponding displacement at the bottom of the hole, standing for preset time to obtain an image containing all hole sites, respectively extracting multiple types of characteristics from each hole site in the image, inputting the characteristics into a corresponding basic classification model for independent judgment, inputting the image and output results of a plurality of basic classification models into three artificial intelligent models for completing training to determine the state of each hole site, and obtaining blood coagulation or blood coagulation inhibition experimental results according to the states of all hole sites. The invention has the advantages of low learning cost for users, low equipment cost, automation, rapid and simultaneous completion of the measurement of a plurality of samples, and higher accuracy by using the fused characteristic engineering technology to highlight the characteristics.

Inventors

  • HUO XIANG
  • YU HUIYAN
  • LIU DEYE
  • WANG SHENJIAO
  • ZHU LIGUO

Assignees

  • 江苏省疾病预防控制中心(江苏省预防医学科学院)

Dates

Publication Date
20260512
Application Date
20260415

Claims (10)

  1. 1. The method for measuring the results of the experiments for blood coagulation and blood coagulation inhibition is characterized by comprising the following steps: Step 1, a 96-hole plate (5) for a blood coagulation to be tested or a blood coagulation inhibition experiment is arranged in a non-light-transmitting closed cavity (1), embedded in a clamping groove on an LED backlight plate (3), wherein the LED backlight plate (3) is controllably inclined, and an industrial camera (6) is arranged vertically above the LED backlight plate (3); Step 2, starting the LED backlight plate (3) and tilting to a preset angle to enable red blood cells to generate corresponding displacement at the bottom of the hole, and after standing for preset time, shooting the 96-hole plate (5) once through the industrial camera (6) to obtain an image containing all hole sites; Step 3, extracting multiple types of features from each hole site in the image respectively, and inputting a corresponding basic classification model for independent judgment; and 4, inputting the output results of the images and the plurality of basic classification models into three artificial intelligent models which are trained, determining the state of each hole site by adopting an integration rule that at least two of the three artificial intelligent models output non-agglutination results and are judged to be non-agglutination, and acquiring the hemagglutination or hemagglutination inhibition experimental result according to the states of all the hole sites.
  2. 2. The method for measuring the experimental results of the hemagglutination and hemagglutination inhibition according to claim 1, wherein the opaque closed cavity (1) is internally provided with an LED backlight plate (3), a fixed supporting rod (7), a telescopic supporting rod (4) and an industrial camera (6), The LED backlight plate (3) adopts a white LED light source and is fixed in the center of the bottom of the cavity, a clamping groove is formed in the upper surface of the LED backlight plate (3) and is used for placing a 96-pore plate (5), the fixed support rod (7) is connected to one side of the bottom of the LED backlight plate (3), the telescopic support rod (4) is connected to the other side of the bottom of the LED backlight plate (3) and is driven by a stepping motor and is used for driving the LED backlight plate (3) and the 96-pore plate (5) above the LED backlight plate to integrally incline to a preset angle; The industrial camera (6) is fixedly arranged in the light-tight closed cavity (1), is positioned right above the 96-hole plate (5) under the state that the LED backlight plate (3) is inclined to a preset angle, and is connected with an external computer.
  3. 3. The method for determining the results of experiments for blood coagulation and blood coagulation inhibition as claimed in claim 1, wherein in step 3, the first type of features are extracted from each hole site in the image, and the first basic classification model is input for determination, comprising: Step 31, for each hole site, acquiring a blue component light intensity gradient and a green component light intensity gradient on a radial path from the upper edge of the hole site to a near-center region; Step 32, fitting the blue component light intensity gradient and the green component light intensity gradient respectively in an exponential decay mode to obtain respective fitting coefficients; and step 33, determining a judging result of the first type of features according to the magnitude relation between the blue component fitting coefficient and a preset blue threshold value and the magnitude relation between the green component fitting coefficient and a preset green threshold value.
  4. 4. The method of claim 3, wherein in step 32, the blue component light intensity gradient and the green component light intensity gradient are fitted by using the following exponential decay forms to obtain respective fitting coefficients : ; Wherein, the The color is represented by a color of the color, Is of a color Intensity component at A gradient is provided at the point of the gradient, Representing the position of the pixel point on the longitudinal path, Is the radius of the hole site, Is the maximum aggregation radius of red blood cells at the center of the hole site; If the fitting coefficient of the blue component is larger than a preset blue threshold value, judging that the blue component of the first type of characteristics is not aggregated; and if the green component fitting coefficient is larger than a preset green threshold value, judging that the first-class characteristic green components are not aggregated.
  5. 5. The method of determining the results of experiments for blood coagulation and blood coagulation inhibition as set forth in claim 1, wherein in step 3, extracting a second type of feature for each hole site in the image, inputting a second basic classification model for determination, comprises: Step 34, judging whether a red deposition area contour exists or not by adopting a contour detection algorithm in the maximum aggregation radius of red cells at the center of each hole site; if the contour is not detected, the second type of feature is judged to be non-agglutinated; If the contour is detected, acquiring the number of long-path pixels and the number of short-path pixels of the contour; and 35, calculating the ratio of the number of the long-path pixels to the number of the short-path pixels, comparing the ratio with a preset ratio threshold, judging that the second type of features are agglutinated if the ratio is larger than the preset ratio threshold, and judging that the second type of features are not agglutinated if the ratio is not larger than the preset ratio threshold.
  6. 6. The method for determining the results of experiments for blood coagulation and blood coagulation inhibition as claimed in claim 1, wherein in step 3, extracting a third type of feature for each hole site in the image, inputting a third basic classification model for determination, comprises: Step 36, making a vertical line at a half position of the distance from the bottom edge to the center of each hole site, wherein the vertical line is intersected with the hole site edge to form a transverse scanning line segment, and acquiring a blue light intensity sequence and a green light intensity sequence point by point from the edge to the center along the scanning line segment; step 37, counting continuous points smaller than a preset low light intensity threshold value in a blue light intensity sequence and a green light intensity sequence respectively, and determining a judging result of the third type of characteristics based on the continuous point threshold value n and a preset element number threshold value; if a low light intensity and substantially continuous region appears in the blue light intensity sequence, determining that the third type of characteristic blue component is agglutinated; If a low light intensity and substantially continuous region occurs in the green light intensity sequence, the third type of characteristic green component is determined to be agglutinated.
  7. 7. The method of determining the results of hemagglutination and hemagglutination inhibition assay of claim 6, wherein step 37 comprises: Step 371, obtaining a light intensity sequence S comprising Individual pixel points Wherein, the method comprises the steps of, The number of the pixel point is indicated, Representing the total number of pixels in the light intensity sequence S, 、 Respectively denoted by the number The position of the transverse path where the pixel point is located and the light intensity; step 372, traverse From 1 to Calculation of Dimension vector Wherein the linear mapping function The linear function passes through the point And Will be positioned Corresponding to the ordinate value on the straight line Constructing into vectors ; Step 373, traversing From 1 to Will be positioned Corresponding light intensity Establishment of Dimension vector ; Step 374, construction Dimension vector Front of the vector The individual elements are And Difference between (a) and (b) back 0 Elements are the vectors are padded; Step 375, traverse From 1 to Construction of Dimension matrix ; Step 376 building a collection Wherein, the method comprises the steps of, Counting threshold value for continuous point number, each pixel point in the light intensity sequence S All corresponding to vectors If (if) In a continuous way The number of the elements is negative and, the pixel point is then Incorporating a set L, wherein the set L represents a region where the light intensity sequence S has a light intensity dip; Step 377, counting the number of elements in the set L, if the number is greater than the preset threshold At this time, the third type of feature is judged to be agglutinated.
  8. 8. The method for measuring the results of experiments for blood coagulation and blood coagulation inhibition as set forth in claim 1, wherein the step 4 comprises: step 41, forming a feature vector by a plurality of judgment values corresponding to the multiple types of features and the RGB channel ordered one-dimensional expansion values of the hole site image; Step 42, respectively inputting the feature vectors into an artificial intelligent model for training, wherein the artificial intelligent model comprises a support vector machine, a random forest and a logistic regression model, and obtaining respective independent prediction results; and 43, finally judging that the hole site is not agglutinated when at least two of the three artificial intelligent models output the non-agglutinated result, otherwise judging that the hole site is agglutinated.
  9. 9. The method for determining the experimental results of the blood coagulation and blood coagulation inhibition as claimed in claim 8, wherein the three artificial intelligent models of the support vector machine, the random forest and the logistic regression are obtained by training in advance by using a labeling data set, and the labeling data set is formed by taking a consistent result after independent judgment of a plurality of professionals.
  10. 10. A blood coagulation and blood coagulation inhibition experiment result measuring system, characterized by comprising: The LED backlight plate (3) is arranged in the light-tight sealed cavity (1), a clamping groove for placing a 96-hole plate (5) is formed in the upper surface of the LED backlight plate, the LED backlight plate (3) is configured to be inclined controllably, and the industrial camera (6) is fixedly arranged in the light-tight sealed cavity (1) and is positioned vertically above the 96-hole plate (5) when the LED backlight plate (3) is in an inclined state; A control and processing unit (2) electrically connected with the LED backlight board (3) and the industrial camera (6), respectively, the control and processing unit (2) comprising: the image acquisition module is used for controlling the LED backlight plate (3) to be started and inclined to a preset angle, and after standing for preset time, controlling the industrial camera (6) to shoot the 96-pore plate (5) once so as to obtain an image containing all pore positions; The basic classification module comprises a plurality of basic classification models and is used for respectively extracting multiple types of features from each hole site in the image and independently judging to obtain a plurality of basic judgment results; The integrated judging module comprises three artificial intelligent models for completing training, is used for inputting the images and the plurality of basic judging results into the three artificial intelligent models for completing training, adopts at least two integrated rules for judging that the three artificial intelligent models are not aggregated when the three artificial intelligent models output the non-aggregation results, determines the state of each hole site, and acquires the experimental result of hemagglutination or hemagglutination inhibition according to the states of all the hole sites.

Description

Method and system for measuring blood coagulation and blood coagulation inhibition experimental result Technical Field The invention belongs to the technical field of medical examination and artificial intelligence intersection, and particularly relates to a system and a method for measuring results of hemagglutination and hemagglutination inhibition experiments of influenza, which are particularly suitable for influenza A-H1N1 and B-Victoria types. Background The HA and HI are classical and core serology and virology experimental techniques in influenza virus detection, identification and research, supplement each other, are indispensable means in influenza prevention and control, laboratory monitoring and basic research, and are important in full-chain work such as influenza virus detection, typing, antibody evaluation, vaccine research and development and effect verification, and are also standard detection methods of WHO influenza monitoring networks of the world health organization. At present, experimental results mainly depend on human eye recognition, are large in workload and limited by technical capability and personal experience, and the recognition of each experimenter has large subjectivity, so that the results are unstable. Therefore, it is highly desirable to develop a device that can replace human power to improve the efficiency and accuracy of the experiment. The artificial intelligence algorithm in the traditional image recognition field can be used for judging the HA and HI results, but in a real situation, the heterogeneity of the HA and HI results is obvious, and the accuracy of the traditional algorithm is relatively low. Therefore, an identification scheme highly matched with the experiment is developed, the accuracy is improved, and the method has practical value. Disclosure of Invention Aiming at the problems that the experiments for inhibiting the blood coagulation and the blood coagulation depend on manual interpretation, have strong subjectivity and poor consistency of results and the defects that the existing image recognition algorithm has insufficient accuracy for judging the heterogeneous results, the invention provides a method and a system for measuring the results of the experiments for inhibiting the blood coagulation and the blood coagulation of human A-H1N1 and B-Victoria influenza types. The technical scheme of the invention is as follows: in a first aspect, the present invention provides a method for determining the results of hemagglutination and hemagglutination inhibition experiments, comprising: step 1, placing a 96-well plate to be tested for hemagglutination or hemagglutination inhibition experiment in an opaque closed cavity, embedding the 96-well plate into a clamping groove on an LED backlight plate, wherein the LED backlight plate can be tilted controllably, and an industrial camera is arranged vertically above the LED backlight plate; Step 2, starting the LED backlight plate and tilting to a preset angle to enable red blood cells to generate corresponding displacement at the bottom of the hole, and after standing for preset time, shooting the 96-hole plate once through the industrial camera to obtain an image containing all hole sites; Step 3, extracting multiple types of features from each hole site in the image respectively, and inputting a corresponding basic classification model for independent judgment; and 4, inputting the output results of the images and the plurality of basic classification models into three artificial intelligent models which are trained, determining the state of each hole site by adopting an integration rule that at least two of the three artificial intelligent models output non-agglutination results and are judged to be non-agglutination, and acquiring the hemagglutination or hemagglutination inhibition experimental result according to the states of all the hole sites. Further, an LED backlight plate, a fixed supporting rod, a telescopic supporting rod and an industrial camera are arranged in the light-tight closed cavity, The LED backlight plate adopts a white LED light source and is fixed in the center of the bottom of the cavity, and the upper surface of the LED backlight plate is provided with a clamping groove for placing a 96 pore plate; the industrial camera is fixedly arranged in the light-tight closed cavity, is positioned right above the 96-hole plate under the state that the LED backlight plate is inclined to a preset angle, and is connected with an external computer. Further, in step 3, extracting a first type of feature from each hole site in the image, and inputting a first basic classification model to determine, including: Step 31, for each hole site, acquiring a blue component light intensity gradient and a green component light intensity gradient on a radial path from the upper edge of the hole site to a near-center region; Step 32, fitting the blue component light intensity gradient and the green component ligh