Prediction Method, Device and Server for Urine Culture Results

Abstract

The present invention provides a method, device and server for predicting urine culture results, including: obtaining a first sample image of a urine sample to be detected; based on the first sample image, identifying a first quantity of each type of microorganism contained in the urine sample to be detected; and through a pre-established urine culture prediction model, performing urine culture detection on the urine sample to be detected based on the first quantity to obtain the urine culture result of the urine sample to be detected. The present invention can significantly improve the detection efficiency of urine culture results.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular, to a method, device and server for predicting urine culture results. Background Art

[0002] Urinary tract infection (UTI) is an infectious disease caused by the growth and reproduction of pathogenic microorganisms in the urinary tract. Its incidence rate is second only to respiratory and digestive tract infections, covering almost all clinical departments. The gold standard for its diagnosis is "urine culture". At present, treating urinary tract infections based on the clinical experience of medical staff may cause problems such as overuse of antibiotics and an increased risk of bacterial drug resistance in urinary tract infections. Therefore, related technologies provide urine culture detection schemes and assist medical staff in treating urinary tract infections through these schemes. However, the traditional urine culture detection scheme takes a long time, and it takes 48 hours to issue a urine culture result report (negative or positive), that is, the urine culture detection efficiency is low, so it is very easy to cause delays in subsequent diagnoses. Summary of the Invention

[0003] In view of this, an object of the present invention is to provide a method, device and server for predicting urine culture results, which can significantly improve the detection efficiency of urine culture results.

[0004] In a first aspect, an embodiment of the present invention provides a method for predicting urine culture results, including: obtaining a first sample image of a urine sample to be detected; based on the first sample image, identifying a first quantity of each type of microorganism contained in the urine sample to be detected; and performing urine culture detection on the urine sample to be detected based on the first quantity through a pre-established urine culture prediction model to obtain the urine culture result of the urine sample to be detected.

[0005] In one implementation, the urine culture prediction model includes multiple urine culture prediction algorithms; the step of performing urine culture detection on the urine sample to be detected based on the first quantity through the pre-trained urine culture prediction model to obtain the urine culture result of the urine sample to be detected includes: for each urine culture prediction algorithm, determining a target microorganism corresponding to the urine culture prediction algorithm from the microorganisms, and judging whether bacteria and / or fungi exist in the urine sample to be detected based on the first quantity of the target microorganism; if the judgment result of any one of the urine culture prediction algorithms is yes, determining that the urine culture result of the urine sample to be detected is positive.

[0006] In one embodiment, the urine culture prediction model includes a first urine culture prediction algorithm; the step of determining, from the microorganisms, the target microorganism corresponding to the urine culture prediction algorithm and determining whether the urine sample to be tested contains bacteria and / or fungi based on the first quantity of the target microorganism includes: determining a first target microorganism from the microorganisms; wherein the first target microorganism includes yeast and / or Candida; if the first quantity of the first target microorganism is greater than a first threshold, determining that the urine sample to be tested contains fungi.

[0007] In one embodiment, the urine culture prediction model includes a second urine culture prediction algorithm; the step of determining, from the microorganisms, the target microorganism corresponding to the urine culture prediction algorithm and determining whether the urine sample to be tested contains bacteria and / or fungi based on the first quantity of the target microorganism includes: determining a second target microorganism and a third target microorganism from the microorganisms; wherein the second target microorganism includes suspected cocci and the third target microorganism includes bacilli; calculating a first product between the first quantity of the third target microorganism and a first parameter; calculating a first difference between the first quantity of the second target microorganism and the first product; if the first difference is less than or equal to a second threshold, determining that the urine sample to be tested contains bacteria.

[0008] In one embodiment, the urine culture prediction model includes a third urine culture prediction algorithm; the step of determining, from the microorganisms, the target microorganism corresponding to the urine culture prediction algorithm and determining whether the urine sample to be tested contains bacteria and / or fungi based on the first quantity of the target microorganism includes: determining a second target microorganism and a third target microorganism from the microorganisms; calculating a second product between the first quantity of the third target microorganism and a second parameter; calculating a sum value among the first quantity of the second target microorganism, the second product, and a third parameter; if the sum value is less than or equal to a third threshold, determining that the urine sample to be tested contains bacteria.

[0009] In one embodiment, the urine culture prediction model includes a fourth urine culture prediction algorithm; the step of determining, from the microorganisms, the target microorganism corresponding to the urine culture prediction algorithm and determining whether the urine sample to be tested contains bacteria and / or fungi based on the first quantity of the target microorganism includes: determining a third target microorganism from the microorganisms; calculating a second difference between the first quantity of the third target microorganism and a fourth parameter; if the second difference is greater than a fourth threshold, determining that the urine sample to be tested contains bacteria.

[0010] In one embodiment, the steps for establishing the urine culture prediction model include: obtaining a second sample image of the training urine sample; based on the second sample image, identifying the second quantity of each type of microorganism contained in the training urine sample; for each of the urine culture prediction algorithms included in the urine culture prediction model, determining the urine culture algorithm to be adjusted and determining candidate parameter values based on the current parameter values of the parameters to be adjusted in the urine culture algorithm to be adjusted; based on the second quantity and the current parameter values, respectively determining the total number of the first samples for which the urine culture results output by each of the urine culture prediction algorithms are positive, and the number of the first samples for which the urine culture results output by the urine culture algorithm to be adjusted are positive; based on the second quantity and the candidate parameter values, respectively determining the total number of the second samples for which the urine culture results output by each of the urine culture prediction algorithms are positive, and the number of the second samples for which the urine culture results output by the urine culture algorithm to be adjusted are positive; based on the total number of the first samples, the number of the first samples, the total number of the second samples, and the number of the second samples, determining whether to use the candidate parameter values to replace the current parameter values to train the urine culture algorithm to be adjusted; wherein, the parameters to be adjusted include one or more of a first parameter, a second parameter, a third parameter, a fourth parameter, a first threshold, a second threshold, a third threshold, and a fourth threshold.

[0011] In one embodiment, the step of, based on the first sample image, identifying the first quantity of each type of bacteria contained in the urine sample to be detected includes: preprocessing the first sample image to determine at least one type of microorganism contained in the first sample image; wherein, the preprocessing includes segmentation processing, feature extraction processing, and microorganism identification processing; and counting the first quantity of each type of the microorganism.

[0012] In a second aspect, an embodiment of the present invention further provides a device for predicting urine culture results, including: a sample acquisition module, configured to acquire a first sample image of a urine sample to be detected; a quantity identification module, configured to, based on the first sample image, identify the first quantity of each type of microorganism contained in the urine sample to be detected; and a urine culture detection module, configured to perform urine culture detection on the urine sample to be detected based on the first quantity through a pre-established urine culture prediction model, to obtain the urine culture result of the urine sample to be detected.

[0013] In a third aspect, an embodiment of the present invention further provides a server, including a processor and a memory, where the memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement the method according to any one of the first aspect.

[0014] Fourthly, an embodiment of the present invention further provides a computer-readable storage medium storing computer-executable instructions, which, when called and executed by a processor, cause the processor to implement the method according to any one of the first aspect.

[0015] The urine culture result prediction method, device and server provided by the embodiments of the present invention first obtain a first sample image of a urine sample to be detected, and based on the first sample image, identify the first quantity of each type of microorganism contained in the urine sample to be detected. Then, through a pre-established urine culture prediction model, urine culture detection is performed on the urine sample to be detected based on the first quantity, and the urine culture result of the urine sample to be detected is obtained. After identifying the first quantity of each type of microorganism in the first sample image of the urine sample to be detected, the above method can efficiently and accurately perform urine culture detection on the urine sample to be detected based on the first quantity by using the urine culture prediction model, and not only can obtain a urine culture result with a relatively high accuracy, but also can significantly improve the detection efficiency of the urine culture result.

[0016] Other features and advantages of the present invention will be described in the following description, and some of them will be obvious from the description, or understood by implementing the present invention. The objectives and other advantages of the present invention are achieved and obtained by the structures specifically pointed out in the description, claims and drawings.

[0017] To make the above objectives, features and advantages of the present invention more obvious and understandable, the following specifically enumerates preferred embodiments and, in conjunction with the accompanying drawings, makes a detailed description as follows. Description of the Drawings

[0018] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 It is a schematic flowchart of a urine culture result prediction method provided by an embodiment of the present invention;

[0020] Figure 2 It is an architecture diagram of a flow-through microscopic imaging artificial intelligence recognition technology provided by an embodiment of the present invention;

[0021] Figure 3 It is a schematic diagram of another urine culture result prediction method provided by an embodiment of the present invention;

[0022] Figure 4Schematic flowchart of a method for constructing a basic framework algorithm provided by an embodiment of the present invention;

[0023] Figure 5 Schematic flowchart of a process for autonomous improvement of an algorithm provided by an embodiment of the present invention;

[0024] Figure 6 Schematic structural diagram of a prediction device for urine culture results provided by an embodiment of the present invention;

[0025] Figure 7 Schematic structural diagram of a server provided by an embodiment of the present invention. Detailed implementation manners

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0027] Currently, the traditional urine culture detection scheme has the problem of low detection efficiency. Based on this, the embodiments of the present invention provide a method, device, and server for predicting urine culture results, which can significantly improve the detection efficiency of urine culture results.

[0028] For the convenience of understanding this embodiment, first, a method for predicting urine culture results disclosed in the embodiments of the present invention will be introduced in detail. Refer to Figure 1 The schematic flowchart of a method for predicting urine culture results shown, and this method mainly includes the following steps S102 to step S106:

[0029] Step S102, obtain a first sample image of the urine sample to be detected. In one implementation manner, an image acquisition device can be used to acquire the first sample image of the urine sample to be detected. Specifically, after sampling the urine of a target object (such as a patient), the urine sample to be detected is obtained. The urine sample to be detected enters the flow cell, and at the same time, the injection pump pushes the laminar fluid into the flow cell, so that the sample enters the thin-layer structure as a single layer of cells under the wrapping of the laminar fluid, and then is imaged by high-speed shooting, and the corresponding first sample image can be acquired.

[0030] Step S104: Based on the first sample image, identify the first quantity of each type of microorganism contained in the urine sample to be detected. In one implementation, the first sample image can be segmented to obtain multiple sample sub-images, feature extraction can be performed on each sample sub-image, the microorganisms contained in the urine sample to be detected can be classified into fungi and bacteria, and the bacteria can be further classified into bacilli and cocci. Then, the first quantity of each type of microorganism in all sample sub-images can be counted respectively.

[0031] Step S106: Through a pre-established urine culture prediction model, perform urine culture detection on the urine sample to be detected based on the first quantity, and obtain the urine culture result of the urine sample to be detected. Among them, the urine culture prediction model includes multiple urine culture prediction algorithms, and the urine culture result includes pass (i.e., positive) and fail (i.e., negative). In one implementation, one or more urine culture prediction algorithms can be used to detect the urine culture result of the urine sample to be detected based on the first quantity. When any urine culture prediction algorithm outputs positive, it is determined that the urine culture result of the urine sample to be detected is positive.

[0032] The prediction method for urine culture results provided by the embodiments of the present invention can, after identifying the first quantity of each type of microorganism in the first sample image of the urine sample to be detected, use the urine culture prediction model to efficiently and accurately perform urine culture detection on the urine sample to be detected based on the first quantity. It can not only obtain a urine culture result with a relatively high accuracy, but also significantly improve the detection efficiency of the urine culture result.

[0033] For the aforementioned step S104, the embodiments of the present invention provide an implementation for identifying the first quantity of each type of microorganism contained in the urine sample to be detected based on the first sample image. The first sample image can be preprocessed to determine at least one type of microorganism contained in the first sample image, and the first quantity of each type of microorganism can be counted. Among them, the preprocessing includes segmentation processing, feature extraction processing, and bacteria identification processing. Optionally, a FUS-360 urine formed element analyzer using flow-through microscopic imaging artificial intelligence recognition technology can be used. This formed element analyzer can complete the detection of urine formed elements within 5 minutes and provide the classification and counting of pathogens such as yeast, candida, bacilli, and cocci.

[0034] In one implementation, for the detection of microorganism counting in urine formed element detection, the following methods can be adopted: Method 1: Based on the principle of flow cytometry fluorescence staining; Method 2: Based on the principle of urine sediment image recognition; Method 3: Based on the principle of flow-through microscopic imaging artificial intelligence recognition. Preferably, the embodiments of the present invention use the principle of flow-through microscopic imaging artificial intelligence recognition to determine at least one type of microorganism and its first quantity contained in the first sample image.

[0035] For better understanding, the principle of artificial intelligence recognition in flow cytometry microscopy is introduced as follows: With the accelerating computer processing speed and the development of artificial intelligence image recognition technology, the artificial intelligence recognition technology in flow cytometry microscopy has been increasingly improved. Specifically: (1) Taking the first sample image: After sampling by the analysis system, the sample enters the flow cell, and at the same time, the injection pump pushes the laminar flow liquid into the flow cell, enabling the sample to enter the thin-layer structure as a single layer of cells under the wrapping of the laminar flow liquid, and then being captured and imaged at high speed. The urine formed elements are classified through artificial intelligence image recognition. (2) Segmenting the first sample image: Refer to Figure 2 the architecture diagram of an artificial intelligence recognition technology in flow cytometry microscopy shown in

[0036] The embodiment of the present invention is based on the principle of artificial intelligence recognition in flow cytometry microscopy. First, compared with flow cytometry fluorescence staining, this technology has the following advantages: 1) Since no fluorescence staining solution is required, the detection cost is reduced; 2) It can provide unique bacterial classification information for cocci and bacilli. Secondly, compared with urine sediment image recognition, this technology has the following advantages: 1) Using flow cytometry technology, since no centrifugation is required to obtain urine sediment, thick urine samples or highly turbid urine samples can be directly detected on the machine without dilution; 2) It can provide unique bacterial classification information for cocci and bacilli.

[0037] Optionally, the FUS-360 urine formed element analyzer can be used. The detection principle of this instrument is based on the artificial intelligence recognition technology in flow cytometry microscopy. The FUS-360 urine formed element analyzer shows good performance in detecting the particle count and classification of urine formed elements, and provides the pathogen classification information of bacilli and suspected cocci for the first time. Specifically, the FUS-360 urine formed element analyzer uses a 40X planachromat objective lens, with an increased magnification, clearer images; the limit resolution is reduced to 0.61μm, enabling more cell details to be seen clearly; the number of captured images is increased to 2,500, improving the detection rate of sample results; in addition, the FUS-360 urine formed element analyzer has also been improved in terms of the flow cell structure and liquid path.

[0038] Based on the foregoing embodiments, embodiments of the present invention are intended to prepare bacterial suspensions with different concentrations using standard strains of common pathogenic bacteria causing urinary tract infections, detect a first quantity using a FUS-360 urine formed element analyzer, and establish multiple urine culture prediction algorithms based on the reported results of bacilli and suspected cocci, combined with the morphological characteristics of different pathogenic bacteria and the principles and performance characteristics of the flow-through microscopic imaging artificial intelligence recognition technology, and construct a urine culture prediction model through each urine culture prediction algorithm.

[0039] For the foregoing step S106, embodiments of the present invention exemplarily provide an implementation manner of performing urine culture detection on a urine sample to be detected based on a first quantity using a pre-trained urine culture prediction model to obtain a urine culture result of the urine sample to be detected. Refer to the following steps 1 to 2:

[0040] Step 1, for each urine culture prediction algorithm, determine the target microorganism corresponding to the urine culture prediction algorithm from the microorganisms, and determine whether bacteria and / or fungi exist in the urine sample to be detected based on the first quantity of the target bacteria. Exemplarily, assume that the urine culture prediction model includes N urine culture prediction algorithms (i.e., algorithm 1, algorithm 2,..., algorithm n), determine the target microorganism corresponding to algorithm 1, the target microorganism corresponding to algorithm 2,..., and the target microorganism corresponding to algorithm n from the microorganisms respectively, and output corresponding judgment results using each algorithm.

[0041] Step 2, if the judgment result of any urine culture prediction algorithm is yes, determine that the urine culture result of the urine sample to be detected is positive. Exemplarily, assume that the judgment results output by algorithm 1 and algorithm n are that bacteria and / or fungi exist in the urine sample to be detected, then it can be determined that the urine culture result of the urine sample to be detected is positive.

[0042] In another implementation manner, each urine culture prediction algorithm can be sorted. When the urine culture result output by the current urine culture prediction algorithm is positive, directly output the urine culture result. When the urine culture result output by the current urine culture prediction algorithm is negative, output the corresponding target microorganism to the next urine culture prediction algorithm according to the foregoing sorting until the judgment result of the last urine culture prediction algorithm is obtained, or a positive urine culture detection result is obtained. For ease of understanding, refer to Figure 3 the schematic diagram of another method for predicting urine culture results shown: (1) Rapid detection: Determine the first quantity of each type of microorganism; (2) Result prediction: Perform urine culture detection using algorithm 1. If the judgment result of algorithm 1 is positive, directly output the result; if the output result of algorithm 1 is negative, perform urine culture detection using algorithm 2 until the result is output by algorithm n finally; (3) Data analysis: Taking the urine culture result as the gold standard, judge the consistency between the urine culture result output by the urine culture prediction model and the true urine culture result.

[0043] Currently, the standard strains of common pathogenic bacteria for urinary tract infections include the Escherichia coli standard strain ATCC25922, the Staphylococcus aureus standard strain ATCC25923, the Enterococcus faecalis standard strain ATCC29212; and the Proteus mirabilis standard strain ATCC3563.

[0044] In a specific implementation, the urine culture prediction model includes a first urine culture prediction algorithm, a second urine culture prediction algorithm, a third urine culture prediction algorithm, and a fourth urine culture prediction algorithm. For ease of understanding, in the embodiments of the present invention, for the above-mentioned urine culture prediction algorithms, specific implementation manners for determining the target microorganism corresponding to the urine culture prediction algorithm from microorganisms and judging whether there are bacteria and / or fungi in the urine sample to be detected based on the first quantity of the target microorganism are respectively provided, as shown in the following (a) to (d):

[0045] (a) For the first urine culture prediction algorithm, refer to the following steps a1 to a2:

[0046] Step a1, determine the first target microorganism from microorganisms. The first target microorganism includes yeast and / or Candida.

[0047] Step a2, if the first quantity of the first target microorganism is greater than the first threshold, determine that there are fungi in the urine sample to be detected. The first threshold can be set to "0". Exemplarily, if yeast and / or Candida > 0, it is considered that there are fungi in the sample, corresponding to a positive urine culture result for fungi, that is, output "predicted urine culture (fungi) positive". Among them, the parameters "0" and ">" can be changed.

[0048] Clinical sample verification: In the clinical verification of urine samples from 146 suspected UTI patients, 2 samples passed the first urine culture prediction algorithm, and their corresponding urine culture results were both positive for fungi, with a consistency rate of 100%, a false negative rate of 0, and a positive predictive value of 100%.

[0049] The basis for establishing the first urine culture prediction algorithm: After serial dilution of the original bacterial solution prepared by diluting the Candida albicans standard strain ATCC24433, it was detected using FUS-360. It was found that the sensitivity of this technology for fungal detection is relatively low, but the specificity is extremely high. That is to say, this technology reports the presence of yeast or Candida, and the probability of positive fungi in the sample is very high.

[0050] (b) For the second urine culture prediction algorithm, refer to the following steps b1 to b4:

[0051] Step b1, determine the second target microorganism and the third target microorganism from microorganisms. The second target microorganism includes suspected cocci, and the third target microorganism includes bacilli.

[0052] Step b2: Calculate the first product between the first quantity of the third target microorganism and the first parameter. Herein, the first parameter can be set to "2".

[0053] Step b3: Calculate the first difference between the first quantity of the second target microorganism and the first product.

[0054] Step b4: If the first difference is less than or equal to the second threshold, determine that the urine sample to be tested has bacteria. Herein, the second threshold can be set to "0". Exemplarily, calculate the number of suspected cocci - the number of bacilli * 2. If the calculated value ≤ 0, it is considered that the number of suspected cocci is caused by bacillus interference, and there are only bacilli in the sample, that is, output "predicted urine culture (bacillus) positive". Herein, the parameters "2", "0", and "≤" can be changed.

[0055] Clinical sample verification: Among the clinical verifications of the urine samples of the remaining 144 suspected UTI patients, 14 samples passed the second urine culture prediction algorithm, and their corresponding urine culture results were all bacilli, with a coincidence rate of 100%, a false negative rate of 0, and a positive predictive value of 100%.

[0056] Basis for establishing this algorithm: After serial dilution of the original bacterial solution prepared by diluting the standard strain of Escherichia coli ATCC25922, FUS-360 was used for detection. Theoretically, there are only bacilli in the serially diluted samples, and the actual coccus level is 0. However, this technique reported the presence of cocci, and there is a proportional relationship of about 2 times between the detected level of cocci and the bacillus level, and this phenomenon is repeatable. Combining the experimental results with the detection principle of this technique, in the process of graphic recognition, theoretically, there is a situation where the two ends of the bacillus are imaged as dots during the focused photography process, and the dot images with diameters meeting the requirements are judged as cocci. Based on this, the maximum noise signal value of cocci generated by bacilli is theoretically about 2 times the bacillus level.

[0057] (c) For the third urine culture prediction algorithm, refer to the following steps c1 to c4:

[0058] Step c1: Determine the second target microorganism and the third target microorganism from the microorganisms, that is, determine the suspected cocci and bacilli.

[0059] Step c2: Calculate the second product between the first quantity of the third target microorganism and the second parameter. Herein, the second parameter can be set to "70.5".

[0060] Step c3: Calculate the sum value among the first quantity of the second target microorganism, the second product, and the third parameter. Herein, the third parameter can be set to "3000".

[0061] Step c4, if the sum value is less than or equal to the third threshold, it is determined that the urine sample to be detected has bacteria. Among them, the third threshold can be set to "0". Exemplarily, calculate the number of suspected cocci - the number of bacilli * 70.5 + 3000. If the calculated value ≤ 0, it is considered that after excluding the interference of streptococcus on the detection, there are still bacilli or streptococcus in the sample, that is, output "predicted urine culture (bacilli or streptococcus) positive". Among them, the parameters "70.5", "3000", "0", and "≤" can be changed.

[0062] Clinical sample verification: Among the remaining 130 samples, 30 samples passed the screening of the third urine culture prediction algorithm. The corresponding urine culture results of 28 samples were positive, and the culture results were all bacilli or streptococcus; 2 samples were negative (the first case was a pleurisy patient undergoing anti-infection treatment with moxifloxacin hydrochloride and meropenem; the second case was a patient with recurrent urinary tract infection undergoing anti-infection treatment with levofloxacin. It should be noted that the use of antibiotics can cause the urine culture results of suspected UTI patients to show negative). Generally speaking, the coincidence rate between the third urine culture prediction algorithm and the urine culture results is 93.3%.

[0063] Basis for establishing this algorithm: After serial dilution of the original bacterial solution prepared by diluting the standard strain of Enterococcus faecalis (a common streptococcus in UTI) ATCC29212, FUS-360 was used for detection, and the values of bacilli and cocci were obtained simultaneously. Theoretically, there are only cocci in the serially diluted samples, and the actual bacillus level is 0. However, this technology reported the presence of bacilli, and the bacillus concentration showed a linear relationship with the coccus concentration, and this phenomenon was repeatable. Therefore, the algorithm formula of the number of suspected cocci - the number of bacilli * 70.5 + 3000 was established. Combining the experimental results with the principle of image recognition of this technology, theoretically, streptococcus forms dots during the focused photography process. Due to the characteristics of the chain arrangement meeting the judgment criteria for the length of bacilli, there is a situation where streptococcus is misjudged as bacilli.

[0064] (d) For the fourth urine culture prediction algorithm, refer to the following steps d1 to d3:

[0065] Step d1, determine the third target microorganism from the microorganisms, that is, determine bacilli.

[0066] Step d2, calculate the second difference between the first quantity of the third target bacteria and the fourth parameter. Among them, the fourth parameter can be set to "70".

[0067] Step d3, if the second difference is greater than the fourth threshold, it is determined that there are bacteria in the urine sample to be detected. Among them, the fourth threshold can be set to "0". Exemplarily, calculate the number of bacilli - 70. If the calculated value > 0, it is considered that after excluding the interference of cocci on the detection of bacilli, there are still bacilli in the sample, that is, output "predicted urine culture (bacilli) positive". Among them, the parameters "70", "0", and ">" can be changed.

[0068] Clinical sample verification: Among the remaining 100 samples, 2 samples passed the screening of the fourth urine culture prediction algorithm, and the corresponding urine culture results were 2 positives, and the urine culture results were all bacilli.

[0069] Basis for establishing this algorithm: After serial dilution of the original bacterial solution prepared by diluting the standard strain of Staphylococcus aureus ATCC25923, it was detected using FUS-360, and the values of bacilli and cocci were obtained simultaneously. Theoretically, only cocci are in the serially diluted sample, and the actual bacillus level is 0. However, this technique reported the presence of bacilli. The concentration of bacilli detected in the sample showed a certain correlation with the concentration of cocci, but the values were always at a relatively low level (40 - 100 / μl). Re-prepare low-concentration and higher-concentration coccus samples and continue to detect them using this technique. It is found that the bacillus noise signal values generated by cocci are concentrated at the 40 - 70 / μl level. Combining the experimental results with the image recognition principle of this technique, when cocci are not dispersed and adhere in a rod-shaped or chain-shaped form, they will be misjudged as bacilli. Based on this, the bacillus noise signal values generated by suspected cocci are concentrated at the 40 - 70 / μl level.

[0070] In practical applications, on the basis of using standard strains for basic experiments to construct the above algorithm framework, clinical samples are collected to accumulate data information, a clinical database is established and continuously expanded to achieve the self-improvement of the algorithm. Specifically, the embodiment of the present invention also provides an implementation manner for establishing a urine culture prediction model, as follows (i) to (vi):

[0071] (i) Obtain the second sample image of the training urine sample.

[0072] (ii) Based on the second sample image, identify the second quantity of each type of microorganism contained in the training urine sample. For the specific implementation manner of identifying the first quantity, reference can be made to the foregoing content, and it will not be elaborated in the embodiment of the present invention.

[0073] (3) For each urine culture prediction algorithm included in the urine culture prediction model, determine the urine culture algorithm to be adjusted, and determine the candidate parameter values based on the current parameter values of the parameters to be adjusted in the urine culture algorithm to be adjusted. Exemplarily, assuming that the first urine culture algorithm is used as the urine culture algorithm to be adjusted, the aforementioned first threshold is the parameter to be adjusted, and the current parameter value is "0". An increment and / or a decrement can be preset, and the current parameter value can be adjusted based on the increment or the decrement to obtain the candidate parameter values.

[0074] (4) Based on the second quantity and the current parameter values, respectively determine the total number of the first samples with positive urine culture results output by each urine culture prediction algorithm, and the number of the first samples with positive urine culture results output by the urine culture algorithm to be adjusted. Among them, for the process of each urine culture prediction algorithm outputting the urine culture results, reference can be made to the foregoing embodiments, and the embodiments of the present invention will not be described again here. Exemplarily, the first urine culture prediction algorithm keeps the current parameter value unchanged. On this basis, assuming that the number of samples with positive urine culture results output by the first urine culture prediction algorithm is x1, the number of samples with positive urine culture results output by the second urine culture prediction algorithm is x2, the number of samples with positive urine culture results output by the third urine culture prediction algorithm is x3, and the number of samples with positive urine culture results output by the fourth urine culture prediction algorithm is x4, then the total number of the first samples is x1 + x2 + x3 + x4, and the number of the first samples is x1.

[0075] (5) Based on the second quantity and the candidate parameter values, respectively determine the total number of the second samples with positive urine culture results output by each urine culture prediction algorithm, and the number of the second samples with positive urine culture results output by the urine culture algorithm to be adjusted. Exemplarily, the first urine culture prediction algorithm is changed to the candidate parameter value. On this basis, assuming that the number of samples with positive urine culture results output by the first urine culture prediction algorithm is y1, the number of samples with positive urine culture results output by the second urine culture prediction algorithm is x2, the number of samples with positive urine culture results output by the third urine culture prediction algorithm is x3, and the number of samples with positive urine culture results output by the fourth urine culture prediction algorithm is x4, then the total number of the second samples is y1 + x2 + x3 + x4, and the number of the second samples is y1.

[0076] (6) Based on the total number of the first samples, the number of the first samples, the total number of the second samples, and the number of the second samples, determine whether to use the candidate parameter value to replace the current parameter value for training the urine culture prediction algorithm to be adjusted. For example, when the urine culture result corresponding to the candidate parameter value is better than the urine culture result corresponding to the current parameter value, it can be determined to use the candidate parameter value to replace the current parameter value. Exemplarily, when the total number of the second samples is greater than the total number of the first samples and the number of the second samples is greater than the number of the first samples, it is determined that the urine culture result corresponding to the candidate parameter value is better than the urine culture result corresponding to the current parameter value.

[0077] In specific implementation, the parameter to be adjusted includes one or more of a first parameter, a second parameter, a third parameter, a fourth parameter, a first threshold, a second threshold, a third threshold, and a fourth threshold.

[0078] For ease of understanding, an application example of establishing a urine culture prediction model is further provided in an embodiment of the present invention. This embodiment is formulated for a urine formed element analyzer based on a flow-through microscopic imaging artificial intelligence recognition principle technology platform:

[0079] This prediction model is implemented through computer software (or can be incorporated into a hospital information system), and depends on a urine analyzer based on the principle of flow-through microscopic imaging artificial intelligence recognition technology. In other words, it is not applicable to urine analyzers based on the principle of flow cytometry fluorescence staining or urine sediment image recognition principle. Specifically:

[0080] First, construct the basic framework algorithm. First, collect the numbers of yeasts, candida, bacilli, and cocci (suspected cocci) in the urine samples of patients detected by this technology through a medical information system; then predict the corresponding urine culture results after 48 hours through a series of algorithms (such as the first urine culture prediction algorithm, the second urine culture prediction algorithm, the third urine culture prediction algorithm, the fourth urine culture prediction algorithm, etc.); finally, prompt the prediction results to the clinic through the information system in a timely manner, and provide the probability data of the prediction model (the reliability of the prediction basis) and the patient's medication situation for clinical judgment. Exemplarily, refer to Figure 4 the flow schematic diagram of a construction method of a basic framework algorithm shown below, and refer to the following steps S402 to step S410:

[0081] Step S402, configure the original bacterial solution with standard strains of common pathogenic bacteria for UTI. Among them, the standard strains of common pathogenic bacteria for UTI include ATCC25922 Escherichia coli standard strain, ATCC25923 Staphylococcus aureus standard strain, ATCC29212 Enterococcus faecalis standard strain, and ATCC3563 Proteus mirabilis standard strain.

[0082] Step S404, perform a series of dilutions on each original bacterial solution, and finally dilute each into 7 concentration samples.

[0083] Step S406: Detect the type of microorganism using the FUS-360 urine formed element analyzer. If the test result is cocci and / or bacilli, execute Step S408; if the test result is yeast or candida, execute Step S410.

[0084] Step S408: Construct a bacterial basic algorithm (including the second urine culture prediction algorithm, the third urine culture prediction algorithm, and the fourth urine culture prediction algorithm).

[0085] Step S410: Construct a fungal basic algorithm (including the first urine culture prediction algorithm).

[0086] Second, algorithm self-improvement. When specifically implemented, refer to Figure 5 the schematic flowchart of an algorithm self-improvement shown, including the following steps S502 to S514:

[0087] Step S502: The patient collects a clean midstream urine sample. Based on the patient's unique identification number and the urine specimen collection time, screen and obtain information on yeast, candida, bacilli, and cocci (suspected cocci) in the patient's urine formed element detection in the medical information system, as well as the real urine culture result after 48 hours (if the urine culture result is positive, the bacterial name information needs to be collected), and at the same time collect the patient's medication information through the electronic medical record system; establish and continuously expand the database.

[0088] Step S504: Detect the type of microorganism using the FUS-360 urine formed element analyzer.

[0089] Step S506: Obtain the predicted urine culture result using the urine culture prediction algorithm and obtain the clinical medication information.

[0090] Step S508: The traditional urine culture protocol determines the real urine culture result.

[0091] Step S510: Determine whether the predicted urine culture result is consistent with the real urine culture result. If so, execute Step S512; if not, execute Step S514.

[0092] Step S512: The consistency is good, and the existing algorithm is maintained.

[0093] Step S514: The consistency is poor, and the algorithm is automatically updated. Specifically, determine whether it is related to the patient's antibiotic use; if not, according to the real urine culture result, combined with the morphological characteristics of the microorganism and the imaging performance characteristics of this technology, the system automatically adjusts or supplements new algorithms, and after being verified by the database, they are incorporated into the prediction model.

[0094] Third, the fungal basic algorithm is automatically adjusted.

[0095] Build a software system to autonomously adjust the fungal prediction algorithm. Compared with the basic algorithm, the system automatically adjusts for different values in the fungal algorithm (such as the judgment threshold "0" in the first urine culture budget algorithm), and different diagnostic fourfold tables can be established, as shown in Table 1 below:

[0096] Table 1

[0097]

[0098] Different fungal prediction algorithms correspond to different sample passing numbers (a + b). First, it is necessary to ensure that the urine fungal culture results corresponding to the samples passing through the fungal algorithm are positive, that is, (a / (a + b)) is the largest, and at the same time, as many samples as possible pass through (a + b) which is larger.

[0099] If the adjusted algorithm is better than the original algorithm, the algorithm will be autonomously updated; otherwise, the original algorithm remains unchanged.

[0100] Fourth, autonomously adjust the bacterial basic algorithm. Taking the second urine culture prediction algorithm as an example, the algorithm established based on the basic experiment is "the number of suspected cocci - the number of bacilli * 2. If the calculated value ≤ 0, then output "predicted urine culture (bacilli) positive". Then the autonomous learning and adjustment process of the algorithm is as follows:

[0101] Compared with the basic algorithm, the system automatically adjusts for the coefficients in the single - bacterium prediction algorithm (such as the coefficient "2" in the second urine culture prediction algorithm, the coefficient "70.5" in the third urine culture prediction algorithm, and the judgment threshold "70" in the fourth urine culture prediction algorithm). Different values can establish different diagnostic fourfold tables, as shown in Table 2 below:

[0102] Table 2

[0103]

[0104] The best algorithm adjustment should satisfy a higher positive predictive value (that is, the percentage of true positives among the positive results predicted by the algorithm), that is, a / (a + b) is the largest; at the same time, as many samples as possible pass through the algorithm, that is, (a + b) / n is higher.

[0105] If the adjusted algorithm is better than the original algorithm, the algorithm will be autonomously updated; otherwise, the original algorithm remains unchanged.

[0106] It should be noted that in a single algorithm adjustment, the positive predictive value was focused on. However, the samples that did not pass the algorithm after a series of algorithm screenings also correspondingly had a relatively high negative predictive value (that is, the probability of true negatives in the negative results that did not pass the series of algorithms). In addition, in the algorithms of this solution, "yeast", "Candida", "coccus", "suspected coccus", and "bacillus" may have other meanings, different expressions with the same referent object, or English names and abbreviations, and equal patent protection should be given.

[0107] Furthermore, the urine formed element analyzer of the embodiment of the present invention is based on the principle of artificial intelligence recognition of flow-through microscopy imaging. The specific algorithm in the embodiment of the present invention is established on the basis of the results detected by the FUS-360 instrument. However, the construction of the prediction model in the embodiment of the present invention does not depend on the FUS-360 instrument, but on the analysis of urine formed elements based on the principle of artificial intelligence recognition of flow-through microscopy imaging.

[0108] As time goes by, the detection technology will be updated and improved, but the detection principle remains unchanged. The technical defects based on this detection principle will always exist. As long as the quantity or classification information of urine pathogens such as fungi, bacilli, and cocci is obtained based on this detection principle, the numbers in the algorithm will change, but the ideas and frameworks for establishing the algorithm will not change. Therefore, it is intended to apply for patent technical protection for predicting urine culture results based on the analysis of urine formed elements using the principle of artificial intelligence recognition of flow-through microscopy imaging.

[0109] Among patients suspected of having UTI, the negative rate of urine culture is as high as over 60%. In the embodiment of the present invention, 146 urine samples from patients suspected of having UTI were detected using a urine formed element analyzer based on the principle of artificial intelligence recognition of flow-through microscopy imaging, and urine culture was carried out simultaneously. Among them, 61 patients had positive urine culture results (42%), and 85 had negative results (58%). Among the 61 positive samples, 2 were positive for fungi, and the remaining 59 were positive for bacteria. Using the urine culture result prediction model constructed in the embodiment of the present invention, 48 positive urine culture results were screened and predicted (the detection rate was 48 / 61 = 78.7%), among which 46 were consistent with the patients' urine culture results (the consistency rate was 46 / 48 = 95.8%); the sensitivity was 75.4%, the specificity was 97.6%, the positive predictive value was 95.8%, and the negative predictive value was 84.7%.

[0110] In the prediction model and computer system constructed by the algorithm of the embodiment of the present invention, with the increase in the number of clinical samples, the improvement of the database, and the automatic update of the algorithm, the prediction value of the analysis of urine formed elements based on the flow-through microscopy artificial intelligence recognition technology for urine culture results (that is, the correctness of the prediction results) will be further improved.

[0111] For the method for predicting urine culture results provided in the foregoing embodiments, the embodiments of the present invention provide a device for predicting urine culture results. Refer to Figure 6 the structural schematic diagram of a device for predicting urine culture results shown in

[0112] A sample acquisition module 602, configured to acquire a first sample image of a urine sample to be detected;

[0113] A quantity recognition module 604, configured to recognize a first quantity of each type of microorganism contained in the urine sample to be detected based on the first sample image;

[0114] A urine culture detection module 606, configured to perform urine culture detection on the urine sample to be detected based on the first quantity through a pre-established urine culture prediction model, and obtain the urine culture result of the urine sample to be detected.

[0115] For the device for predicting urine culture results provided by the embodiments of the present invention, after recognizing the first quantity of each type of bacteria in the first sample image of the urine sample to be detected, the urine culture prediction model can be used to efficiently and accurately perform urine culture detection on the urine sample to be detected based on the first quantity, not only can a urine culture result with a relatively high accuracy be obtained, but also the detection efficiency of the urine culture result can be significantly improved.

[0116] In one implementation, the urine culture prediction model includes multiple urine culture prediction algorithms; the urine culture detection module 606 is further configured to: for each urine culture prediction algorithm, determine a target microorganism corresponding to the urine culture prediction algorithm from the microorganisms, and determine whether there are bacteria and / or fungi in the urine sample to be detected based on the first quantity of the target microorganism; if the judgment result of any urine culture prediction algorithm is yes, determine that the urine culture result of the urine sample to be detected is positive.

[0117] In one implementation, the urine culture prediction model includes a first urine culture prediction algorithm; the urine culture detection module 606 is further configured to: determine a first target microorganism from the microorganisms; wherein, the first target microorganism includes yeast and / or Candida; if the first quantity of the first target microorganism is greater than a first threshold, determine that there are fungi in the urine sample to be detected.

[0118] In one implementation, the urine culture prediction model includes a second urine culture prediction algorithm; the urine culture detection module 606 is further configured to: determine a second target microorganism and a third target microorganism from the microorganisms; wherein, the second target microorganism includes suspected cocci, and the third target microorganism includes bacilli; calculate a first product between the first quantity of the third target microorganism and a first parameter; calculate a first difference between the first quantity of the second target microorganism and the first product; if the first difference is less than or equal to a second threshold, determine that there are bacteria in the urine sample to be detected.

[0119] In one embodiment, the urine culture prediction model includes a third urine culture prediction algorithm; the urine culture detection module 606 is further configured to: determine a second target microorganism and a third target microorganism from the microorganisms; calculate a second product between the first quantity of the third target microorganism and a second parameter; calculate a sum value among the first quantity of the second target microorganism, the second product, and a third parameter; and if the sum value is less than or equal to a third threshold, determine that the urine sample to be detected has bacteria.

[0120] In one embodiment, the urine culture prediction model includes a fourth urine culture prediction algorithm; the urine culture detection module 606 is further configured to: determine a third target microorganism from the microorganisms; calculate a second difference between the first quantity of the third target microorganism and a fourth parameter; and if the second difference is greater than a fourth threshold, determine that the urine sample to be detected has bacteria.

[0121] In one embodiment, the above device further includes a building module, configured to: obtain a second sample image of a training urine sample; based on the second sample image, identify a second quantity of each type of bacteria included in the training urine sample; for each urine culture prediction algorithm included in the urine culture prediction model, determine the urine culture algorithm to be adjusted as the to-be-adjusted urine culture prediction algorithm, and determine a candidate parameter value based on the current parameter value of the parameter to be adjusted in the to-be-adjusted urine culture prediction algorithm; based on the second quantity and the current parameter value, respectively determine a first total number of samples with a positive urine culture result output by each urine culture prediction algorithm, and a first number of samples with a positive urine culture result output by the to-be-adjusted urine culture prediction algorithm; based on the second quantity and the candidate parameter value, respectively determine a second total number of samples with a positive urine culture result output by each urine culture prediction algorithm, and a second number of samples with a positive urine culture result output by the to-be-adjusted urine culture prediction algorithm; and based on the first total number of samples, the first number of samples, the second total number of samples, and the second number of samples, determine whether to replace the current parameter value with the candidate parameter value to train the to-be-adjusted urine culture prediction algorithm; wherein, the parameter to be adjusted includes one or more of a first parameter, a second parameter, a third parameter, a fourth parameter, a first threshold, a second threshold, a third threshold, and a fourth threshold.

[0122] In one embodiment, the quantity recognition module 604 is further configured to: preprocess the first sample image to determine at least one type of microorganism included in the first sample image; wherein, the preprocessing includes segmentation processing, feature extraction processing, and microorganism recognition processing; and count the first quantity of each type of microorganism.

[0123] The device provided in the embodiments of the present invention has the same implementation principle and the same technical effects as those in the foregoing method embodiments. For the sake of brief description, for the parts not mentioned in the device embodiments, reference may be made to the corresponding contents in the foregoing method embodiments.

[0124] An embodiment of the present invention provides a server. Specifically, the server includes a processor and a storage device; a computer program is stored on the storage device, and when the computer program is run by the processor, it executes the method described in any one of the above-described embodiments.

[0125] Figure 7 FIG. 4 is a schematic structural diagram of a server provided by an embodiment of the present invention. The server 100 includes: a processor 70, a memory 71, a bus 72, and a communication interface 73. The processor 70, the communication interface 73, and the memory 71 are connected through the bus 72; the processor 70 is configured to execute an executable module stored in the memory 71, such as a computer program.

[0126] Among them, the memory 71 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory. Through at least one communication interface 73 (which may be wired or wireless), a communication connection between the system network element and at least one other network element is realized, and the Internet, a wide area network, a local area network, a metropolitan area network, etc. can be used.

[0127] The bus 72 may be an ISA bus, a PCI bus, an EISA bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For the sake of simplicity of representation, Figure 7 only a bidirectional arrow is used in FIG. 4, but it does not mean that there is only one bus or one type of bus.

[0128] Among them, the memory 71 is used to store a program. After receiving an execution instruction, the processor 70 executes the program. The method executed by the device defined by the flow process disclosed in any one of the foregoing embodiments of the present invention can be applied to the processor 70 or implemented by the processor 70.

[0129] The processor 70 may be an integrated circuit chip with the ability to process signals. In the implementation process, the steps of the above method can be completed by the integrated logic circuit of the hardware in the processor 70 or the instructions in the form of software. The above-mentioned processor 70 may be a general-purpose processor, including a central processing unit (CPU for short), a network processor (NP for short), etc.; it may also be a digital signal processor (DSP for short), an application specific integrated circuit (ASIC for short), a field-programmable gate array (FPGA for short), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present invention can be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of the hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. This storage medium is located in the memory 71, and the processor 70 reads the information in the memory 71 and combines its hardware to complete the steps of the above method.

[0130] The computer program product of the readable storage medium provided by the embodiments of the present invention includes a computer-readable storage medium storing program code, and the instructions included in the program code can be used to execute the methods described in the foregoing method embodiments. For specific implementation, reference can be made to the foregoing method embodiments, which will not be elaborated here.

[0131] If the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs.

[0132] Finally, it should be noted that the above-mentioned embodiments are only specific implementation manners of the present invention, used to illustrate the technical solutions of the present invention, rather than limiting it. The protection scope of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed by the present invention can still modify the technical solutions recorded in the foregoing embodiments, or can easily think of changes, or perform equivalent replacements for some of the technical features; and these modifications, changes, or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A method for predicting the result of urine culture, characterized in that, it includes: Obtaining a first sample image of the urine sample to be detected; Based on the first sample image, identifying the first quantity of each type of microorganism contained in the urine sample to be detected; Through a pre-established urine culture prediction model, performing urine culture detection on the urine sample to be detected based on the first quantity, and obtaining the urine culture result of the urine sample to be detected; The urine culture prediction model includes multiple urine culture prediction algorithms; The step of performing urine culture detection on the urine sample to be detected based on the first quantity through a pre-established urine culture prediction model and obtaining the urine culture result of the urine sample to be detected includes: For each of the urine culture prediction algorithms, determining the target microorganism corresponding to the urine culture prediction algorithm from the microorganisms, and judging whether bacteria and / or fungi exist in the urine sample to be detected based on the first quantity of the target microorganism; If the judgment result of any one of the urine culture prediction algorithms is yes, determining that the urine culture result of the urine sample to be detected is positive; The establishment step of the urine culture prediction model includes: Obtaining a second sample image of the training urine sample; Based on the second sample image, identifying the second quantity of each type of microorganism contained in the training urine sample; For each of the urine culture prediction algorithms included in the urine culture prediction model, determining the urine culture prediction algorithm to be adjusted as the urine culture algorithm, and determining the candidate parameter value based on the current parameter value of the parameter to be adjusted in the urine culture prediction algorithm to be adjusted; Based on the second quantity and the current parameter value, respectively determining the total number of the first samples with positive urine culture results output by each urine culture prediction algorithm, and the number of the first samples with positive urine culture results output by the urine culture prediction algorithm to be adjusted; Based on the second quantity and the candidate parameter value, respectively determining the total number of the second samples with positive urine culture results output by each urine culture prediction algorithm, and the number of the second samples with positive urine culture results output by the urine culture prediction algorithm to be adjusted; Based on the total number of the first samples, the number of the first samples, the total number of the second samples, and the number of the second samples, judging whether to use the candidate parameter value to replace the current parameter value to train the urine culture prediction algorithm to be adjusted; Wherein, the parameter to be adjusted includes one or more of a first parameter, a second parameter, a third parameter, a fourth parameter, a first threshold, a second threshold, a third threshold, and a fourth threshold.

2. The method according to claim 1, characterized in that, the urine culture prediction model includes a first urine culture prediction algorithm; The step of determining the target microorganism corresponding to the urine culture prediction algorithm from the microorganisms and judging whether bacteria and / or fungi exist in the urine sample to be detected based on the first quantity of the target microorganism includes: Determining a first target microorganism from the microorganisms; wherein, the first target microorganism includes yeast and / or Candida. If the first quantity of the first target microorganism is greater than the first threshold, it is determined that the urine sample to be detected contains fungi.

3. The method according to claim 1, wherein, the urine culture prediction model includes a second urine culture prediction algorithm; the step of determining the target microorganism corresponding to the urine culture prediction algorithm from the microorganisms and judging whether the urine sample to be detected contains bacteria and / or fungi based on the first quantity of the target microorganism includes: determining a second target microorganism and a third target microorganism from the microorganisms; wherein, the second target microorganism includes suspected cocci, and the third target microorganism includes bacilli; calculating a first product between the first quantity of the third target microorganism and a first parameter; calculating a first difference between the first quantity of the second target microorganism and the first product; If the first difference is less than or equal to a second threshold, it is determined that the urine sample to be detected contains bacteria.

4. The method according to claim 1, wherein, the urine culture prediction model includes a third urine culture prediction algorithm; the step of determining the target microorganism corresponding to the urine culture prediction algorithm from the microorganisms and judging whether the urine sample to be detected contains bacteria and / or fungi based on the first quantity of the target microorganism includes: determining a second target microorganism and a third target microorganism from the microorganisms; calculating a second product between the first quantity of the third target microorganism and a second parameter; calculating a sum value among the first quantity of the second target microorganism, the second product, and a third parameter; If the sum value is less than or equal to a third threshold, it is determined that the urine sample to be detected contains bacteria.

5. The method according to claim 1, wherein, the urine culture prediction model includes a fourth urine culture prediction algorithm; the step of determining the target microorganism corresponding to the urine culture prediction algorithm from the microorganisms and judging whether the urine sample to be detected contains bacteria and / or fungi based on the first quantity of the target microorganism includes: determining a third target microorganism from the microorganisms; calculating a second difference between the first quantity of the third target microorganism and a fourth parameter; If the second difference is greater than a fourth threshold, it is determined that the urine sample to be detected contains bacteria.

6. The method according to claim 1, wherein, the step of identifying the first quantity of each type of microorganism contained in the urine sample to be detected based on the first sample image includes: preprocessing the first sample image to determine at least one type of microorganism contained in the first sample image; wherein, the preprocessing includes segmentation processing, feature extraction processing, and microorganism recognition processing; counting the first quantity of each type of the microorganisms.

7. A prediction device for urine culture results, wherein, it includes: a sample acquisition module for acquiring a first sample image of a urine sample to be detected; a quantity recognition module for identifying the first quantity of each type of microorganism contained in the urine sample to be detected based on the first sample image; A urine culture detection module, configured to perform urine culture detection on the urine sample to be detected based on the first quantity through a pre-established urine culture prediction model, so as to obtain the urine culture result of the urine sample to be detected; The urine culture prediction model includes a plurality of urine culture prediction algorithms; The urine culture detection module is further configured to: for each of the urine culture prediction algorithms, determine the target microorganism corresponding to the urine culture prediction algorithm from the microorganisms, and determine whether bacteria and / or fungi exist in the urine sample to be detected based on the first quantity of the target microorganism; if the judgment result of any one of the urine culture prediction algorithms is yes, determine that the urine culture result of the urine sample to be detected is positive; When establishing the urine culture prediction model in the urine culture detection module, it is further configured to: obtain a second sample image of the training urine sample; based on the second sample image, identify the second quantity of each type of microorganism included in the training urine sample; for each of the urine culture prediction algorithms included in the urine culture prediction model, determine the urine culture algorithm to be adjusted, and determine a candidate parameter value based on the current parameter value of the parameter to be adjusted in the urine culture prediction algorithm to be adjusted; based on the second quantity and the current parameter value, respectively determine the total number of the first samples with a positive urine culture result output by each urine culture prediction algorithm, and the number of the first samples with a positive urine culture result output by the urine culture prediction algorithm to be adjusted; based on the second quantity and the candidate parameter value, respectively determine the total number of the second samples with a positive urine culture result output by each urine culture prediction algorithm, and the number of the second samples with a positive urine culture result output by the urine culture prediction algorithm to be adjusted; based on the total number of the first samples, the number of the first samples, the total number of the second samples and the number of the second samples, judge whether to use the candidate parameter value to replace the current parameter value to train the urine culture prediction algorithm to be adjusted; wherein, the parameter to be adjusted includes one or more of a first parameter, a second parameter, a third parameter, a fourth parameter, a first threshold, a second threshold, a third threshold and a fourth threshold.

8. A server, Characterized in that, It includes a processor and a memory, the memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, Characterized in that, The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are called and executed by a processor, the computer-executable instructions cause the processor to implement the method according to any one of claims 1 to 6.

Images