Method for evaluating sepsis lung injury by detecting crosslinking degree of type VI collagen

Abstract

The invention discloses a method for evaluating sepsis lung injury by detecting the crosslinking degree of type VI collagen, and relates to the technical field of medical diagnosis and monitoring. The method at least comprises the following steps: S1, firstly, carrying out sample collection and multi-index detection; s2, performing imaging feature extraction, and automatically identifying fine structure changes in the chest CT image by using a deep learning algorithm; and S3, based on the feature extraction in the S2, establishing a corresponding quality control algorithm, a basic damage scoring model, a dynamic progress prediction model and a comprehensive risk assessment model. The accurate, early-stage and dynamic sepsis lung injury assessment method is provided through multi-dimensional biomarker detection, iconography feature extraction and a dynamic algorithm model, the early diagnosis rate of sepsis lung injury can be remarkably improved, the disease progress can be effectively monitored, the treatment effect can be effectively assessed, and the application prospect is wide. The method has remarkable clinical application prospects and economic benefits, and particularly has important value in intensive care and formulation of personalized treatment schemes.

Description

Technical Field

[0001] This invention relates to the field of medical diagnostics and monitoring technology, specifically a method for assessing sepsis-related lung injury by detecting the degree of type VI collagen crosslinking. Background Technology

[0002] Sepsis is a systemic inflammatory response syndrome caused by infection, and the resulting lung injury is one of the leading causes of death in patients in the intensive care unit. Currently, clinical assessment of septic lung injury mainly relies on the following methods:

[0003] Traditional biomarker detection methods include C-reactive protein, procalcitonin, and inflammatory factors (IL-6, TNF-α), but these markers lack tissue specificity and cannot accurately reflect the degree of damage to the lung tissue itself.

[0004] Imaging assessment: Chest CT and X-ray examinations can show lung infiltration and consolidation, but cannot quantify early microstructural changes and pose a risk of radiation exposure.

[0005] Physiological parameter monitoring: Although oxygenation index, respiratory mechanics parameters, etc. can reflect the lung function status, their sensitivity and specificity are limited, making it difficult to achieve early warning.

[0006] Existing collagen detection technologies: Although collagen detection has applications in areas such as liver fibrosis, a method for detecting the degree of cross-linking of type VI collagen has not yet been established, and research on its association with septic lung injury is still lacking.

[0007] The limitations of these methods are mainly reflected in:

[0008] Lung tissue damage cannot be identified at the molecular level in its early stages;

[0009] Lack of quantitative standards for assessing the extent of damage;

[0010] Unable to dynamically monitor disease progression and treatment effectiveness;

[0011] The evaluation results are greatly influenced by subjective factors, which is a shortcoming of existing technologies;

[0012] Therefore, a new solution is needed to address the above problems. Summary of the Invention

[0013] The purpose of this invention is to provide a method for assessing septic lung injury by detecting the degree of type VI collagen crosslinking, so as to solve the technical problems mentioned in the background art.

[0014] To achieve the above objectives, the present invention provides the following technical solution: a method for assessing septic lung injury by detecting the degree of type VI collagen crosslinking, comprising at least the following steps:

[0015] S1: First, sample collection and multi-indicator testing are carried out;

[0016] S2: Perform imaging feature extraction, use deep learning algorithms to automatically identify subtle structural changes in chest CT images, and extract lung tissue density distribution, texture features and morphological parameters;

[0017] S3: Based on the feature extraction of S2, build the corresponding quality control algorithm, basic damage scoring model, dynamic progression prediction model and comprehensive risk assessment model;

[0018] The quality control algorithm optimizes the detection quality by detecting parameters such as the coefficient of variation and the degree of fit.

[0019] The baseline lung injury scoring model generates a baseline lung injury score (LIS) based on CLI values ​​and other biomarkers. b ;

[0020] The dynamic progression prediction model uses CLI changes, MMP activity, and the rate of change of imaging features to predict the progression of damage.

[0021] The comprehensive risk assessment model integrates basic injury score, injury change rate and complication correction factor to generate comprehensive risk index R and evaluate treatment effect;

[0022] S4: Conduct clinical validation and parameter optimization. Based on large-sample clinical data, establish a parameter database and perform dynamic optimization through machine learning to ensure the model's broad applicability.

[0023] Furthermore, the sample collection includes at least blood, urine, and bronchoalveolar lavage fluid;

[0024] The multi-index detection includes main detection indicators, auxiliary detection indicators, and quality control indicators;

[0025] The main detection index is the type VI collagen crosslinking index (CLI).

[0026] The auxiliary detection indicators include at least MP activity and tissue inhibitor concentration;

[0027] The quality control indicator is the sample integrity index.

[0028] Furthermore, the quality control algorithm is used to calculate the type VI collagen crosslinking index (CLI), and the calculation process is as follows:

[0029]

[0030] Where concentrationi is the concentration of the type VI collagen cross-linking product; weighti is the weight of the type VI cross-linking product, based on its contribution to lung injury.

[0031] Furthermore, the basic injury scoring model establishes a basic lung injury scoring model based on changes in the type VI collagen crosslinking index and other biomarkers.

[0032]

[0033] Among them, w1, w2 and w3 are the weighting coefficients of each biomarker on the damage score; CLI: type VI collagen crosslinking index; MMP: matrix metalloproteinase activity; other factors are other possible biomarkers (such as IL-6, CRP, etc.).

[0034] Furthermore, the dynamic progression prediction model is used to predict the dynamic progression of lung injury in sepsis patients, employing a time-series-based autoregressive model.

[0035]

[0036] Where: LIS(t) is the lung injury score at time t; α is a constant term; βi is the autoregressive coefficient obtained from the training data; LIS(t−i) is the lung injury score at the previous time step; εt is the error term, representing random fluctuations that the model cannot explain.

[0037] Furthermore, the comprehensive risk index R combines the scores from various dimensions to provide an overall risk of lung injury for sepsis patients;

[0038]

[0039] in, Basic lung injury score; The rate of change of the lung injury score, i.e. Clinical factors are clinical factors analyzed using an autoregressive model. These are the weighting coefficients for each factor;

[0040] Finally, the treatment effect is evaluated based on the changes in lung injury scores before and after treatment.

[0041]

[0042] in, Lung injury score after treatment; Pre-treatment lung injury score;

[0043] when This indicates that the treatment was effective and the damage had healed to some extent.

[0044] when If the result is negative, it indicates poor treatment effectiveness and no improvement in the damage.

[0045] Compared with the prior art, the beneficial effects of the present invention are:

[0046] This invention provides a precise, early, and dynamic assessment method for sepsis-induced lung injury through multi-dimensional biomarker detection, imaging feature extraction, and dynamic algorithm models. It can not only significantly improve the early diagnosis rate of sepsis-induced lung injury, but also effectively monitor disease progression and assess treatment effects. It has significant clinical application prospects and economic benefits, and is especially valuable in the development of intensive care and personalized treatment plans. Attached Figure Description

[0047] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a schematic diagram of the overall process of the present invention. Detailed Implementation

[0049] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0050] This invention presents a method for accurately assessing the degree of lung injury in sepsis patients using a dynamic algorithm model. It integrates biomarker detection, imaging feature extraction, and clinical parameter analysis to establish a multi-level assessment system, providing a comprehensive solution for the early diagnosis, severity grading, and prognostic prediction of sepsis-related lung injury.

[0051] Please see Figure 1 A method for assessing septic lung injury by detecting the degree of type VI collagen crosslinking, comprising at least the following steps:

[0052] S1: First, sample collection and multi-indicator testing are carried out;

[0053] S2: Perform imaging feature extraction, use deep learning algorithms to automatically identify subtle structural changes in chest CT images, and extract lung tissue density distribution, texture features and morphological parameters;

[0054] S3: Based on the feature extraction of S2, build the corresponding quality control algorithm, basic damage scoring model, dynamic progression prediction model and comprehensive risk assessment model;

[0055] The quality control algorithm optimizes the detection quality by detecting parameters such as the coefficient of variation and the goodness of fit.

[0056] The baseline injury scoring model generates a baseline lung injury score (LIS) based on CLI values ​​and other biomarkers. b ;

[0057] The dynamic progression prediction model uses CLI changes, MMP activity, and the rate of change of imaging features to predict the progression of damage.

[0058] The comprehensive risk assessment model integrates basic injury scores, injury change rates, and complication correction factors to generate a comprehensive risk index R and evaluate treatment effectiveness.

[0059] S4: Conduct clinical validation and parameter optimization. Based on large-sample clinical data, establish a parameter database and perform dynamic optimization through machine learning to ensure the model's broad applicability.

[0060] Sample collection should include at least blood, urine, and bronchoalveolar lavage fluid;

[0061] Multi-indicator testing includes primary testing indicators, auxiliary testing indicators, and quality control indicators;

[0062] The main testing indicator is the type VI collagen crosslinking index (CLI).

[0063] Auxiliary detection indicators should include at least MP activity and tissue inhibitor concentration;

[0064] The quality control indicator is the sample integrity index.

[0065] The quality control algorithm is used to calculate the type VI collagen crosslinking index (CLI), and the calculation process is as follows:

[0066]

[0067] Where concentrationi is the concentration of the type VI collagen cross-linking product; weighti is the weight of the type VI cross-linking product, based on its contribution to lung injury.

[0068] The basic injury scoring model establishes a basic lung injury scoring model based on changes in type VI collagen crosslinking index and other biomarkers.

[0069]

[0070] Among them, w1, w2 and w3 are the weighting coefficients of each biomarker on the damage score; CLI: type VI collagen crosslinking index; MMP: matrix metalloproteinase activity; other factors are other possible biomarkers (such as IL-6, CRP, etc.).

[0071] The dynamic progression prediction model uses a time-series-based autoregressive model to predict the dynamic progression of lung injury in sepsis patients.

[0072]

[0073] Where: LIS(t) is the lung injury score at time t; α is a constant term; βi is the autoregressive coefficient obtained from the training data; LIS(t−i) is the lung injury score at the previous time step; εt is the error term, representing random fluctuations that the model cannot explain.

[0074] The comprehensive risk index R combines scores from various dimensions to provide an overall risk of lung injury for sepsis patients.

[0075]

[0076] in, Basic lung injury score; The rate of change of the lung injury score, i.e. Clinical factors are clinical factors analyzed using an autoregressive model. These are the weighting coefficients for each factor;

[0077] Finally, the treatment effect is evaluated based on the changes in lung injury scores before and after treatment.

[0078]

[0079] in, Lung injury score after treatment; Pre-treatment lung injury score;

[0080] when This indicates that the treatment was effective and the damage had healed to some extent.

[0081] when If the result is negative, it indicates poor treatment effectiveness and no improvement in the damage.

[0082] Further specific application examples are as follows:

[0083] In routine screening applications:

[0084] Collect 3 mL of venous blood from the patient;

[0085] CLI values ​​were detected using an optimized ELISA kit.

[0086] Calculate LIS using the basic scoring model b ;

[0087] Risk classification is performed based on preset thresholds;

[0088] In intensive care applications:

[0089] Monitor CLI and MMP activity daily;

[0090] Perform a chest CT scan every 48 hours;

[0091] A comprehensive evaluation is conducted using a complete algorithm system.

[0092] Generate dynamic risk trend reports.

[0093] In the evaluation of treatment effectiveness:

[0094] Comparison of CLI changes before and after treatment;

[0095] Calculate ΔLIS / Δt to assess treatment response;

[0096] The treatment plan is adjusted based on algorithmic predictions.

[0097] In summary, the present invention has the following significant advantages:

[0098] 1. Improved early diagnostic capabilities enable the detection of lung injury changes 24-48 hours before the onset of clinical symptoms; the detection sensitivity is 3-5 times higher than that of traditional inflammatory markers; and it can better achieve early warning of microscopic damage.

[0099] 2. The accuracy of the assessment has been significantly improved. Through cross-validation of multi-dimensional indicators, the misdiagnosis rate has been reduced by more than 40%, and the ability to dynamically monitor and provide continuous assessment of the condition has been enhanced. Individualized correction has improved the reliability of the results.

[0100] 3. It has outstanding clinical application value, enabling non-invasive detection to reduce patient suffering, automated analysis to reduce operational difficulty, and providing support for treatment decision-making.

[0101] 4. By combining collagen crosslinking detection with imaging features, a complete quality control system is established, and a proprietary dynamic prediction algorithm is developed accordingly.

[0102] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for assessing the degree of cross-linking of collagen type VI for detecting sepsis-induced lung injury, characterized in that: At least comprising the following steps: S1: First, sample collection and multi-index detection are performed; S2: Image features are extracted, and deep learning algorithms are used to automatically identify subtle structural changes in chest CT images, extract lung tissue density distribution, texture features, and morphological parameters; S3: Based on the feature extraction of S2, corresponding quality control algorithms, basic injury scoring models, dynamic progression prediction models, and comprehensive risk assessment models are built; The quality control algorithm optimizes the detection quality by detecting the coefficient of variation, fitting degree, and other parameters; The base injury score model generates a base lung injury score LIS from the CLI value and other biomarkers b ; The dynamic progression prediction model uses CLI changes, MMP activity, and the rate of change of image features to predict the progression of injury; The comprehensive risk assessment model integrates basic injury scores, injury change rates, and complication correction factors to generate a comprehensive risk index R and evaluate treatment effectiveness; S4: Clinical verification and parameter optimization are performed, and based on large sample clinical data, a parameter database is established and machine learning is dynamically optimized to ensure the wide applicability of the model.

2. The method for evaluating sepsis-induced lung injury by detecting the cross-linking degree of collagen type V according to claim 1, characterized in that: The sample collection at least includes blood, urine, and bronchoalveolar lavage fluid; The multi-index detection includes main detection indicators, auxiliary detection indicators, and quality control indicators; The main detection indicator is the VI collagen cross-linking index; The auxiliary detection indicator at least includes MP activity and tissue inhibitor concentration; The quality control indicator is the sample integrity index.

3. The method for evaluating sepsis-induced lung injury by detecting the cross-linking degree of collagen type V according to claim 1, characterized in that: The quality control algorithm is calculated for the VI collagen cross-linking index, and the calculation process is as follows: ， Wherein, concentrationi is the concentration of the ith VI collagen cross-linking product; weighti is the weight of the ith cross-linking product, based on its contribution to lung injury.

4. The method for evaluating sepsis-induced lung injury by detecting the cross-linking degree of collagen type V according to claim 3, characterized in that: The basic injury scoring model is established according to the changes of VI collagen cross-linking index and other biomarkers; ， Wherein, w1, w2 and w3 are the weight coefficients of each biomarker on injury score; CLI: is the VI collagen cross-linking index; MMP: is the matrix metalloproteinase activity; other factors are other possible biomarkers.

5. The method for evaluating sepsis-induced lung injury by detecting the cross-linking degree of collagen type V according to claim 4, characterized in that: The dynamic progression prediction model is a time series-based autoregressive model for predicting the dynamic progression of lung injury in sepsis patients; ， Wherein: LIS(t) is the lung injury score at time t; α is a constant term; βi is the autoregressive coefficient obtained by training data; LIS(t−i) is the lung injury score at the previous time; εt is the error term, representing the random fluctuations that the model cannot explain.

6. The method for evaluating sepsis-induced lung injury by detecting the cross-linking degree of collagen type VI according to claim 5, characterized in that: The comprehensive risk index R combines the scores of each dimension to provide the overall lung injury risk of sepsis patients; ， wherein, is a baseline lung injury score; is a rate of change in lung injury score, i.e. ; Clinical factors are clinical factors regressed by an autoregressive model; is a weight coefficient of each factor; Finally, the treatment effectiveness is evaluated according to the changes of lung injury score before and after treatment: ， wherein, is the post-treatment lung injury score; is the pre-treatment lung injury score; When then the treatment is good and the damage is recovered; When then the treatment is poor and the injury is not improved.

Images