Test paper inspection analysis method and system for medical inspection

Abstract

The invention discloses a test paper inspection analysis method and system for medical inspection, and relates to the technical field of medical inspection, and the method comprises the following steps: collecting a spectral signal, filtering data, and performing multi-dimensional feature optimization; constructing a depth feature decomposition model and a multi-objective optimization prediction model; and dynamically correcting the test paper data, and generating a visual report according to a detection result. According to the test paper inspection analysis method for the medical inspection, the surface of the test paper is subjected to high-precision scanning to obtain optical characteristic information, the sensitivity of test paper detection to low-concentration target substances is improved, collected multi-dimensional spectral data are decomposed into different characteristic signals, and a dynamic optimization model is combined, so that the accuracy of the test paper inspection analysis is improved. The characteristics of each target substance in the sample are comprehensively analyzed, the adaptability of the detection system is improved, the availability and usability of the detection result are improved by automatically correcting optical deviation and environmental noise in the collection process, and the method has better effects in the aspects of sensitivity, adaptability and availability.

Description

Technical Field

[0001] The present invention relates to the field of medical testing technology, and in particular to a test paper testing and analysis method and system for medical testing. Background Art

[0002] In recent years, medical testing technology has developed rapidly, especially in the field of biomarker detection and quantitative analysis. Test paper testing has been widely used due to its advantages of low cost, convenient operation and rapid results. From the early colorimetric method of naked eye observation, to the later simple optical colorimetric detection, to the currently used multi-spectral analysis and computer-assisted digital detection technology, test paper detection technology has undergone a leapfrog development from qualitative analysis to semi-quantitative analysis. The introduction of optical detection technology has significantly improved the sensitivity and accuracy of test paper detection. At the same time, the automated analysis system combined with the machine learning model has further improved the detection efficiency. However, in the face of multi-target analysis needs in complex biological samples, such as the coexistence of multiple biomarkers or the detection of low-concentration target substances, the existing technology still has certain limitations in accuracy, resolution and adaptability, and cannot fully meet the growing needs of precision medicine.

[0003] Although test strip detection technology has been widely used in the medical field, the current mainstream methods still have many shortcomings. First, the traditional colorimetric method is highly dependent on the color change of the test strip surface. This method is not only easily interfered by ambient light, but also often results in inaccurate results due to cross-interference of different markers in complex samples. Secondly, existing optical detection methods are mostly focused on single-band or simple optical feature extraction, which is difficult to fully reflect the multi-dimensional spectral characteristics of the test strip reaction area, resulting in insufficient sensitivity in the detection of low-concentration target substances. In addition, existing technologies generally lack the ability to dynamically calibrate and multi-target optimization of detection data. For example, when multiple markers coexist, traditional analysis methods are difficult to achieve high-precision qualitative and quantitative analysis at the same time. Finally, the interpretation of test results often relies on manual experience and lacks intelligent automated analysis tools, which not only increases the risk of subjective errors, but also makes it difficult to meet the needs of high-throughput detection. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is: the existing test paper inspection and analysis method has insufficient detection sensitivity, high sensitivity to ambient light, difficulty in dealing with cross-interference of multiple markers, and how to achieve optimization problems of high sensitivity, high accuracy and intelligent detection.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a test strip inspection and analysis method for medical testing, comprising collecting spectral signals, filtering and multi-dimensional feature optimization of data; constructing a deep feature decomposition model and a multi-objective optimization prediction model; dynamically correcting the test strip data, and generating a visual report based on the test results.

[0007] As a preferred solution of the test paper test analysis method for medical testing described in the present invention, the spectral signal acquisition includes acquiring the reflection, absorption and scattering signals of the test paper surface in different wavebands through a high-precision optical scanning device, capturing the color change, light absorption intensity and scattering characteristics on the test paper by scanning the reaction area on the test paper surface, and controlling and correcting the interference of the ambient light conditions, the inherent deviation of the scanning device and the optical property difference of the test paper material in real time during the acquisition process, and analyzing the characteristic differences between different spectral bands. The spectral signal acquisition is expressed as:

[0008]

[0009] I(λ,t)=R(λ)e -α(λ)t +S(λ)sin(βλt)

[0010] Wherein, I(λ,t) is the optical signal intensity of the test paper reaction area at wavelength λ and time t, R(λ) is the reflection spectrum intensity at wavelength λ, α(λ) is the light absorption coefficient at wavelength λ, t is the reaction time, S(λ) is the scattering spectrum intensity at, β is the coupling coefficient in the scattering characteristics, D(λ,t) is the normalized optical signal intensity under wavelength λ and time t, G(λ,t) is the interference intensity of the ambient background light, t1 and t2 are the starting point and end point of the time interval for optical signal acquisition, respectively.

[0011] As a preferred solution of the test paper test analysis method for medical testing described in the present invention, the filtering and multi-dimensional feature optimization of the data include optimizing through data processing after collecting preliminary spectral data, removing interference signals caused by equipment noise, ambient light changes and non-target components in the sample, and extracting the most representative characteristic signals for the reaction of the target substance from the original data through multi-dimensional feature analysis. The reaction time, the characteristics of different spectral bands and the potential complexity in the sample are considered in the process. The data filtering and multi-dimensional feature optimization are expressed as:

[0012]

[0013] in, is the calculated value of the preliminary intensity of the characteristic signal in the i-th scan, N is the number of data feature dimensions, and F j is the signal weight of the jth dimension, H jis the strength of the j-th dimension signal, T j is the reaction time value of the jth signal, δ is the time compensation coefficient, γ is the time sensitivity adjustment coefficient, and Y i is the normalized signal quality factor, M(λ) is the spectral noise signal function, dλ is the differential element in the integral, which represents the infinitesimal change of the wavelength variable λ during the integral process, and λ min and λ max are the minimum and maximum values ​​of the spectral wavelength range, respectively.

[0014] As a preferred solution of the test paper test analysis method for medical testing described in the present invention, the deep feature decomposition model includes the spectral signal detected by the test paper, which decomposes the complex spectral signal into components that are easy to analyze through the feature decomposition model. The feature components reflect the dynamic behavior of the target substance reaction and reveal the characteristics of different reaction stages. By comprehensively analyzing the error distribution of the multidimensional spectral signal during the decomposition process, the feature extraction result is optimized. The deep feature decomposition model is expressed as:

[0015]

[0016] Among them, F h (x, y) is the value of the two-dimensional characteristic signal after the gradient and the second-order derivative are calculated, F(x, y) is the two-dimensional characteristic distribution function of the test paper reaction area, x and y are the spatial coordinates of the reaction area, P k is the normalized eigendecomposition value, x1 and x2 are the starting point and end point of the spatial range of the feature distribution, E i is the error value of the i-th dimension, W i is the weight of the error in the i-th dimension, and m is the total number of error dimensions.

[0017] As a preferred solution of the test strip test analysis method for medical testing described in the present invention, the multi-objective optimization prediction model includes a comprehensive analysis of multiple parameters and a model for predicting the state of the target substance. The prediction model integrates various variables in the test strip detection, dynamically adjusts the weight, and outputs results that reflect the characteristics of the target substance. The results are adjusted according to the detection conditions and background conditions of different samples. The complex chemical reaction process in the test strip detection is predicted through the multi-objective optimization model, which is expressed as:

[0018]

[0019] Among them, Z t is the intermediate calculated value adjusted based on time, T p is the time variable of the pth reaction, θ is the dynamic adjustment index of time, φ p is the phase difference of the pth reaction signal, Z is the normalized optimized prediction value, U qis the mean of the qth signal feature, κ q is the attenuation coefficient of the qth signal feature.

[0020] As a preferred embodiment of the test strip test analysis method for medical testing described in the present invention, the dynamic correction of the test strip data includes eliminating various equipment errors and environmental interferences in the test strip detection process. The dynamic calibration automatically adjusts the detection parameters by real-time monitoring of the detection conditions and comparing the multi-dimensional spectral signals actually collected with the preset calibration reference data. The deviation is corrected by adjusting the sensitivity of the detection equipment and the weight of the signal processing. The dynamic calibration is expressed as:

[0021]

[0022] Among them, C x is the intermediate signal value of dynamic calibration, K is the amplitude constant of the oscillation signal, ω is the frequency of the oscillation signal, L is the amplitude constant of the attenuation signal, η is the exponential attenuation coefficient of the signal, C is the dynamic calibration factor, N k is the kth noise factor, and n is the total number of noise factors.

[0023] As a preferred embodiment of the test strip test analysis method for medical testing of the present invention, the generating of a visual report according to the test results includes, after the test analysis is completed, quantitatively analyzing the characteristics of the test signal, and generating a score in combination with the specific background information of the sample, and the score output is expressed as:

[0024]

[0025] Among them, S x is the intermediate calculated score signal value, R(t) is the calibrated signal strength, ρ is the growth rate of the time signal, T(t) is the time noise signal, γ is the signal phase adjustment coefficient, S is the output score value, V i is the variance of the ith signal feature, and o is the total number of dimensions of the signal feature.

[0026] Another object of the present invention is to provide a test paper testing and analysis system for medical testing, which can construct a deep feature decomposition model and a multi-objective optimization prediction model through a model building module, and is used to perform high-precision analysis on the collected and processed data, thereby solving the current problem of difficulty in effectively separating the characteristic signals of target substances when multiple markers coexist, and prone to cross-interference.

[0027] As a preferred solution of the test paper inspection and analysis system for medical inspection described in the present invention, it includes: an acquisition and processing module, a model building module, and a dynamic correction module; the acquisition and processing module is used to acquire spectral signals, filter the data and optimize multi-dimensional features; the model building module is used to build a deep feature decomposition model and a multi-objective optimization prediction model; the dynamic correction module is used to dynamically correct the test paper data and generate a visual report based on the test results.

[0028] A computer device includes a memory and a processor, wherein the memory stores a computer program, and wherein the processor executes the computer program to implement a test paper test analysis method for medical testing.

[0029] A computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the steps of a test strip test analysis method for medical testing.

[0030] Beneficial effects of the present invention: The test paper inspection and analysis method for medical inspection provided by the present invention performs high-precision scanning on the test paper surface, obtains multi-band optical characteristic information including reflected light, absorbed light and scattered light, improves the sensitivity of the test paper detection to low-concentration target substances, and enhances the resolution capability of multi-target characteristics in complex biological samples. Through the deep feature decomposition model, the collected multi-dimensional spectral data is decomposed into different characteristic signals, and combined with the dynamic optimization model, the characteristics of each target substance in the sample are comprehensively analyzed, thereby improving the adaptability of the detection system. By automatically correcting the optical deviation and environmental noise in the collection process, the availability and ease of use of the detection results are improved. The present invention achieves better results in terms of sensitivity, adaptability and availability. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0032] Figure 1 The present invention provides an overall flow chart of a test paper testing and analysis method for medical testing provided by the first embodiment of the present invention.

[0033] Figure 2 This is an overall module diagram of a test paper testing and analysis system for medical testing provided in the third embodiment of the present invention. DETAILED DESCRIPTION

[0034] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0035] Example 1

[0036] Reference Figure 1 , as an embodiment of the present invention, provides a test paper test analysis method for medical testing, comprising:

[0037] S1: Collect spectral signals, filter the data and optimize multi-dimensional features.

[0038] Furthermore, collecting the spectral signal includes using a high-precision optical scanning device.

[0039] It should be noted that the reflection, absorption and scattering signals of the test paper surface are collected within different bands. By scanning the reaction area on the test paper surface, the color change, light absorption intensity and scattering characteristics on the test paper are captured. During the collection process, the interference of ambient light conditions, the inherent deviation of the scanning equipment and the optical property differences of the test paper material are controlled and corrected in real time, and the characteristic differences between different spectral bands are analyzed. The spectral signal collection is expressed as:

[0040]

[0041] I(λ,t)=R(λ)e -α(λ)t +S(λ)sin(βλt)

[0042] Wherein, I(λ,t) is the optical signal intensity of the test paper reaction area at wavelength λ and time t, R(λ) is the reflection spectrum intensity at wavelength λ, α(λ) is the light absorption coefficient at wavelength λ, t is the reaction time, S(λ) is the scattering spectrum intensity at, β is the coupling coefficient in the scattering characteristics, D(λ,t) is the normalized optical signal intensity under wavelength λ and time t, G(λ,t) is the interference intensity of the ambient background light, t1 and t2 are the starting point and end point of the time interval for optical signal acquisition, respectively.

[0043] It should also be noted that during the test paper detection process, the optical scanning equipment scans the surface of the test paper with a multi-band light source to capture the optical response signal of the target area. The optical response signal includes multi-dimensional characteristics such as the intensity of reflected light, absorption of specific wavelengths, and angular distribution of scattered light. Through high-resolution sensors, these characteristic signals can be collected in real time at different wavelengths, time, and spatial regions. In order to eliminate interference caused by ambient light, the equipment dynamically adjusts the light source intensity and noise shielding mechanism to ensure that the collected signal truly reflects the chemical properties of the test paper reaction area. The collection process can capture slight color changes or optical properties to enhance the detection capability of low-concentration target substances.

[0044] Furthermore, data filtering and multidimensional feature optimization are included after collecting preliminary spectral data.

[0045] It should be noted that through data processing optimization, interference signals caused by equipment noise, ambient light changes and non-target components in the sample are removed, and the most representative characteristic signals for the reaction of the target substance are extracted from the original data through multi-dimensional feature analysis. The reaction time, characteristics of different spectral bands and potential complexity in the sample are considered in the process. Data filtering and multi-dimensional feature optimization are expressed as:

[0046]

[0047] in, is the calculated value of the preliminary intensity of the characteristic signal in the i-th scan, N is the number of data feature dimensions, and F j is the signal weight of the jth dimension, H j is the strength of the j-th dimension signal, T j is the reaction time value of the jth signal, δ is the time compensation coefficient, γ is the time sensitivity adjustment coefficient, and Y i is the normalized signal quality factor, M(λ) is the spectral noise signal function, dλ is the differential element in the integral, which represents the infinitesimal change of the wavelength variable λ during the integral process, and λ min and λ max are the minimum and maximum values ​​of the spectral wavelength range, respectively.

[0048] It should also be noted that the collected raw spectral signals may contain interference factors such as ambient light, equipment noise and test paper base reflection. It is necessary to perform quality filtering on the spectral signals through data preprocessing. According to the characteristics of different spectral bands, appropriate filtering mechanisms are applied to remove background noise. At the same time, signals directly related to the target reaction characteristics are separated and normalized to reduce deviations caused by differences in test paper batches or detection environments. Through feature extraction algorithms, core features closely related to the reaction characteristics of the target substance are extracted from multi-band signals. The optimization process improves the resolution and sensitivity of the signal.

[0049] S2: Build a deep feature decomposition model and a multi-objective optimization prediction model.

[0050] Furthermore, the deep feature decomposition model includes the spectral signal of the test paper detection through the feature decomposition model.

[0051] It should be noted that the complex spectral signal is decomposed into components that are easy to analyze. The characteristic components reflect the dynamic behavior of the target substance reaction and reveal the characteristics of different reaction stages. By comprehensively analyzing the error distribution of the multidimensional spectral signal during the decomposition process, the feature extraction results are optimized. The deep feature decomposition model is expressed as:

[0052]

[0053] Among them, F h (x, y) is the value of the two-dimensional characteristic signal after the gradient and the second-order derivative are calculated, F(x, y) is the two-dimensional characteristic distribution function of the test paper reaction area, x and y are the spatial coordinates of the reaction area, P k is the normalized eigendecomposition value, x1 and x2 are the starting point and end point of the spatial range of the feature distribution, E i is the error value of the i-th dimension, W i is the weight of the error in the i-th dimension, and m is the total number of error dimensions.

[0054] It should also be noted that spectral signals contain multi-dimensional complex data, which need to be decomposed to extract key information. Feature decomposition is used to separate complex signals into single characteristic signals that are easy to analyze, such as changes in absorbance peaks or reflectivity at specific wavelengths. When decomposing signals, the interaction between different dimensions must also be considered, such as the coupling relationship between temporal dynamic changes and spectral characteristics. At the same time, to ensure the accuracy of the decomposition results, the error distribution in the signal is modeled, such as signal deviations caused by equipment optical deviations or differences in sample characteristics. The chemical reaction characteristics of the target substance can be accurately identified through the decomposed characteristic signals.

[0055] Furthermore, the multi-objective optimization prediction model includes a comprehensive analysis of multiple parameters.

[0056] It should be noted that a model for predicting the state of the target substance is constructed. The prediction model integrates various variables in the test strip detection, dynamically adjusts the weights, and outputs results that reflect the characteristics of the target substance. It is adjusted according to the detection conditions and background conditions of different samples. Through the multi-objective optimization model, the complex chemical reaction process in the test strip detection is predicted, which is expressed as:

[0057]

[0058] Among them, Z t is the intermediate calculated value adjusted based on time, T p is the time variable of the pth reaction, θ is the dynamic adjustment index of time, φ p is the phase difference of the pth reaction signal, Z is the normalized optimized prediction value, U q is the mean of the qth signal feature, κ q is the attenuation coefficient of the qth signal feature.

[0059] It should also be noted that the goal of test paper detection is not only to identify the presence of the target substance, but also to simultaneously detect multiple components or analyze their concentration changes. By constructing a multi-objective optimization model, the characteristic signals of different target substances are comprehensively analyzed. The model can adapt to the coexistence of multiple markers in complex samples, while optimizing the identification and quantification results of each target. The model dynamically adjusts the weights of different features to enhance weak signals and suppress strong interference signals. Combined with the time dynamic characteristics of the sample, the model can predict the completion state of the chemical reaction and the final concentration of the target substance, thereby improving the accuracy and adaptability of the detection and reducing errors caused by cross-interference.

[0060] S3: Dynamically calibrate the test strip data and generate a visual report based on the test results.

[0061] Furthermore, dynamic correction of the test strip data includes eliminating various equipment errors and environmental interferences during the test strip detection process.

[0062] It should be noted that dynamic calibration monitors the detection conditions in real time and compares the multi-dimensional spectral signals actually collected with the preset calibration reference data to automatically adjust the detection parameters and correct the deviation by adjusting the sensitivity of the detection equipment and the weight of signal processing. Dynamic calibration is expressed as:

[0063]

[0064] Among them, C x is the intermediate signal value of dynamic calibration, K is the amplitude constant of the oscillation signal, ω is the frequency of the oscillation signal, L is the amplitude constant of the attenuation signal, η is the exponential attenuation coefficient of the signal, C is the dynamic calibration factor, Nk is the kth noise factor, and n is the total number of noise factors.

[0065] Furthermore, generating a visual report based on the test results includes, after completing the test analysis, performing a quantitative analysis on the characteristics of the test signal and generating a score based on the specific background information of the sample. The score output is expressed as:

[0066]

[0067] Among them, S x is the intermediate calculated score signal value, R(t) is the calibrated signal strength, ρ is the growth rate of the time signal, T(t) is the time noise signal, γ is the signal phase adjustment coefficient, S is the output score value, V i is the variance of the ith signal feature, and o is the total number of dimensions of the signal feature.

[0068] Example 2

[0069] An embodiment of the present invention provides a test paper testing and analysis method for medical testing. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0070] Multiple groups of experiments were conducted, covering different environmental conditions and sample types. Six groups of samples were selected for the experiment, each of which contained target substances in different concentrations to simulate real medical testing scenarios. The experimental equipment included optical scanners, data processing platforms and dynamic calibration systems. The optical scanner supports multi-band spectral acquisition (including reflection, scattering and absorption signals), and the data processing platform integrates deep feature decomposition models and multi-objective optimization prediction algorithms.

[0071] During the experiment, each group of samples was first spectrally scanned, and the original signals were collected in the reaction area through multi-dimensional optical technology, and the reflected light intensity and environmental noise level were recorded. Then, the collected data was input into the data integration platform, and the characteristic decomposition model was used to extract the target characteristic peaks and filter the noise. On this basis, the multi-objective optimization algorithm was used to dynamically fit the characteristics of the target substance to generate a predicted concentration. To ensure the accuracy of the results, the experiment combined with dynamic calibration technology to adjust in real time the deviations introduced by ambient light, equipment errors and test paper batch differences. Finally, a test report was generated based on the calibration results, and key parameters such as accuracy and environmental interference level were recorded.

[0072] As shown in Table 1, after dynamic calibration, the signal strength of each sample is significantly improved, and the interference of environmental noise on the detection results is reduced compared with the initial signal strength. The signal strength of sample A after calibration is from 12.5W / m 2 Optimized to 11.3W / m 2 , background noise level from 1.2W / m2 The calibrated signal not only enhances the detectability of the target feature, but also improves the detection accuracy of low-concentration target substances (such as sample F, 4.7 mg / L) to 98.3%. From the accuracy data, the average accuracy of each sample detection exceeds 97.9%, which is significantly higher than the traditional detection method (usually 90%-93%). In addition, the degree of ambient light interference is effectively controlled, with the highest interference value being only 0.8%, which proves the adaptability of the method of the present invention to changes in ambient light.

[0073] Table 1 Experimental data table

[0074]

[0075] Example 3

[0076] Reference Figure 2 , which is an embodiment of the present invention, provides a test paper inspection and analysis system for medical inspection, including: an acquisition and processing module, a model building module, and a dynamic correction module.

[0077] The acquisition and processing module is used to collect spectral signals, filter the data and optimize multi-dimensional features; the model building module is used to build a deep feature decomposition model and a multi-objective optimization prediction model; the dynamic correction module is used to dynamically correct the test paper data and generate a visual report based on the test results.

[0078] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0079] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0080] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0081] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc. It should be noted that the above embodiments are only used to illustrate the technical solution of the present invention and are not limited. Although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

[0082] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A test paper test analysis method for medical examination, characterized in that: include: Collect spectral signals, filter the data and optimize multi-dimensional features; Construct deep feature decomposition models and multi-objective optimization prediction models; Dynamically calibrate the test strip data and generate a visual report based on the test results.

2. The test paper analysis method for medical examination as claimed in claim 1, characterized in that: The spectral signal collection includes collecting the reflection, absorption and scattering signals of the test paper surface in different wavebands through high-precision optical scanning equipment, capturing the color change, light absorption intensity and scattering characteristics on the test paper by scanning the reaction area on the test paper surface, and controlling and correcting the interference of the ambient light conditions, the inherent deviation of the scanning equipment and the optical property difference of the test paper material in real time during the collection process, and analyzing the characteristic differences between different spectral bands. The spectral signal collection is expressed as: I(λ,t)=R(λ)e -α(λ)t +S(λ)sin(βt) Wherein, I(λ,t) is the optical signal intensity of the test paper reaction area at wavelength λ and time t, R(λ) is the reflection spectrum intensity at wavelength λ, α(λ) is the light absorption coefficient at wavelength λ, t is the reaction time, S(λ) is the scattering spectrum intensity at, β is the coupling coefficient in the scattering characteristics, D(λ,t) is the normalized optical signal intensity under wavelength λ and time t, G(λ,t) is the interference intensity of the ambient background light, t1 and t2 are the starting point and end point of the time interval for optical signal acquisition, respectively.

3. The test paper analysis method for medical examination as claimed in claim 2, characterized in that: The filtering and multi-dimensional feature optimization of the data include optimizing through data processing after collecting preliminary spectral data, removing interference signals caused by equipment noise, ambient light changes and non-target components in the sample, and extracting the most representative characteristic signals for the reaction of the target substance from the original data through multi-dimensional feature analysis. The process takes into account the reaction time, the characteristics of different spectral bands and the potential complexity in the sample. The data filtering and multi-dimensional feature optimization are expressed as follows: in, is the calculated value of the preliminary intensity of the characteristic signal in the i-th scan, N is the number of data feature dimensions, and F j is the signal weight of the jth dimension, H j is the strength of the j-th dimension signal, T j is the reaction time value of the jth signal, δ is the time compensation coefficient, γ is the time sensitivity adjustment coefficient, and Y i is the normalized signal quality factor, M(λ) is the spectral noise signal function, dλ is the differential element in the integral, which represents the infinitesimal change of the wavelength variable λ during the integral process, and λ min and λ max are the minimum and maximum values ​​of the spectral wavelength range, respectively.

4. The test paper analysis method for medical examination as claimed in claim 3, characterized in that: The deep feature decomposition model includes a spectral signal detected by a test strip, which decomposes the complex spectral signal into components that are easy to analyze. The characteristic components reflect the dynamic behavior of the target substance reaction and reveal the characteristics of different reaction stages. The error distribution of the multidimensional spectral signal in the decomposition process is comprehensively analyzed to optimize the feature extraction result. The deep feature decomposition model is expressed as: Among them, F h (x, y) is the value of the two-dimensional characteristic signal after the gradient and the second-order derivative are calculated, F(x, y) is the two-dimensional characteristic distribution function of the test paper reaction area, x and y are the spatial coordinates of the reaction area, P k is the normalized eigendecomposition value, x1 and x2 are the starting point and end point of the spatial range of the feature distribution, E i is the error value of the i-th dimension, W i is the weight of the error in the i-th dimension, and m is the total number of error dimensions.

5. The test paper analysis method for medical examination as claimed in claim 4, characterized in that: The multi-objective optimization prediction model includes comprehensive analysis of multiple parameters and building a model for predicting the state of the target substance. The prediction model integrates various variables in the test strip detection, dynamically adjusts the weight, and outputs results that reflect the characteristics of the target substance. It is adjusted according to the detection conditions and background conditions of different samples. Through the multi-objective optimization model, the complex chemical reaction process in the test strip detection is predicted, which is expressed as: Among them, Z t is the intermediate calculated value adjusted based on time, T p is the time variable of the pth reaction, θ is the dynamic adjustment index of time, φ p is the phase difference of the pth reaction signal, Z is the normalized optimized prediction value, U q is the mean of the qth signal feature, k q is the attenuation coefficient of the qth signal feature.

6. The test paper analysis method for medical examination as claimed in claim 5, characterized in that: The dynamic calibration of the test paper data includes eliminating various equipment errors and environmental interferences in the test paper detection process. The dynamic calibration automatically adjusts the detection parameters by monitoring the detection conditions in real time and comparing the multi-dimensional spectral signals actually collected with the preset calibration reference data. The deviation is corrected by adjusting the sensitivity of the detection equipment and the weight of signal processing. The dynamic calibration is expressed as: Among them, C x is the intermediate signal value of dynamic calibration, K is the amplitude constant of the oscillation signal, ω is the frequency of the oscillation signal, L is the amplitude constant of the attenuation signal, η is the exponential attenuation coefficient of the signal, C is the dynamic calibration factor, N k is the kth noise factor, and n is the total number of noise factors.

7. The test paper analysis method for medical examination as claimed in claim 6, characterized in that: Generating a visual report based on the test results includes, after completing the test analysis, performing a quantitative analysis on the features of the test signal and generating a score in combination with the specific background information of the sample. The score output is expressed as: Among them, S x is the intermediate calculated score signal value, R(t) is the calibrated signal strength, ρ is the growth rate of the time signal, T(t) is the time noise signal, γ is the signal phase adjustment coefficient, S is the output score value, V i is the variance of the ith signal feature, and o is the total number of dimensions of the signal feature.

8. A system using the test paper testing and analysis method for medical testing as claimed in any one of claims 1 to 7, characterized in that: It includes acquisition and processing module, model building module and dynamic correction module; The acquisition and processing module is used to collect spectral signals, filter the data and optimize multi-dimensional features; The model building module is used to build a deep feature decomposition model and a multi-objective optimization prediction model; The dynamic correction module is used to dynamically correct the test paper data and generate a visual report based on the test results.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the test paper test analysis method for medical testing described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the test paper test analysis method for medical testing described in any one of claims 1 to 7 are implemented.

Images