阅读说明：本技术 基于荧光免疫层析法的荧光曲线的检测方法与装置 (Fluorescence curve detection method and device based on fluorescence immunochromatography ) 是由 陈喆 陈秋强 王燕青 沈明程 颜楚楚 于 2021-08-16 设计创作，主要内容包括：本发明提供了一种基于荧光免疫层析法的荧光曲线的检测方法与装置,该方法涉及荧光数据检测的技术领域,该方法包括：获取初步检测结果,其中,初步检测结果包括荧光曲线以及荧光曲线的配套信息；根据预设的规则及数据库,对荧光曲线进行有效性判断,得到判断结果；根据判断结果确定标识信息,将标识信息与荧光曲线关联保存至数据库；其中,数据库用于训练用于进行荧光曲线有效性判断及标识信息的确定的模型。本发明能够通过实时更新的数据库与特定的检验算法,自动区分荧光曲线的有效性、自动对荧光曲线进行分析、标识,进而提高对荧光曲线的检测效率与检测精度。(The invention provides a fluorescence curve detection method and a fluorescence curve detection device based on a fluorescence immunochromatography method, and relates to the technical field of fluorescence data detection, wherein the method comprises the following steps: acquiring a preliminary detection result, wherein the preliminary detection result comprises a fluorescence curve and matched information of the fluorescence curve; according to preset rules and a database, carrying out validity judgment on the fluorescence curve to obtain a judgment result; determining identification information according to the judgment result, and storing the identification information and the fluorescence curve in a database in a correlation manner; the database is used for training a model for judging the effectiveness of the fluorescence curve and determining the identification information. The invention can automatically distinguish the effectiveness of the fluorescence curve and automatically analyze and mark the fluorescence curve through the real-time updated database and the specific detection algorithm, thereby improving the detection efficiency and the detection precision of the fluorescence curve.)

1. A method for detecting a fluorescence curve based on fluorescence immunochromatography, comprising:

acquiring a preliminary detection result, wherein the preliminary detection result comprises a fluorescence curve and matched information of the fluorescence curve;

according to preset rules and a database, carrying out validity judgment on the fluorescence curve to obtain a judgment result;

determining identification information according to the judgment result, and storing the identification information and the fluorescence curve in the database in a correlation manner;

the database is used for training a model for judging the effectiveness of the fluorescence curve and determining the identification information.

2. The method according to claim 1, wherein the step of determining identification information according to the determination result, and storing the identification information in the database in association with the fluorescence curve comprises:

and if the judgment result indicates that the fluorescence curve is valid data, taking the judgment result, the ID information and the remark information of the fluorescence curve as first identification information, and storing the first identification information and the fluorescence curve into the database in a correlation manner.

3. The method according to claim 2, wherein the step of determining identification information according to the determination result, storing the identification information in association with the fluorescence curve in the database, further comprises:

if the judgment result indicates that the fluorescence curve is invalid data, judging the invalid reason of the fluorescence curve, classifying and analyzing the invalid reason to obtain invalid reason classification and standard operation suggestions;

saving the reasons for inefficiency, the reasons for inefficiency classification, and the normative operating recommendations in association with the fluorescence curve to the database.

4. The method of claim 3, further comprising:

and acquiring a manual reinspection detection result of the preliminary detection result at a specified time, and if the manual reinspection detection result indicates that the preliminary detection result is incorrect, taking the manual reinspection detection result as an updated preliminary detection result.

5. The method of claim 4, further comprising:

acquiring a manual review judgment result of the judgment result at a specified time, and if the manual review judgment result indicates that the judgment result is incorrect, taking the manual review detection result as an updated judgment result;

and determining the identification information according to the updated judgment result, and storing the identification information and the fluorescence curve in the database in a correlation manner.

6. The method of claim 5, further comprising:

and if the judgment result cannot be obtained according to the preset rule and the database, obtaining a manual judgment result.

7. The method according to claim 1, wherein the step of determining the validity of the fluorescence curve according to a preset rule and a database to obtain a determination result comprises:

performing smoothing and normalization processing on the fluorescence curve to obtain a curve processing result;

performing peak discrimination on the curve processing result by a peak discrimination method to obtain a curve peak;

and carrying out validity check on the curve processing result and the curve wave crest by combining a heterogeneous characteristic check rule, a machine self-learning algorithm and the database to obtain the judgment result.

8. The method of claim 1, further comprising: and generating a report in a specified form according to the customer requirements through the database.

9. A fluorescence curve detection device based on fluorescence immunochromatography, comprising:

a data acquisition module to: acquiring a preliminary detection result, wherein the preliminary detection result comprises a fluorescence curve and matched information of the fluorescence curve;

a determination module configured to: according to preset rules and a database, carrying out validity judgment on the fluorescence curve to obtain a judgment result;

a database perfecting module for: determining identification information according to the judgment result, and storing the identification information and the fluorescence curve in the database in a correlation manner;

the database is used for training a model for judging the effectiveness of the fluorescence curve and determining the identification information.

10. The apparatus of claim 9, wherein the database perfecting module is further configured to: and if the judgment result indicates that the fluorescence curve is valid data, taking the judgment result, the ID information and the remark information of the fluorescence curve as first identification information, and storing the first identification information and the fluorescence curve into the database in a correlation manner.

Technical Field

The invention relates to the technical field of fluorescence data detection, in particular to a fluorescence curve detection method and device based on a fluorescence immunochromatography method.

Background

The fluorescence analysis technology based on the fluorescence immunochromatography method depends on the specific combination of an antibody or an antigen marked in an immunoreagent and a detected object, and takes a nitrocellulose membrane as a carrier to detect substances such as protein, small molecules and the like in the detected object. The determination of the detection result and the screening result usually requires analysis by means of an immunofluorescence curve.

Because the traditional fluorescence instrument can not judge the effectiveness of the immunofluorescence curve, and has no capability of analyzing data and identifying the data, users are difficult to obtain timely and correct detection results. Based on the above problems, how to extract, identify and interpret the curve characteristics of the fluorescence curve based on the fluorescence immunochromatography, thereby ensuring the reliability of the detection result, is a technical problem to be solved urgently.

Disclosure of Invention

The invention aims to provide a fluorescence curve detection method and a fluorescence curve detection device based on a fluorescence immunochromatography method, so as to improve the interpretation accuracy of an immunofluorescence chromatography detection result and improve the detection efficiency.

In a first aspect, the embodiments of the present invention provide a method for detecting a fluorescence curve based on fluorescence immunochromatography, the method including: acquiring a preliminary detection result, wherein the preliminary detection result comprises a fluorescence curve and matched information of the fluorescence curve; according to preset rules and a database, carrying out validity judgment on the fluorescence curve to obtain a judgment result; determining identification information according to the judgment result, and storing the identification information and the fluorescence curve in a database in a correlation manner; wherein the database is used for training a model for judging the effectiveness of the fluorescence curve and determining the identification information.

Further, the step of storing the identification information in association with the fluorescence curve in the database includes: and if the judgment result indicates that the fluorescence curve is valid data, taking the judgment result, the ID information of the fluorescence curve and the remark information as first identification information, and storing the first identification information and the fluorescence curve in a database in a correlation manner.

Further, the step of determining the identification information according to the judgment result and storing the identification information and the fluorescence curve in a database in a correlated manner further includes: if the judgment result indicates that the fluorescence curve is invalid data, judging the invalid reason of the fluorescence curve, classifying and analyzing the invalid reason to obtain invalid reason classification and standard operation suggestions; and storing the invalid reason, the invalid reason classification and the standard operation suggestion in a database in association with the fluorescence curve.

Further, the method further comprises: and acquiring a manual reinspection detection result of the preliminary detection result at a specified time, and taking the manual reinspection detection result as an updated preliminary detection result if the manual reinspection detection result indicates that the preliminary detection result is incorrect.

Further, the method further comprises: acquiring a manual rechecking judgment result of the judgment result at a specified time, and if the manual rechecking judgment result indicates that the judgment result is incorrect, taking the manual rechecking detection result as an updated judgment result; and determining identification information according to the updated judgment result, and storing the identification information and the fluorescence curve in a database in a correlation manner.

Further, the method further comprises: and if the judgment result cannot be obtained according to the preset rule and the database, obtaining a manual judgment result.

Further, the step of judging the validity of the fluorescence curve according to the preset rule and the database to obtain a judgment result includes: performing smoothing and normalization processing on the fluorescence curve to obtain a curve processing result; performing peak discrimination on the curve processing result by a peak discrimination method to obtain a curve peak; and (4) carrying out validity check on the curve processing result and the curve wave crest by combining a heterogeneous characteristic checking rule, a machine self-learning algorithm and a database to obtain a judgment result.

Further, the method further comprises: and generating a report in a specified form according to the requirements of customers through a database.

In a second aspect, the embodiments of the present invention further provide a fluorescence curve detection device based on fluorescence immunochromatography, the device including: a data acquisition module to: acquiring a preliminary detection result, wherein the preliminary detection result comprises a fluorescence curve and matched information of the fluorescence curve; a determination module configured to: according to preset rules and a database, carrying out validity judgment on the fluorescence curve to obtain a judgment result; a database perfecting module for: determining identification information according to the judgment result, and storing the identification information and the fluorescence curve in a database in a correlation manner; the database is used for training a model for judging the effectiveness of the fluorescence curve and determining the identification information.

Further, the database perfecting module is further configured to: and if the judgment result indicates that the fluorescence curve is valid data, taking the judgment result, the ID information of the fluorescence curve and the remark information as first identification information, and storing the first identification information and the fluorescence curve in a database in a correlation manner.

The embodiment of the invention has the following beneficial effects:

the invention provides a fluorescence curve detection method and a fluorescence curve detection device based on a fluorescence immunochromatography method, and relates to the technical field of fluorescence data detection, wherein the method comprises the following steps: acquiring a preliminary detection result, wherein the preliminary detection result comprises a fluorescence curve and matched information of the fluorescence curve; according to preset rules and a database, carrying out validity judgment on the fluorescence curve to obtain a judgment result; determining identification information according to the judgment result, and storing the identification information and the fluorescence curve in a database in a correlation manner; the database is used for training a model for judging the effectiveness of the fluorescence curve and determining the identification information. The invention can automatically distinguish the effectiveness of the fluorescence curve and automatically analyze the fluorescence curve through the real-time updated database and the specific detection algorithm, thereby improving the detection efficiency and the detection precision of the fluorescence curve.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by the practice of the invention as set forth above.

In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.

FIG. 1 is a flow chart of a fluorescence curve detection method based on fluorescence immunochromatography according to an embodiment of the present invention;

FIG. 2 is a schematic view of a fluorescence curve scanned by a test paper card according to an embodiment of the present invention;

FIG. 3 is a flow chart of another fluorescence curve detection method based on fluorescence immunochromatography according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of a fluorescence curve detection device based on fluorescence immunochromatography according to an embodiment of the present invention;

FIG. 5 is a flowchart of a method for analyzing fluorescence detection results based on fluorescence immunochromatography and self-constructed database according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.

Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The traditional fluorescence instrument basically has no capability of distinguishing valid data from invalid data and no capability of analyzing causes of the invalid data, so that the interpretation result can be accurate only under the condition that the test data are valid. Invalid curve data can distort an instrument interpretation result, so that a user cannot obtain timely and correct feedback, and the use experience of the user is seriously influenced. The actual detection condition cannot be clearly reflected by the currently uploaded detection data result. The invention carries out automatic AI detection on the fluorescence curve through the process of 'cleaning' data, namely curve characteristic extraction, identification and interpretation, and the technology can obviously improve the after-sale service efficiency and greatly save the after-sale cost.

Example one

To facilitate understanding of this embodiment, a detailed description will be given of a fluorescence curve detection method based on fluorescence immunochromatography disclosed in this embodiment, as shown in fig. 1.

Step S102, obtaining a preliminary detection result, wherein the preliminary detection result comprises a fluorescence curve and matching information of the fluorescence curve.

In a specific implementation, the preliminary detection result is obtained manually or identified automatically through the database AI. The more data in the database, the higher the accuracy of the primary detection result, and the database improves the AI curve interpretation and identification precision and the precision and efficiency of the detection result output through huge data volume. The preliminary detection result is downloaded and derived from the cloud platform or obtained through an API (Application Programming Interface) Interface or the like. The preliminary detection result comprises fluorescence curve data, detection time, detection machine model, detected person information and other fluorescence curve matching information, and the invention mainly detects the validity and error (invalidation) reasons of the fluorescence curve, wherein fig. 2 is a schematic diagram of the fluorescence curve scanned by the test paper card, the abscissa is the fluorescence value of the fluorescence point, and the unit is one; the ordinate reflects the relative magnitude of the fluorescence value, and the numerical value of the ordinate is automatically read by the fluorescence machine. Wherein, the information matched with the fluorescence curve includes but not limited to: test items (biomarkers such as antigens, antibodies, proteins and small molecules), test samples (biological samples such as blood, urine, saliva, hair and feces), test time, serial numbers of instruments generating the test data, places of putting the instruments, reagent batch numbers, negative and positive results and the like.

And step S104, judging the effectiveness of the fluorescence curve according to a preset rule and a database to obtain a judgment result.

In specific implementation, the preset rules include N heterogeneous inspection methods, an euclidean distance matching method, a convolutional neural network algorithm, a KNN (k-Nearest Neighbors) algorithm, a BP (back propagation) neural network algorithm, an SVM (support vector machines) algorithm, and the like, and the rules and the algorithms may be updated or changed in real time according to user requirements. The database is updated in real time, and the accuracy of the judgment result is improved through a large amount of resources in the database.

Step S106, determining identification information according to the judgment result, and storing the identification information and the fluorescence curve in a database in a correlation manner; wherein the database is used for training a model for judging the effectiveness of the fluorescence curve and determining the identification information.

In specific implementation, the judgment result comprises two types of 'effective' and 'ineffective', if the judgment result is 'effective', the 'effective' is used as identification information, and then the identification information is associated with the fluorescence curve and is stored in a database. And if the judgment result is 'invalid', analyzing the fluorescence curve through a database and a preset rule, taking the analysis result as identification information, associating the identification information with the fluorescence curve, and storing the identification information in the database. The database is used for training a model for judging the effectiveness of the fluorescence curve and determining the identification information, and helps to improve the AI curve interpretation and identification precision and improve the precision and efficiency of detection result output by virtue of the advantages of data volume. And the data in the database is used as a machine learning sample set, and the AI identification precision is improved by continuously increasing the data volume.

The embodiment of the invention provides a fluorescence curve detection method based on a fluorescence immunochromatography method, which comprises the following steps: acquiring a preliminary detection result, wherein the preliminary detection result comprises a fluorescence curve and matched information of the fluorescence curve; according to preset rules and a database, carrying out validity judgment on the fluorescence curve to obtain a judgment result; determining identification information according to the judgment result, and storing the identification information and the fluorescence curve in a database in a correlation manner; the database is used for training a model for judging the effectiveness of the fluorescence curve and determining the identification information. The invention can automatically distinguish the effectiveness of the fluorescence curve and automatically analyze the fluorescence curve through the real-time updated database and the specific detection algorithm, thereby improving the detection efficiency and the detection precision of the fluorescence curve.

Example two

For the convenience of understanding this example, another fluorescence curve detection method based on fluorescence immunochromatography disclosed in this example will be described in detail, as shown in FIG. 3.

Step S302, obtaining a preliminary detection result, wherein the preliminary detection result comprises a fluorescence curve and matching information of the fluorescence curve.

And step S304, according to preset rules and a database, carrying out validity judgment on the fluorescence curve to obtain a judgment result.

During specific implementation, firstly, smoothing and normalizing a fluorescence curve to obtain a curve processing result; then, performing peak discrimination on the curve processing result by a peak discrimination method to obtain a curve peak; and finally, AI automatic processing, namely validity inspection, is carried out on the curve processing result and the curve wave crest through a heterogeneous characteristic inspection rule, a machine self-learning algorithm and a database, so as to obtain a judgment result.

Specifically, the specific judgment index of the above heterogeneous characteristic check rule includes the following contents:

firstly, a 1 × N discrimination matrix reason is set, each number is 1 or 0, and the number respectively represents whether the N heterogeneous features pass or not.

1) The fluorescence curve is checked for undersize. In the curve fluorescence value raw data, if the value less than a certain critical value reaches a certain proportion, the item does not pass. Typically, when the fluorescence value of a point above [ 50-100% ] is less than a predetermined threshold value X (X is typically within the interval [0,200000 ]), this term does not pass.

2) And (5) checking whether the peak has a valid characteristic peak. There is at least one valid peak in the preset area of the T line or the C line, otherwise, the item does not pass.

3) It is checked whether the wrong test item was selected. And (4) judging according to the item number and curve characteristics in the input original data. For example, if a single test is mistakenly selected as a double test item, the item will not pass if no item is selected or an empty item is selected.

4) And (5) checking whether a quality control line exists. And (4) judging according to the input test record and curve characteristics. If the "C value" in the test record is less than a certain threshold and no peak is identified in the curve feature, then the term fails.

5) And (5) checking whether the non-peak area at the left end of the curve is tilted. If a bulge or lift occurs, the item does not pass.

6) It is checked whether the value between the left and right peaks of the curve is too high. If the trough is too high then the item does not pass.

7) And (5) checking whether the non-peak area at the right end of the curve is tilted. If a bulge or lift occurs, the item does not pass.

8) And (5) checking whether the right end of the curve is suddenly dropped. If so, the entry does not pass.

9) And (4) checking whether a section of low wave appears at the left end of the curve, and if so, not passing the term.

10) It is checked whether another peak or bulge appears in the closer distance of the curve peak. Typically another peak occurs within a certain peak distance and the term does not pass.

11) The left end of the curve is checked for the presence of a very low line and if so, the term fails.

12) It is checked whether the bottom line of the whole non-peak area of the curve is sufficiently low. If the bottom line is too high, the item does not pass.

The criteria include, but are not limited to, the above items, which may be adjusted according to the actual situation.

And (4) establishing a discrimination matrix, wherein if all values are 1 (all pass), the curve data is valid data, and otherwise, the curve data is invalid data.

Step S306, if the judgment result indicates that the fluorescence curve is valid data, the judgment result, the ID information of the fluorescence curve and the remark information are used as first identification information, and the first identification information and the fluorescence curve are stored in a database in a correlation mode.

Specifically, the data structure of the fluorescence curve is the fluorescence value of 1 × X fluorescence points. In the fluorescence project based on fluorescence immunochromatography, X is generally 350. And packaging the judgment result, the ID information of the fluorescence curve, the remark information and the fluorescence curve into a whole, and storing the whole into a database. The remark information includes, but is not limited to, a record of the communication between the user and the client. In order to prevent duplication and redundancy of data, the data newly stored in the database should generate unique ID information including the time when the data was generated, the serial number of the instrument generating the data, the place where the instrument was put, the test item, the reagent lot number, the negative and positive results, and the like. And continuously storing the information with the judgment result identification into the database, so that the database can be used as a learning sample of machine learning, and the AI recognition model can be continuously optimized.

Step S308, if the judgment result indicates that the fluorescence curve is invalid data, judging the invalid reason of the fluorescence curve, classifying and analyzing the invalid reason to obtain the invalid reason classification and standard operation suggestion; and storing the invalid reason, the invalid reason classification and the standard operation suggestion in a database in association with the fluorescence curve.

Specifically, the classification and analysis of the invalidation reason is performed based on a database, a heterogeneous inspection algorithm and a heterogeneous characteristic inspection rule. Specifically, the null analysis is based first on N heterogeneous tests that yield a 1 × N decision matrix (e.g.[1,1,0,1,1,1,0,1,0,1,1,1,1]). If the matrix values are all 1, the curve data is valid, otherwise, the curve data is invalid. If the decision matrices are not all 1, error classification (miscause analysis) is first performed by the values of the decision matrices. The value of the decision matrix may theoretically have a value of 2^NThe possible forms of the error curve include all the error curve data, but in practice, the forms of the error curves are strange, and a small part of error curve data is difficult to be accurately distinguished by using the heterogeneous decision matrix. For problematic error curve data, in addition to direct human intervention, we use self-built databases for further identification, using methods including but not limited to: a least square Euclidean distance matching method based on a curve database, a curve data classification method based on image recognition algorithms such as a convolutional neural network, a KNN proximity algorithm, a BP neural network and an SVM support vector machine, and the like. The most important roles of the above database are: by the aid of the advantages of the data quantity, the AI curve interpretation and identification precision is improved, and the precision and the efficiency of detection result output are improved.

And step S310, acquiring a manual reinspection detection result of the preliminary detection result at a specified time, and taking the manual reinspection detection result as an updated preliminary detection result if the manual reinspection detection result indicates that the preliminary detection result is incorrect.

In a specific implementation, the preliminary detection result includes a fluorescence curve, unique ID information, marking information (identification information) and remarks. The method can backtrack the abnormal reason, and the abnormal information and the abnormal reason are manually added into the remarks. And (3) finding result abnormality: mainly discovered and fed back by after-sales personnel, operation and maintenance personnel or users.

Step S312, acquiring a manual recheck judgment result of the judgment result at a specified time, and if the manual recheck judgment result indicates that the judgment result is incorrect, taking the manual recheck detection result as an updated judgment result; and determining identification information according to the updated judgment result, and storing the identification information and the fluorescence curve in a database in a correlation manner.

During specific implementation, the accuracy of automatic effectiveness judgment, error classification and invalid classification of AI reaches more than 99% through the current experimental data. There are two main methods for knowing the inaccurate detection result: firstly, further testing the stability and accuracy of the algorithm; secondly, for the case that the interpretation in the newly generated detection data is inaccurate, the case is mainly found and fed back by operation and maintenance personnel, after-sales personnel or users. If the judgment result is wrong, the identification information is updated manually, and the updated identification information is stored in a database. During manual calibration, the data entries that the open authority allows the operator to manually alter are: curve data validity, invalid curve data type and remark. The structure of the identification information is as follows: the identification information is N isomerous detection judgment matrix + curve data validity; or the identification information is N heterogeneous test discrimination matrix + invalid curve data type. The ID information cannot be changed manually.

In step S314, if the determination result cannot be obtained according to the preset rule and the database, a manual determination result is obtained.

And step S316, generating a report in a specified form according to the requirement of the customer through the database.

Specifically, the invention can automatically generate the detection result by one key and periodically output the report in the designated form. The report includes single detection results, namely, the date of single detection, test items, equipment serial numbers, equipment putting sites, reagent batch numbers, C values, T values, validity interpretation results, invalid error classification interpretation results, detection target concentrations, negative and positive, curve pictures of single detection result curve data and the like. If the curve data is invalid, outputting a corresponding standard operation suggestion according to the error type. The report can be a daily report, a weekly report, a monthly report or other forms of reports customized according to the requirements of customers.

According to the embodiment, the identification precision and the identification efficiency of the fluorescence curve are improved through methods such as an AI model based on heterogeneous characteristic discrimination, an improved database, machine learning and the like. After the recognition accuracy of the machine learning model reaches a certain degree, the workload of the after-sale service department can be greatly saved. The method can greatly optimize user experience, improve the after-sale efficiency, promote the realization of industrial digital transformation and further feedback the whole flow of business by specialization of the fluorescence curve and development and arrangement of a system (effective and ineffective judgment, error type judgment, self-built database and automatic report output).

EXAMPLE III

The embodiment of the present invention further provides a fluorescence curve detection apparatus based on fluorescence immunochromatography, as shown in fig. 4, the apparatus includes:

a data acquisition module 41 configured to: and acquiring a preliminary detection result, wherein the preliminary detection result comprises a fluorescence curve and matched information of the fluorescence curve.

A determining module 42, configured to: and according to a preset rule and a database, judging the effectiveness of the fluorescence curve to obtain a judgment result.

A database perfecting module 43 for: determining identification information according to the judgment result, and storing the identification information and the fluorescence curve in a database in a correlation manner; the database is used for training a model for judging the effectiveness of the fluorescence curve and determining the identification information.

A database perfecting module 43, further configured to: and if the judgment result indicates that the fluorescence curve is valid data, taking the judgment result, the ID information of the fluorescence curve and the remark information as first identification information, and storing the first identification information and the fluorescence curve in a database in a correlation manner.

The determining module 42 is further configured to: if the judgment result indicates that the fluorescence curve is invalid data, judging the invalid reason of the fluorescence curve, classifying and analyzing the invalid reason to obtain invalid reason classification and standard operation suggestions; and storing the invalid reason, the invalid reason classification and the standard operation suggestion in a database in association with the fluorescence curve.

And the manual detection module is used for acquiring a manual re-detection result of the preliminary detection result at the specified time, and if the manual re-detection result indicates that the preliminary detection result is incorrect, taking the manual re-detection result as an updated preliminary detection result.

The manual detection module is further used for: acquiring a manual rechecking judgment result of the judgment result at a specified time, and if the manual rechecking judgment result indicates that the judgment result is incorrect, taking the manual rechecking detection result as an updated judgment result; and determining identification information according to the updated judgment result, and storing the identification information and the fluorescence curve in a database in a correlation manner.

The manual detection module is further used for: and if the judgment result cannot be obtained according to the preset rule and the database, obtaining a manual judgment result.

The determining module 42 is further configured to: performing smoothing and normalization processing on the fluorescence curve to obtain a curve processing result; performing peak discrimination on the curve processing result by a peak discrimination method to obtain a curve peak; and (4) carrying out validity check on the curve processing result and the curve wave crest by combining a heterogeneous characteristic checking rule, a machine self-learning algorithm and a database to obtain a judgment result.

An output module to: and generating a report in a specified form according to the requirements of customers through a database.

The fluorescence curve detection device based on the fluorescence immunochromatography provided by the embodiment of the invention has the same technical characteristics as the fluorescence curve detection method based on the fluorescence immunochromatography provided by the embodiment, so that the same technical problems can be solved, the same technical effects can be achieved, and the details are not repeated herein.

Example four

This example describes a method for analyzing fluorescence detection results based on fluorescence immunochromatography based on self-constructed databases, as shown in FIG. 5.

A method for analyzing fluorescence detection results of a self-constructed database based on fluorescence immunochromatography, comprising the following steps:

1) and acquiring detection data.

2) And (4) carrying out AI automatic processing on the data to judge the validity of the data. If the data is invalid, the error type is continuously judged.

And (2.1) preprocessing the data, wherein the processing mode comprises but is not limited to smoothing and normalization.

And (2.2) identifying the peak of the curve by using a maximum peak discrimination method.

And (2.3) carrying out immune curve validity test on the curve data based on an N-term isomerism characteristic test algorithm.

And (2.4) if the curve data is invalid, carrying out error classification according to the numerical value of the discrimination matrix.

3) The interpretation result is used as an identifier and is uploaded to a database together with the detection data. The data structure of the data result of the primary detection is 1TEST ═ fluorescence curve data + unique ID + marking information + remarks.

And (3.1) uploading the detection data and the interpretation identification result to a database. The data structure of the data result of the primary detection is as follows: 1TEST is fluorescence curve data + unique ID + marking information + remarks. The structure of the fluorescence curve data is as follows: fluorescence values of 1X fluorescence points. In the fluorescence project based on fluorescence immunochromatography, X is generally 350.

And (3.2) carrying out manual marking correction on the difficult data with inaccurate AI interpretation or wrong AI interpretation, and carrying out manual marking on the detection data. During the marking process, the data items that the openable right allows the operator to change are as follows: curve data validity, invalid curve data type and remark. The structure of the identification information is as follows: and identifying the N heterogeneous test discrimination matrix, curve data validity and invalid curve data type.

(3.3) in order to prevent data duplication and redundancy, the data newly stored in the database should be generated with a unique ID. The information that needs to be included in the ID serial number includes, but is not limited to: time of data generation, instrument serial number of generated data, instrument place of delivery, test item, reagent lot number, negative and positive results.

(3.4) the remark information includes but is not limited to: records of user-to-client communications, etc.

4) And (4) putting the data with the correct interpretation result identification into a warehouse as a learning sample of machine learning to optimize an AI recognition model, and continuously perfecting the process.

And (4.1) taking the data in the database as a machine learning sample set, and increasing the data volume continuously to improve the AI identification precision.

And (4.2) besides the AI model based on the heterogeneous feature discrimination, the AI model can be optimized by a machine learning method, and other machine learning models can be established by using a database so as to further enhance the identification efficiency.

And (4.3) after the recognition accuracy of the machine learning model reaches a certain degree, the workload of the after-sale service department can be greatly saved.

5) Automatically generating a detection result and outputting a report periodically.

And (5.1) reserving a function of outputting a single detection result by one key in the server. The output information items are: the date of single detection, test items, equipment serial numbers, equipment putting places, reagent batch numbers, C values, T values, effectiveness interpretation results, invalid error classification interpretation results, detection target concentration, negative and positive, curve pictures of single detection result curve data and the like. If the curve data is invalid, outputting a corresponding standard operation suggestion according to the error type.

And (5.2) keeping a one-key report output function in the server. The report mainly comprises: customizing a custom daily, weekly, monthly or other form report of the equipment according to the requirement of a customer.

The method for analyzing the fluorescence detection result based on the fluorescence immunochromatography provided by the embodiment of the invention has the same technical characteristics as the method for detecting the fluorescence curve based on the fluorescence immunochromatography provided by the embodiment, so that the same technical problems can be solved, the same technical effects can be achieved, and further description is omitted.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.