Automatic pathogenic fungus identification method based on image

Abstract

The invention relates to the field of pathogenic microorganism detection, and discloses an image-based pathogenic fungus automatic identification method, which comprises the following steps of: using a serial number in a fungus classification catalog table in an interpersonal infectious pathogenic microorganism catalog (2023 edition) as a serial number of each fungus; acquiring parameters in the fitting formula; and unknown pathogenic fungi cultured from a clinical specimen are automatically identified. The method has the advantages that the pathogenic fungi can be automatically identified, and the accuracy and efficiency of pathogenic fungi detection are greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of pathogenic microorganism detection, and particularly relates to an automatic recognition method of pathogenic fungi based on images. Background Art

[0002] The detection of pathogenic fungi is one of the important tasks in microbiological inspection.

[0003] Currently, the detection of pathogenic fungi mainly relies on manual operation. However, manual detection not only has low efficiency but also often leads to misjudgment. At the same time, it takes a lot of time to train a professional who can proficiently detect pathogenic fungi, while the clinical demand for professional pathogen detection personnel is relatively large. Summary of the Invention

[0004] The present invention proposes an automatic recognition method of pathogenic fungi based on images, which solves the problems of low efficiency and frequent misjudgment in manual detection.

[0005] An automatic recognition method of pathogenic fungi based on images according to the present invention includes the following steps:

[0006] Step S01: Use the serial numbers in the fungal classification catalog table in the Catalog of Pathogenic Microorganisms Transmissible among Humans (2023 Edition) as the numbers of each fungus;

[0007] Step S02: Obtain the parameters in the fitting formula. The specific process is as follows:

[0008] Step S02-01: For the fungi cultured from clinically collected specimens, stain them with lactophenol cotton blue staining solution, observe them every 6 hours using an optical microscope for 30 consecutive times, and collect image data f R (x, y), f G (x, y), f B (x, y), where f R (x, y) is the gray value of the R channel of the color image, f G (x, y) is the gray value of the G channel of the color image, f B (x, y) is the gray value of the B channel of the color image, x is an integer greater than 0 and less than 1921, and y is an integer greater than 0 and less than 1081;

[0009] Step S02-02: Have professional inspectors identify the cultured fungi and give the number Z of the cultured fungi according to the numbering in Step S01;

[0010] Step S02-03: For each piece of collected image data, construct a binary function f 1 (x, y) about x and y using the following formula:

[0011]

[0012] Where x is an integer greater than 0 and less than 1921, and y is an integer greater than 0 and less than 1081;

[0013] Step S02-04: Construct a binary function f of x and y using the following formula 2 (x, y):

[0014]

[0015] Where x is an integer greater than 0 and less than 1921, and y is an integer greater than 0 and less than 1921;

[0016] Step S02-05: Calculate six characteristic parameters C of the binary function f 2 (x, y) 1 , C 2 , C 3 , C 4 , C 5 , C 6 , specifically the process is as follows:

[0017] Step S02-05-01: Construct the following matrix A 1 :

[0018]

[0019] Step S02-05-02: Construct the following matrix B 1 ;

[0020]

[0021] Step S02-05-03: Calculate the characteristic parameter C 1 :

[0022]

[0023] Step S02-05-04: Construct the following matrix A 2 :

[0024]

[0025] Step S02-05-05: Construct the following matrix B 2 :

[0026]

[0027] Step S02-05-06: Calculate the parameter C 2 :

[0028]

[0029] Step S02-05-06: Calculate parameter C 2 :

[0030]

[0031] Step S02-05-07: Construct the following matrix A 3 :

[0032]

[0033] Step S02-05-08: Construct the following matrix B 2 :

[0034]

[0035] Step S02-05-09: Calculate parameter C 3 :

[0036]

[0037] Step S02-05-10: Construct the following matrix A 4 :

[0038]

[0039] Step S02-05-11: Construct the following matrix B 4 :

[0040]

[0041] Step S02-05-12: Calculate parameter C 4 :

[0042]

[0043] Step S02-05-13: Construct the following matrix A 5 :

[0044]

[0045] Step S02-05-14: Construct the following matrix B 5 :

[0046]

[0047] Step S02-05-15: Calculate parameter C 5 :

[0048]

[0049] Step S02-05-16: Construct the following matrix A 6 :

[0050]

[0051] Step S02-05-17: Construct the following matrix B 6 :

[0052]

[0053] Step S02-05-18: Calculate the parameter C 6 :

[0054]

[0055] Step S02-06: Use the scipy.optimize.curve_fit function in the SciPy library of Python to perform parameter fitting on the following formula:

[0056]

[0057] where F i , E i , D ij are the parameters to be fitted, j is an integer from 1 to 6, i is an integer from 1 to N, and N is an integer greater than 5;

[0058] where X 1 , X 2 , X 3 , X 4 , X 5 , X 6 correspond to the parameters C 1 , C 2 , C 3 , C 4 , C 5 , C 6 , and Y corresponds to the number Z in Step S02-02;

[0059] Step S03: Automatically identify the unknown pathogenic fungi cultured from clinical specimens. The specific process is as follows;

[0060] Step S03-01: For the fungi cultured from the clinically collected specimens, stain them with lactophenol cotton blue staining solution, observe them using an optical microscope, and use the parameters F i , E i , D ij fitted in Step S02 and formula (1) to calculate the corresponding Y, denoted as Y 1 ;

[0061] Step S03-02: Round Y 1 and record the calculation result as Y 2 ;

[0062] Step S03-03: Use Y 2 as the number to retrieve the corresponding fungus in the fungal classification catalog table in the Catalog of Pathogenic Microorganisms Transmissible Among Humans (2023 Edition).

[0063] wherein, N is 5 to 15.

[0064] Compared with the existing method, the advantage of the present invention is that it can automatically identify pathogenic fungi, greatly improving the accuracy and efficiency of pathogenic fungi detection. Detailed implementation manners

[0065] The present invention will be further described in detail below in conjunction with embodiments.

[0066] Embodiment: An automatic identification method for pathogenic fungi based on images, comprising the steps of:

[0067] Step S01; Use the serial numbers in the fungal classification catalog table in the Catalog of Pathogenic Microorganisms Transmissible Among Humans (2023 Edition) as the numbers of each fungus;

[0068] Step S02: Obtain the parameters in the fitting formula, and the specific process is as follows:

[0069] Step S02-01: For the fungi cultured from the clinically collected specimens, stain them with lactophenol cotton blue staining solution, observe them using an optical microscope, re-prepare the slides for staining and observe every 6 hours, continuously observe 30 times, and collect the image data f R (x, y), f G (x, y), f B (x, y), where f R (x, y) is the gray value of the R channel of the color image, f G (x, y) is the gray value of the G channel of the color image, f B (x, y) is the gray value of the B channel of the color image, x is an integer greater than 0 and less than 1921, and y is an integer greater than 0 and less than 1081;

[0070] Step S02-02: Have professional inspectors identify the cultured fungi and give the number Z of the cultured fungi according to the number in Step S01;

[0071] Step S02-03: For each collected image data, construct a binary function f 1 (x, y) about x and y using the following formula:

[0072]

[0073] where x is an integer greater than 0 and less than 1921, and y is an integer greater than 0 and less than 1081;

[0074] Step S02 - 04: Construct a binary function f of x and y using the following formula 2 (x, y):;

[0075]

[0076] where x is an integer greater than 0 and less than 1921, and y is an integer greater than 0 and less than 1921;

[0077] Step S02 - 05: Calculate six characteristic parameters C of the binary function f 2 (x, y), and the specific process is as follows: 1 、C 2 、C 3 、C 4 、C 5 、C 6

[0078] Step S02 - 05 - 01: Construct the following matrix A 1 :

[0079]

[0080] Step S02 - 05 - 02: Construct the following matrix B 1 :

[0081]

[0082] Step S02 - 05 - 03: Calculate the characteristic parameter C 1 :

[0083]

[0084] Step S02 - 05 - 04: Construct the following matrix A 2 :

[0085]

[0086] Step S02 - 05 - 05: Construct the following matrix B 2 :

[0087]

[0088] Step S02 - 05 - 06: Calculate the parameter C 2 :

[0089]

[0090] Step S02-05-06: Calculate parameter C 2 :

[0091]

[0092] Step S02-05-07: Construct the following matrix A 3 :

[0093]

[0094] Step S02-05-08: Construct the following matrix B 2 :

[0095]

[0096] Step S02-05-09: Calculate parameter C 3 :

[0097]

[0098] Step S02-05-10: Construct the following matrix A 4 :

[0099]

[0100] Step S02-05-11: Construct the following matrix B 4 :

[0101]

[0102] Step S02-05-12: Calculate parameter C 4 :

[0103]

[0104] Step S02-05-13: Construct the following matrix A 5 :

[0105]

[0106] Step S02-05-14: Construct the following matrix B 5 :

[0107]

[0108] Step S02-05-15: Calculate parameter C 5 :

[0109]

[0110] Step S02-05-16: Construct the following matrix A 6 :

[0111]

[0112] Step S02-05-17: Construct the following matrix B 6 :

[0113]

[0114] Step S02-05-18: Calculate the parameter C 6 :

[0115]

[0116] Step S02-06: Use the scipy.optimize.curve_fit function in the SciPy library of Python to perform parameter fitting on the following formula:

[0117]

[0118] where F i , E i , D ij are the parameters to be fitted, j is an integer from 1 to 6, i is an integer from 1 to N, and N is an integer greater than 5;

[0119] where X 1 , X 2 , X 3 , X 4 , X 5 , X 6 correspond to the parameters C 1 , C 2 , C 3 , C 4 , C 5 , C 6 , and Y corresponds to the number Z in Step S02-02;

[0120] Step S03: Automatically identify the unknown pathogenic fungi cultured from clinical specimens. The specific process is as follows

[0121] Step S03-01: For the fungi cultured from the clinically collected specimens, stain them with lactophenol cotton blue staining solution, observe them using an optical microscope, and calculate the corresponding Y using the F i , E i , D ij parameters and formula (1), and denote it as Y 1 ;

[0122] Step S03-02: For Y1 Perform rounding calculation, and record the calculation result as Y 2 ;

[0123] Step S03-03: Take Y 2 As the number, retrieve the corresponding fungus in the fungus classification catalog table in the Catalog of Pathogenic Microorganisms Transmitted Among Humans (2023 Edition).

[0124] Wherein, the N is 5 to 15.

[0125] It should be noted that in step S02-01, the clinical specimens are observed every 6 hours, and observations and collection of image data are also required at night. At night, the on-duty laboratory technicians complete the preparation of slides, staining, observation, and collection of image data.

[0126] In addition, it should be noted that in step S02-01, during the cultivation process, SDA medium is uniformly used to ensure that the conditions for collecting image data are consistent with the experimental conditions of clinical specimens.

[0127] Furthermore, it should be noted that in step S02-01, during the cultivation process, the temperature of 37°C is first used for cultivation. If the fungus does not grow at this temperature, the temperature of 25°C is then used for cultivation, and image collection is only carried out for the cases where the fungus grows.

[0128] Meanwhile, it should be noted that in step S02-02, to ensure the accuracy and reliability of the identified fungus, the laboratory technicians observing the clinical specimens should have a professional title of intermediate or above and have been engaged in pathogenic microorganism work for more than 3 years. In addition, at least two or more people are required to detect and identify the same specimen.

[0129] It should also be noted that steps S02-03 to S02-06 are to establish a relationship model between the fungus image and the fungus number; among them, the six characteristic parameters C 1 、C 2 、C 3 、C 4 、C 5 、C 6 in step S02-05 represent the characteristic parameters of the fungus image at different scales; the formula (1) in step S02-06 is the fitting formula between the fungus number and the characteristic parameters of the fungus image.

[0130] In this embodiment, in order to determine the optimal value of N in formula (1), all the collected image data are tested, and the following table of the selection of the value of N and the accuracy rate is obtained:

[0131] As can be seen from the table, when the value of N is 11, the accuracy rate has reached 100.00%. Increasing the value of N will increase the calculation time for detection. Therefore, the optimal value of N is taken as 11.

Claims

1. An image-based automatic identification method for pathogenic fungi, characterized in that: Includes steps: Step S01: Use the serial numbers in the fungal classification catalog table in the "Catalogue of Pathogenic Microorganisms Transmitted in Humans" (2023 Edition) as the numbers of each fungus; Step S02: Obtain the parameters in the fitting formula. The specific process is as follows: Step S02-01: For the fungi cultured from the clinical specimens, stain them with lactic acid phenol cotton blue staining solution, observe them once every 6 hours using an optical microscope for 30 consecutive observations, and collect image data. R (x,y),f G (x,y),f B (x, y), where f R (x, y) is the grayscale value of the R channel of the color image, f G (x, y) is the gray value of the G channel of the color image, f B (x, y) is the grayscale value of the B channel of the color image, x is an integer greater than 0 and less than 1921, and y is an integer greater than 0 and less than 1081; Step S02-02: A professional inspector identifies the cultured fungus and gives the cultured fungus a number z according to the number in step S01; Step S02-03: For each acquired image data, a binary function f1(x, y) about x and y is constructed using the following formula: Wherein x is an integer greater than 0 and less than 1921, and y is an integer greater than 0 and less than 1081; Step S02-04: Use the following formula to construct a binary function f2(x, y) about x and y: Wherein x is an integer greater than 0 and less than 1921, and y is an integer greater than 0 and less than 1921; Step S02-05: Calculate the six characteristic parameters C1, C2, C3, C4, C5, and C6 of the binary function f2(x, y). The specific process is as follows: Step S02-05-01: Construct the following matrix A1: Step S02-05-02: Construct the following matrix B1: Step S02-05-03: Calculate characteristic parameter C1: Step S02-05-04: Construct the following matrix A2: Step S02-05-05: Construct the following matrix B2: Step S02-05-06: Calculate parameter C2: Step S02-05-07: Construct the following matrix A3: Step S02-05-08: Construct the following matrix B2: Step S02-05-09: Calculate parameter C3: Step S02-05-10: Construct the following matrix A4: Step S02-05-11: Construct the following matrix B4: Step S02-05-12: Calculate parameter C4: Step S02-05-13: Construct the following matrix A5: Step S02-05-14: Construct the following matrix B5: Step S02-05-15: Calculate parameter C5: Step S02-05-16: Construct the following matrix A6: Step S02-05-17: Construct the following matrix B6: Step S02-05-18: Calculate parameter C6: Step S02-06: Use the scipy.optimize.curve_fit function in Python's SciPy library to perform parameter fitting on the following formula; Among them, F i 、E i , D ij is the parameter to be fitted, i is an integer from 1 to N, j is an integer from 1 to 6, and N>5; Where X1, X2, X3, X4, X5, X6 correspond to parameters C1, C2, C3, C4, C5, C6, and Y corresponds to the number Z in step S02-02; Step S03: Automatically identify unknown pathogenic fungi cultured from clinical specimens. The specific process is as follows: Step S03-01: For the fungi cultured from the clinical specimens, stain them with lactic acid phenol cotton blue staining solution, observe them under an optical microscope, and use the F fitted in step S02 i 、E i , D ij The corresponding Y is calculated by using the parameters and formula (1), denoted as Y1; Step S03-02: rounding Y1 to the nearest integer, and recording the result as Y2; Step S03-03: Use Y2 as the number and retrieve the corresponding fungus from the fungal classification catalog table in the "Catalogue of Pathogenic Microorganisms Transmitted in Humans" (2023 edition).

Images