A method and system for evaluating the effect of liver stem cells combined with intestinal flora in treating diabetes

Abstract

The application discloses a kind of for liver stem cell and intestinal flora combined treatment diabetes effect evaluation method and system, it is related to medical data processing technical field.The method is by quality detection to liver stem cell suspension and intestinal flora suspension before treatment, in-vivo liver stem cell and intestinal flora detection are carried out in cycle during treatment, and subsequent metabolic influence detection, diabetes related index detection, risk probability identification and auxiliary treatment scheme determination, realize from before treatment to the whole process accurate evaluation in treatment, ensure the quality of treatment material, timely grasp in-vivo treatment component state, accurately judge metabolism influence and treatment effect, effectively identify risk and provide pertinence auxiliary treatment scheme, can solve the problem that traditional diabetes treatment effect evaluation lacks comprehensive, dynamic evaluation to liver stem cell and intestinal flora combined treatment, cannot accurately judge treatment quality, in-vivo condition, metabolic influence, treatment effect and potential risk in time.

Description

Technical Field

[0001] This invention relates to the field of medical data processing technology, specifically to a method and system for evaluating the efficacy of combined treatment of diabetes with liver stem cells and gut microbiota. Background Technology

[0002] Diabetes mellitus, a metabolic disorder of carbohydrates, fats, and proteins characterized by hyperglycemia, can lead to various vascular and neurological complications affecting the heart, brain, kidneys, feet, and eyes if left uncontrolled, seriously endangering patients' lives and health. Chinese patent application CN117174327A discloses a system and storage medium for evaluating the treatment efficacy of diabetes based on a matrix factorization model. This system includes a data acquisition module, a data processing module, and a human-computer interaction device. The data processing module acquires evaluation results of the treatment efficacy of diabetes and its complications based on information and data from before and after treatment and prevention of complications in diabetic patients. This includes: labeling the evaluation criteria and classifications of the treatment efficacy of diabetes and its complications based on the acquired information and data; establishing a database of evaluation index standards for the treatment efficacy of diabetes and its complications in diabetic patients; and evaluating the sample data provided by the diabetic patients to be tested using the constructed evaluation model to obtain the evaluation results of the treatment efficacy of diabetes and its complications in diabetic patients.

[0003] However, existing methods for evaluating the effectiveness of diabetes treatment lack a comprehensive and dynamic assessment of the combined treatment of liver stem cells and gut microbiota, making it impossible to promptly and accurately determine the quality of treatment, the patient's condition, metabolic effects, treatment efficacy, and potential risks. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method and system for evaluating the efficacy of combined liver stem cell and gut microbiota therapy for diabetes. This solves the problem that traditional diabetes treatment efficacy evaluation methods lack comprehensive and dynamic assessment of combined liver stem cell and gut microbiota therapy, and cannot timely and accurately determine treatment quality, in vivo condition, metabolic effects, treatment efficacy, and potential risks.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for evaluating the efficacy of combined liver stem cell and gut microbiota therapy for diabetes, comprising the following steps: pre-treatment quality testing of liver stem cell suspension and gut microbiota suspension to determine whether the quality meets the standards; if the standards are not met, the liver stem cell suspension and gut microbiota suspension are replaced, and pre-treatment quality testing is performed again; if the standards are met, in vivo liver stem cell testing and gut microbiota testing are performed at set intervals during treatment to determine whether the in vivo liver stem cell testing and gut microbiota testing meet the standards; if both in vivo liver stem cell testing and gut microbiota testing meet the standards, treatment continues; if either in vivo liver stem cell testing or gut microbiota testing fails to meet the standards, in vivo liver stem cell metabolic impact testing and gut microbiota metabolic impact testing are performed.

[0006] If both the in vivo liver stem cell metabolism impact test and the gut microbiota metabolism impact test are normal, then diabetes-related indicators will be tested to obtain the treatment effect evaluation coefficient; if either the in vivo liver stem cell metabolism impact test or the gut microbiota metabolism impact test is abnormal, then abnormal signs risk identification will be performed to determine the probability of liver risk and the probability of intestinal risk, and at the same time, an adjuvant treatment plan will be determined.

[0007] Furthermore, pre-treatment quality testing was performed on liver stem cell suspensions and gut microbiota suspensions to determine whether the quality met the standards. This included the following steps: obtaining pre-treatment characteristic data of liver stem cells and gut microbiota, wherein the pre-treatment characteristic data of liver stem cells included the liver stem cell activity value Gx. hx Cell purity percentage (Gx) cd telomere length Gx dl Expression levels of genes related to glucose metabolism (Gx) td Insulin sensitivity-related cytokine levels Gx yd Pretreatment gut microbiota characteristics data included Akkermania relative abundance (CA). xf Bifidobacteria relative abundance Sq xf Shannon diversity index Jq dy Simpson dominance index Yq dy Short-chain fatty acid content Dx aj and branched-chain amino acid content Zx aj Pre-treatment characteristic data of liver stem cells and gut microbiota were obtained from the database. The pre-treatment characteristic data of liver stem cells included the liver stem cell activity value (Gc). hx Cell purity percentage reference value Gc cd Telomere length parameter Gc dl Parameters for the expression levels of genes related to glucose metabolism (Gc) tdReference value of insulin sensitivity-related cytokine content Gc yd Pretreatment gut microbiota characteristic data included the relative abundance value of Akkermansia (Cc). xf Bifidobacterium relative abundance parameter Sc xf Shannon diversity index reference value Jc dy Simpson dominance index parameter Yc dy Short-chain fatty acid content reference value Dc aj The reference value Zc for branched-chain amino acid content aj ;

[0008] Based on pre-treatment characteristic data and pre-treatment characteristic parameter data of liver stem cells, stem cell characteristic detection and stem cell metabolic secretion detection were performed to obtain stem cell characteristic evaluation coefficients Gx. tx And stem cell metabolic secretion assessment coefficient Gx fm The liver stem cell quality detection coefficient Gx was obtained based on the stem cell characteristic evaluation coefficient and the stem cell metabolic secretion evaluation coefficient. jc If the liver stem cell quality detection coefficient is greater than the liver stem cell quality detection threshold set in the database, then the liver stem cell quality meets the standard; if the liver stem cell quality detection coefficient is not greater than the liver stem cell quality detection threshold set in the database, then the liver stem cell quality does not meet the standard.

[0009] Based on pre-treatment gut microbiota characteristic data and pre-treatment gut microbiota characteristic parameter data, gut microbiota characteristics and gut microbiota metabolism were detected, and gut microbiota characteristic evaluation coefficients Cx were obtained respectively. tx And gut microbiota metabolic assessment coefficient Cx fm The gut microbiota quality detection coefficient Cx was obtained based on the gut microbiota characteristic assessment coefficient and the gut microbiota metabolism assessment coefficient. jc If the gut microbiota quality detection coefficient is greater than the gut microbiota quality detection threshold set in the database, then the gut microbiota quality meets the standard; if the gut microbiota quality detection coefficient is not greater than the gut microbiota quality detection threshold set in the database, then the gut microbiota quality does not meet the standard.

[0010] Furthermore, the liver stem cell quality detection coefficient Gx jc The calculation formula is:

[0011]

[0012] Where α1 is Gx tx The weighting factor, α2 is Gx fm Weighting factors;

[0013] Gut microbiota quality detection coefficient Cx jc :

[0014]

[0015] Where α3 is Cx tx The weighting factor, α4 is Cx fm Weighting factors.

[0016] Further, determining whether the in vivo liver stem cell detection criteria are met includes the following steps: obtaining stem cell factor concentration parameters based on blood tests, including hepatocyte growth factor concentration and insulin-like growth factor-1 concentration; obtaining stem cell behavior parameters based on labeling and tracing technology, including the stem cell survival rate Gx. ch The study included: a set of organ migration data (A1) and a set of distribution data of stem cells in various organs within the body (B1); standardized stem cell factor concentration parameters, followed by weighted summation to obtain a comprehensive coefficient for stem cell factor concentration; and retrieved stem cell behavioral parameters stored in the database, combining these parameters to obtain a stem cell behavior evaluation coefficient (Gx). xw ;

[0017]

[0018] Among them, Gc ch A1 represents the proportion of stem cell survival in the stem cell behavioral parameters, A2 represents the set of organs in which stem cells migrate in the stem cell behavioral parameters, B2 represents the set of distribution data of stem cells in various organs in the body in the stem cell behavioral parameters, σ(·) is the cosine similarity function, and e is the natural constant.

[0019] The in vivo liver stem cell detection index is obtained by weighted summing of the comprehensive coefficient of stem cell factor concentration and the stem cell behavior evaluation coefficient: if the in vivo liver stem cell detection index is greater than the in vivo liver stem cell detection threshold set in the database, the in vivo liver stem cell detection meets the standard; if the in vivo liver stem cell detection index is not greater than the in vivo liver stem cell detection threshold set in the database, the in vivo liver stem cell detection does not meet the standard.

[0020] Determining whether the gut microbiota testing meets the standards includes the following steps: obtaining the concentration ratio of bile acid metabolites to short-chain fatty acids, the rate of change of the relative abundance of key bacterial species, and the trend of change of the relative content of acetic acid, propionic acid, and butyric acid in short-chain fatty acids; if the concentration ratio of bile acid metabolites to short-chain fatty acids and the rate of change of the relative abundance of key bacterial species are both within the corresponding set range, and the trend of change of the relative content of acetic acid, propionic acid, and butyric acid in short-chain fatty acids all meet the expected trend, then the gut microbiota testing meets the standards; otherwise, the gut microbiota testing does not meet the standards.

[0021] Further, in vivo detection of the metabolic impact of liver stem cells is performed, including the following steps: acquiring data on the metabolic impact of stem cells, including intrinsic parameters of liver stem cells, parameters related to stem cell metabolites, and parameters related to overall metabolism in the body; normalizing the data and organizing it into a feature vector for detecting the metabolic impact of stem cells; inputting the feature vector into a trained decision tree, starting from the root node and traversing downwards according to the node partitioning conditions until a leaf node is reached, and counting the number of samples with normal metabolism and the number of samples with abnormal metabolism in the leaf nodes; calculating the proportion of samples with normal metabolism and the proportion of samples with abnormal metabolism based on the number of samples with normal metabolism and the number of samples with abnormal metabolism, respectively; if the proportion of samples with normal metabolism is greater than the proportion of samples with abnormal metabolism, then the detection of the metabolic impact of liver stem cells in vivo is normal; otherwise, it is abnormal.

[0022] The detection of the impact of gut microbiota metabolism includes the following steps: acquiring gut microbiota metabolic impact detection data, including gut microbiota structural parameters, gut microbiota metabolite-related parameters, and overall body metabolic correlation parameters; normalizing the gut microbiota metabolic impact detection data and organizing it into a gut microbiota metabolic impact detection feature vector; inputting the gut microbiota metabolic impact detection feature vector into a trained gradient decision tree model to obtain the final predicted value; converting the final predicted value into a probability value through a logistic function. If the probability value is less than a set probability threshold, the gut microbiota metabolic impact detection is normal; otherwise, it is abnormal.

[0023] Furthermore, obtaining the treatment efficacy evaluation coefficient includes the following steps: performing a change analysis on the currently acquired diabetes-related indicators and the previously acquired diabetes-related indicators to obtain the change factor bh. yz :

[0024]

[0025] Among them, Ssx i Dsx represents the difference between the i-th diabetes-related indicator that exceeds the set value and its corresponding set value from the previously acquired diabetes-related indicators. i To be with Ssx i The difference between the corresponding diabetes index and the corresponding reference value, where n is the total number of diabetes-related indicators that exceeded the reference value in the previous acquisition; Sx j Ddx represents the absolute value of the difference between the j-th diabetes-related indicator below the reference value and its corresponding reference value from the previously obtained diabetes-related indicators. j To be with Sx j The absolute value of the difference between the corresponding diabetes index and the corresponding reference value, where m is the total number of diabetes indices below the reference value among the previously obtained diabetes-related indices.

[0026] After normalizing the currently acquired diabetes-related indicators, a cosine similarity analysis was performed between the normalized indicators and the expected normalized diabetes-related indicators to obtain the similarity factor xs. yz The system retrieves current comprehensive body parameters and compares them with the comprehensive indicators in the database's comprehensive indicator-comprehensive risk value mapping table using cosine similarity. It then identifies the comprehensive indicator most similar to the current parameters and obtains the corresponding comprehensive risk value. Current comprehensive body parameters include metabolic syndrome-related indicators, insulin secretion and sensitivity indicators, inflammation and oxidative stress indicators, kidney function indicators, and liver function indicators. The system also considers the variable factor bh. yz Similarity factor xs yz Comprehensive risk value fx The proportion of people with normal metabolism (zc) bl and probability value P gl The data is input into the treatment efficacy evaluation model to obtain the treatment efficacy evaluation coefficient ZL. xs The treatment efficacy evaluation model takes the following form:

[0027]

[0028] Furthermore, the steps for obtaining the comprehensive indicator-comprehensive risk value mapping table are as follows: Obtain historical comprehensive physical parameter indicators for each treatment cycle; perform cluster analysis on the historical comprehensive physical parameter indicators based on the DBSCAN clustering algorithm to obtain the cluster centers corresponding to each treatment cycle; average the historical comprehensive physical parameter indicators corresponding to each cluster center to obtain the averaged historical comprehensive physical parameter indicators, denoted as comprehensive indicators, with each cluster center corresponding to one comprehensive indicator; count the number of samples where the historical comprehensive physical parameter indicators corresponding to each cluster center are abnormal in the next detection cycle, and calculate the abnormality probability; map each comprehensive indicator to the abnormality probability to obtain the comprehensive indicator-comprehensive risk value mapping table for each treatment cycle.

[0029] Further, determining the probability of liver and intestinal risk includes the following steps: obtaining and standardizing risk indicators for diabetic liver disease. The standardized risk indicators for diabetic liver disease include metabolic indicators and liver injury / fibrosis markers. Metabolic indicators include glycated hemoglobin (HbA1c) levels. hl Triglyceride content (GY) hl Low-density lipoprotein cholesterol (LDL-C) content DG hl Non-esterified fatty acid content ZH hl Liver damage and fibrosis markers include alanine aminotransferase (BA) levels. hl Aspartate aminotransferase content TD hl and liver stiffness value (GZ) hlBased on the trained Logistic regression model, the standardized risk indicators of diabetic liver disease were processed to obtain the first probability of liver disease, P1.

[0030] P1′=β0′+β1′*TH hl +β2′*GY hl +β3′*DG hl +β4′*ZH hl ;

[0031] P1″=β0″+β1″*BA hl +β2″*TD hl +β3″*GZ hl ;

[0032]

[0033] Where β0′ and β0″ are intercepts, β1′, β2′, β3′, β4′, β1″, β2″ and β3″ are regression coefficients, and P1′ and P1″ are transit functions;

[0034] Based on a trained liver disease decision tree model, the standardized risk indicators for diabetic liver disease were processed to obtain the second liver disease probability P2; the first liver disease probability P1, the second liver disease probability P2, and the proportion of metabolic abnormalities yc were also analyzed. bl After processing, the probability of liver risk is obtained:

[0035] P gz = (δ1*P1+δ2*P2)*(1+yc bl ); where δ1 is the weighting factor of P1 and δ2 is the weighting factor of P2;

[0036] Diabetic gut risk indicators were acquired and standardized. These standardized indicators included gut microbiota characteristics, inflammation and oxidative stress markers, and gut function indicators. A trained gut disease decision tree model was used to process the standardized gut risk indicators to obtain a first gut disease probability. Weighted scoring rules stored in a database were then used to process the standardized gut risk indicators to obtain a total risk score, which was mapped to a second gut disease probability. Finally, the first and second gut disease probabilities were weighted and summed to obtain the gut risk probability.

[0037] Further, the determination of the adjunctive treatment plan includes the following steps: Obtaining the diabetic patient's vital signs data, including height, age, body mass index (BMI), diabetes-related indicators, and current comprehensive body parameters; comparing the diabetic patient's vital signs data with the vital signs matching data-adjunctive treatment plan set mapping set stored in the database, determining the vital signs matching data closest to the diabetic patient's vital signs data, and obtaining the adjunctive treatment plan set, which includes multiple historical adjunctive treatment plans, each corresponding to a priority value; using the historical adjunctive treatment plan corresponding to the highest priority value in the adjunctive treatment plan set as the adjunctive treatment plan corresponding to the current diabetic patient's vital signs data; after setting the adjunctive treatment period, re-performing in vivo liver stem cell metabolism impact testing and gut microbiota metabolism impact testing: if both in vivo liver stem cell metabolism impact testing and gut microbiota metabolism impact testing are normal, adjunctive treatment is stopped; if either in vivo liver stem cell metabolism impact testing or gut microbiota metabolism impact testing is abnormal, the diabetic patient's vital signs data is re-obtained, and the adjunctive treatment plan is re-determined, while simultaneously updating the priority value of the current adjunctive treatment plan.

[0038] An efficacy evaluation system for the combined treatment of diabetes with liver stem cells and gut microbiota includes a pre-treatment quality testing module for testing liver stem cell suspensions and gut microbiota suspensions to determine if the quality meets standards; if not, the liver stem cell suspension and gut microbiota suspension are replaced, and the pre-treatment quality testing is repeated; and an in vivo testing module for performing in vivo liver stem cell and gut microbiota testing at set intervals during treatment, provided the liver stem cell and gut microbiota suspensions meet the standards, to determine if the in vivo liver stem cell and gut microbiota tests meet the standards; if both in vivo liver stem cell and gut microbiota tests meet the standards, treatment continues.

[0039] The metabolic impact detection module is used to detect the metabolic impact of liver stem cells and gut microbiota when either of these tests fails to meet the target. If both tests are normal, diabetes-related indicators are tested to obtain a treatment efficacy evaluation coefficient. The risk probability and auxiliary treatment plan determination module is used to identify the risk of abnormal signs, determine the liver risk probability and the gut risk probability, and determine the auxiliary treatment plan when either test is abnormal.

[0040] The present invention has the following beneficial effects:

[0041] This method for evaluating the efficacy of combined liver stem cell and gut microbiota therapy for diabetes involves pre-treatment quality testing of liver stem cell and gut microbiota suspensions, periodic in vivo testing of liver stem cells and gut microbiota during treatment, and subsequent metabolic impact testing, diabetes-related indicator testing, risk probability identification, and determination of adjuvant treatment plans. This achieves precise evaluation throughout the entire process, from pre-treatment to treatment, ensuring the quality of materials used in treatment, timely monitoring of the in vivo status of therapeutic components, accurate assessment of metabolic impact and treatment efficacy, effective risk identification, and provision of targeted adjuvant treatment plans. It addresses the shortcomings of traditional diabetes treatment efficacy evaluation methods, which lack comprehensive and dynamic assessment of combined liver stem cell and gut microbiota therapy, and are unable to promptly and accurately determine treatment quality, in vivo condition, metabolic impact, treatment efficacy, and potential risks.

[0042] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0043] Figure 1 This is a flowchart of the method for evaluating the efficacy of the combined treatment of diabetes with liver stem cells and gut microbiota according to the present invention.

[0044] Figure 2 This is a flowchart of the efficacy evaluation system for the combined treatment of diabetes using liver stem cells and gut microbiota according to the present invention. Detailed Implementation

[0045] Please see Figure 1 The present invention provides a technical solution: a method for evaluating the efficacy of combined treatment of diabetes with liver stem cells and intestinal flora, comprising the following steps: performing pre-treatment quality testing on liver stem cell suspension and intestinal flora suspension to determine whether the quality meets the standards; if the quality does not meet the standards, replacing the liver stem cell suspension and intestinal flora suspension, and re-performing the pre-treatment quality testing;

[0046] Pre-treatment characteristic data of liver stem cells and gut microbiota were obtained. The pre-treatment characteristic data of liver stem cells included the liver stem cell activity value (Gx). hx Cell purity percentage (Gx) cd telomere length Gx dl Expression levels of genes related to glucose metabolism (Gx) td Insulin sensitivity-related cytokine levels Gx yd Pretreatment gut microbiota characteristics data included Akkermania relative abundance (CA). xf Bifidobacteria relative abundance Sq xf Shannon diversity index Jq dy Simpson dominance index Yq dy Short-chain fatty acid content Dxaj and branched-chain amino acid content Zx aj .

[0047] The activity and purity of liver stem cells directly affect their differentiation and metabolic functions; telomere length reflects the degree of cellular senescence; and glucose metabolism genes and insulin-related factors are associated with their ability to regulate blood glucose. Characteristic data of the gut microbiota (such as the abundance of Akkermansia and Bifidobacterium, and the content of short-chain fatty acids) are closely related to intestinal barrier function and metabolic regulation. This comprehensive approach covers key indicators affecting liver stem cell function and gut microbiota metabolism, ensuring that subsequent evaluations are based on complete and accurate baseline data. It avoids biases in quality judgment due to missing data, thus guaranteeing the reliability of therapeutic substances from the source.

[0048] Pre-treatment characteristic parameters of liver stem cells and gut microbiota were obtained from the database. The pre-treatment characteristic parameters of liver stem cells included the liver stem cell activity value (Gc). hx Cell purity percentage reference value Gc cd Telomere length parameter Gc dl Parameters for the expression levels of genes related to glucose metabolism (Gc) td Reference value of insulin sensitivity-related cytokine content Gc yd Pretreatment gut microbiota characteristic data included the relative abundance value of Akkermansia (Cc). xf Bifidobacterium relative abundance parameter Sc xf Shannon diversity index reference value Jc dy Simpson dominance index parameter Yc dy Short-chain fatty acid content reference value Dc aj The reference value Zc for branched-chain amino acid content aj .

[0049] Pretreatment characteristic data of liver stem cells and gut microbiota, such as activity values ​​and microbiota diversity indices, were obtained from databases, providing clear quantitative standards for quality testing. By comparing actual test data with the reference data, it is possible to intuitively determine whether the therapeutic substance meets the preset quality requirements, making quality assessment objective and operable, and avoiding errors caused by subjective judgment.

[0050] Based on pre-treatment characteristic data and pre-treatment characteristic parameter data of liver stem cells, stem cell characteristic detection and stem cell metabolic secretion detection were performed to obtain stem cell characteristic evaluation coefficients Gx. tx And stem cell metabolic secretion assessment coefficient Gx fm The liver stem cell quality detection coefficient Gx was obtained based on the stem cell characteristic evaluation coefficient and the stem cell metabolic secretion evaluation coefficient. jcIf the liver stem cell quality detection coefficient is greater than the liver stem cell quality detection threshold set in the database, then the liver stem cell quality meets the standard; if the liver stem cell quality detection coefficient is not greater than the liver stem cell quality detection threshold set in the database, then the liver stem cell quality does not meet the standard.

[0051] Characteristic testing focuses on the biological functions of the cells themselves (such as activity and purity), while metabolic secretion testing focuses on their ability to regulate glucose metabolism. Combining the two allows for the assessment of liver stem cell quality from both structural and functional dimensions. A weighted calculation is used to generate a comprehensive coefficient, transforming the complex quality assessment into concrete numerical values ​​for easy and intuitive judgment of whether the cells meet therapeutic needs.

[0052] Based on pre-treatment gut microbiota characteristic data and pre-treatment gut microbiota characteristic parameter data, gut microbiota characteristics and gut microbiota metabolism were detected, and gut microbiota characteristic evaluation coefficients Cx were obtained respectively. tx And gut microbiota metabolic assessment coefficient Cx fm The gut microbiota quality detection coefficient Cx was obtained based on the gut microbiota characteristic assessment coefficient and the gut microbiota metabolism assessment coefficient. jc If the gut microbiota quality detection coefficient is greater than the gut microbiota quality detection threshold set in the database, then the gut microbiota quality meets the standard; if the gut microbiota quality detection coefficient is not greater than the gut microbiota quality detection threshold set in the database, then the gut microbiota quality does not meet the standard.

[0053] Characteristic assays (such as gut microbiota diversity indices) reflect the stability of gut microbiota structure, while metabolic assays (such as short-chain fatty acid content) are directly related to their regulatory effects on host glucose metabolism (e.g., short-chain fatty acids can improve insulin resistance). Quantitative calculations can identify gut microbiota with structural imbalances or insufficient metabolic capacity. For example, a low Shannon diversity index may indicate a weak regulatory function of the gut microbiota on host metabolism, requiring re-preparation to avoid affecting the efficacy of combination therapy.

[0054] Ensuring zero-risk access to therapeutic substances: Through a closed-loop process of "testing-replacement-retesting", unqualified therapeutic substances with insufficient activity or imbalanced flora are forcibly excluded, eliminating treatment ineffectiveness or safety risks caused by quality problems from the source (such as low-activity stem cells being unable to differentiate and repair pancreatic islets, and abnormal flora potentially causing inflammation), providing a basic guarantee for the safety and effectiveness of subsequent combination therapies.

[0055] Liver stem cell quality detection coefficient Gx jc The calculation formula is:

[0056]

[0057] Where α1 is Gx tx The weighting factor, α2 is Gx fmThe weighting factors; the formula combines "structure" (characteristics) and "function" (metabolism) quantitatively through weighted summation, avoiding the one-sidedness of evaluation by a single indicator.

[0058] Gut microbiota quality detection coefficient Cx jc :

[0059]

[0060] Where α3 is Cx tx The weighting factor, α4 is Cx fm The weighting factor combines structure and metabolic potential to identify bacteria with "normal structure but weak metabolic capacity" (such as Bifidobacterium abundance meeting the standard but short-chain fatty acid content being insufficient), thus avoiding poor treatment effects due to metabolic defects.

[0061] If the target is met, in vivo liver stem cell testing and gut microbiota testing will be performed according to the set cycle during the treatment process to determine whether the target is met. If both the in vivo liver stem cell testing and gut microbiota testing are within the target range, the treatment will continue. The liver stem cell suspension is administered intravenously and transplanted with blood, while the gut microbiota suspension is administered orally.

[0062] Determining whether in vivo liver stem cell testing meets the standards includes the following steps: Based on blood tests, obtain stem cell factor concentration parameters, including hepatocyte growth factor concentration (HGF, which promotes hepatocyte proliferation and differentiation, and its concentration reflects the liver stem cell's ability to regulate the liver microenvironment) and insulin-like growth factor-1 concentration (IGF-1, which improves glucose metabolism through the insulin-like signaling pathway and is directly related to the stem cell's regulatory effect on blood glucose); Based on labeling and tracing technology, obtain stem cell behavioral parameters, including the stem cell survival rate (Gx). ch The study analyzed two datasets: A1, a set of organ migration pathways for stem cells (normal stem cells should migrate to target organs such as the liver and pancreas; abnormal migration may pose safety risks); and B1, a set of data on stem cell distribution in various organs within the body (quantifying the colonization density of stem cells in target organs, directly affecting treatment efficacy). After standardizing stem cell factor concentration parameters, a weighted summation was performed to obtain a comprehensive coefficient for stem cell factor concentration. Finally, stem cell behavior parameters stored in the database were acquired, and combined with these parameters to obtain a stem cell behavior evaluation coefficient, Gx. xw ;

[0063]

[0064] Among them, Gc chA1 represents the proportion of stem cell survival in the stem cell behavioral parameters, A2 represents the set of organs in which stem cells migrate in the stem cell behavioral parameters, B2 represents the set of distribution data of stem cells in various organs in the body in the stem cell behavioral parameters, σ(·) is the cosine similarity function, and e is the natural constant.

[0065] Cosine similarity measures the directional similarity between vectors and is suitable for assessing the degree of matching between stem cell behavioral parameters (such as the migration organ set being transformed into an organ distribution vector) and reference data. For example, if the cosine value of the measured migration organ set and the reference set is close to 1, it indicates that the stem cell migration path is in line with expectations; otherwise, it suggests abnormal migration (such as aggregation towards inflammatory sites). Simultaneous consideration of survival, migration, and distribution parameters avoids misjudgment based on a single indicator. For instance, if the stem cell survival rate meets the target but the cells migrate to non-target organs, the behavioral assessment coefficient will decrease due to the low cosine value of the migration set, thus exposing potential risks.

[0066] The in vivo liver stem cell detection index is obtained by weighted summing of the comprehensive coefficient of stem cell factor concentration and the stem cell behavior evaluation coefficient: if the in vivo liver stem cell detection index is greater than the in vivo liver stem cell detection threshold set in the database, the in vivo liver stem cell detection meets the standard; if the in vivo liver stem cell detection index is not greater than the in vivo liver stem cell detection threshold set in the database, the in vivo liver stem cell detection does not meet the standard.

[0067] This testing process utilizes a combined design of "blood factors (functional metabolism) + labeling and tracing (behavioral tracking)" to achieve a comprehensive assessment of the in vivo status of liver stem cells. It integrates factor concentration and behavioral parameters into a single detection indicator, comparing it to a threshold (e.g., an indicator > 0.8 indicates compliance), thus avoiding subjective judgment. For example, when two groups of patients have similar stem cell factor concentrations but significantly different behavioral parameters, the comprehensive indicator can clearly distinguish their in vivo activity, providing a quantitative basis for clinical decision-making.

[0068] Determining whether the gut microbiota test meets the standards includes the following steps: obtaining the concentration ratio of bile acid metabolites to short-chain fatty acids, the rate of change of the relative abundance of key bacterial species (such as Akkermansia and Bifidobacterium), and the trend of change of the relative content of acetic acid, propionic acid, and butyric acid in short-chain fatty acids; if the concentration ratio of bile acid metabolites to short-chain fatty acids and the rate of change of the relative abundance of key bacterial species are both within the corresponding set range, and the trend of change of the relative content of acetic acid, propionic acid, and butyric acid in short-chain fatty acids all meet the expected trend, then the gut microbiota test meets the standards; otherwise, the gut microbiota test does not meet the standards.

[0069] Bile acid metabolism and gut microbiota are mutually regulated. Normal microbiota can convert primary bile acids into secondary bile acids, regulating glucose and lipid metabolism. SCFAs (such as butyrate) can improve insulin resistance by inhibiting bile acid reabsorption. An imbalance in the ratio (such as increased bile acids / decreased SCFAs) often indicates dysbiosis (such as a reduction in SCFA-producing bacteria), which may exacerbate insulin resistance.

[0070] During normal treatment, the abundance of key bacterial species should increase gradually (e.g., 10%-15% per week). A slow growth rate (e.g., <5%) indicates poor colonization; a rapid growth rate (e.g., >20%) may lead to excessive proliferation (e.g., gut microbiota dysbiosis). Dynamic rate monitoring can promptly detect problems such as "insufficient activation" or "abnormal proliferation" of the microbiota, avoiding fluctuations in treatment efficacy due to imbalances in microbiota dynamics.

[0071] In combination therapy, SCFAs levels should show a continuous upward trend with the treatment cycle (e.g., an increase of 8%-12% per week). If the trend stagnates or declines, it indicates weakened gut microbiota metabolic function (e.g., reduced activity of acid-producing bacteria), directly affecting the hypoglycemic effect (e.g., insufficient butyrate can exacerbate intestinal inflammation and worsen insulin resistance). Trend assessment can verify the positive regulatory effect of gut microbiota metabolites on the host in real time.

[0072] If either the in vivo liver stem cell test or the gut microbiota test fails to meet the standard, then the in vivo liver stem cell metabolism impact test and the gut microbiota metabolism impact test will be performed; if both the in vivo liver stem cell metabolism impact test and the gut microbiota metabolism impact test are normal, then diabetes-related indicators will be tested to obtain the treatment effect evaluation coefficient.

[0073] The in vivo detection of the metabolic impact of liver stem cells includes the following steps: acquiring data on the metabolic impact of stem cells, including intrinsic parameters of liver stem cells (cell viability, percentage of cell purity, telomere length, expression levels of hepatocyte differentiation markers), parameters related to stem cell metabolites (HGF concentration, IGF-1 concentration, glucokinase activity), and parameters related to overall metabolism in the body (blood glucose levels, serum insulin concentration, insulin resistance index); and performing a three-level data correlation analysis from "cell-product-body" to avoid misjudgment from a single dimension.

[0074] After normalizing the data on the impact of stem cell metabolism, a feature vector for detecting the impact of stem cell metabolism was generated. This feature vector was then input into a trained decision tree. Starting from the root node, the tree was traversed downwards according to the node partitioning conditions until a leaf node was reached. The number of samples with normal metabolism and the number of samples with abnormal metabolism were counted in each leaf node. Based on the number of samples with normal metabolism and the number of samples with abnormal metabolism, the proportions of samples with normal metabolism and abnormal metabolism were calculated. If the proportion of samples with normal metabolism was greater than the proportion of samples with abnormal metabolism, the detection of the impact of liver stem cell metabolism was considered normal, indicating that liver stem cell metabolism was normal; otherwise, it was considered abnormal.

[0075] Decision trees, through hierarchical division (e.g., "IGF-1 concentration > threshold → blood glucose decrease rate > threshold"), are highly interpretable and can clearly identify key influencing factors of metabolic abnormalities (e.g., the concentration of a certain metabolite is too low). Using the proportion of metabolically normal samples in the leaf nodes as the basis for judgment aligns with the clinical logic that "if most samples are normal, it is considered that the whole is normal," reducing the interference of individual abnormal data.

[0076] The detection of the metabolic impact of gut microbiota includes the following steps: acquiring data on the metabolic impact of gut microbiota, including gut microbiota structural parameters (relative abundance of key species, microbiota diversity index), parameters related to gut microbiota metabolites (short-chain fatty acid content, bile acid metabolite concentration, branched-chain amino acid content), and parameters related to overall metabolism (inflammation and oxidative stress indicators, intestinal function indicators); the gut microbiota affects diabetes through "metabolic products → intestinal barrier repair → systemic metabolic regulation," and the detection indicators cover microbiota structure, metabolites, and systemic inflammation / metabolic indicators, forming a complete chain of evidence.

[0077] After normalizing the data on the impact of gut microbiota metabolism, a feature vector for the detection of the impact of gut microbiota metabolism is formed. This feature vector is then input into a trained gradient decision tree model to obtain the final predicted value. The final predicted value is converted into a probability value through a logical function. If the probability value is less than the set probability threshold, the detection of the impact of gut microbiota metabolism is considered normal, indicating that gut microbiota metabolism is normal; otherwise, it is not.

[0078] The logical function is: X represents the final predicted value.

[0079] Obtaining the treatment efficacy evaluation coefficient includes the following steps: performing a change analysis on the currently acquired diabetes-related indicators and the previously acquired diabetes-related indicators to obtain the change factor bh. yz :

[0080]

[0081] Among them, Ssx iDsx represents the difference between the i-th diabetes-related indicator that exceeds the set value and its corresponding set value from the previously acquired diabetes-related indicators. i To be with Ssx i The difference between the corresponding diabetes index and the corresponding reference value, where n is the total number of diabetes-related indicators that exceeded the reference value in the previous acquisition; Sx j Ddx represents the absolute value of the difference between the j-th diabetes-related indicator below the reference value and its corresponding reference value from the previously obtained diabetes-related indicators. j To be with Sx j The absolute value of the difference between the corresponding diabetes index and the corresponding reference value, where m is the total number of diabetes indices below the reference value among the previously obtained diabetes-related indices.

[0082] By using positive and negative values ​​to distinguish between "too high" and "too low" attributes of indicators, we can avoid uniform calculations from masking the true changes.

[0083] After normalizing the currently acquired diabetes-related indicators, a cosine similarity analysis was performed between the normalized indicators and the expected normalized diabetes-related indicators to obtain the similarity factor xs. yz The method calculates the directional similarity between indicator vectors (such as blood glucose, insulin, and inflammatory factors) and expected vectors, regardless of the numerical magnitude (e.g., blood glucose mmol / L and inflammatory factors pg / mL can be uniformly assessed), focusing on the synergistic trends of changes between indicators. Expected indicators are set based on clinical criteria for diabetes remission (e.g., blood glucose <7.0 mmol / L, HOMA-IR <2.5), and similarity analysis can directly quantify "how far the treatment effect is from the target."

[0084] The system obtains current comprehensive body parameters and compares them with the comprehensive indicators in the comprehensive indicator-comprehensive risk value mapping table stored in the database using cosine similarity. It then identifies the comprehensive indicator most similar to the current comprehensive body parameters and obtains the corresponding comprehensive risk value. The current comprehensive body parameters include metabolic syndrome-related indicators, insulin secretion and sensitivity indicators, inflammation and oxidative stress indicators, kidney function indicators, and liver function indicators. The system also considers the variable factor bh. yz Similarity factor xs yz Comprehensive risk value fx The proportion of people with normal metabolism (zc) bl and probability value P gl The data is input into the treatment efficacy evaluation model to obtain the treatment efficacy evaluation coefficient ZL. xs The treatment efficacy evaluation model takes the following form:

[0085]

[0086] The steps for obtaining the comprehensive indicator-comprehensive risk value mapping table are as follows: First, obtain historical comprehensive body parameter indicators for each treatment cycle. Then, perform cluster analysis on these historical comprehensive body parameter indicators using the DBSCAN clustering algorithm to obtain the cluster centers corresponding to each treatment cycle. Next, average the historical comprehensive body parameter indicators corresponding to each cluster center to obtain the averaged historical comprehensive body parameter indicators, denoted as the comprehensive indicator. Each cluster center corresponds to one comprehensive indicator. Finally, count the number of samples where the historical comprehensive body parameter indicators corresponding to each cluster center show abnormalities in the next detection cycle, and calculate the probability of abnormality. Finally, map each comprehensive indicator to its probability of abnormality to obtain the comprehensive indicator-comprehensive risk value mapping table for each treatment cycle.

[0087] By using the DBSCAN algorithm to cluster historical comprehensive indicators, patients with similar health conditions are grouped into one category, the abnormality probability of each category is calculated, and the statistical regularity of historical data is used to map the current patient's multidimensional indicators (metabolism, inflammation, liver and kidney function) to the most similar risk category.

[0088] If either the detection of the impact of liver stem cell metabolism or the detection of the impact of gut microbiota metabolism are abnormal, then the risk of abnormal signs will be identified to determine the probability of liver risk and the probability of gut risk, and at the same time, the adjuvant treatment plan will be determined.

[0089] Determining liver and intestinal risk probabilities involves the following steps: obtaining and standardizing risk indicators for diabetic liver disease. Standardized risk indicators for diabetic liver disease include metabolic indicators and liver damage / fibrosis markers. Metabolic indicators include glycated hemoglobin (HbA1c) levels. hl Triglyceride content (GY) hl Low-density lipoprotein cholesterol (LDL-C) content DG hl Non-esterified fatty acid content ZH hl By focusing on "dysregulation of glucose and lipid metabolism," we can identify the early causes of diabetic liver disease and avoid the lagging assessment caused by focusing only on liver damage markers.

[0090] Biomarkers for liver damage and fibrosis include alanine aminotransferase (BA) levels. hl Aspartate aminotransferase content TD hl and liver stiffness value (GZ) hl It directly reflects organic liver damage and, combined with metabolic indicators, forms a complete chain of evidence of "cause-effect" (such as abnormal glucose and lipid metabolism → hepatocellular damage → fibrosis).

[0091] Based on the trained Logistic regression model, the standardized risk indicators of diabetic liver disease were processed to obtain the first probability of liver disease, P1:

[0092] P1′=β0′+β1′*TH hl +β2′*GY hl +β3′*DG hl +β4′*ZH hl ;

[0093] P1″=β0″+β1″*BA hl +β2″*TD hl +β3″*GZ hl ;

[0094]

[0095] Where β0′ and β0″ are intercepts, β1′, β2′, β3′, β4′, β1″, β2″ and β3″ are regression coefficients, and P1′ and P1″ are transit functions;

[0096] Logistic regression model results are highly interpretable, making it easy for doctors to trace the source of risk (e.g., "P1 elevation is mainly due to HbA1c exceeding the reference value by 2.3%)", and providing clear targets for individualized intervention (e.g., prioritizing enhanced blood glucose control).

[0097] The liver disease decision tree model, after being trained, is used to process the standardized risk indicators of diabetic liver disease to obtain the second probability of liver disease, P2. The liver disease decision tree model identifies the interaction between indicators through hierarchical division (e.g., "LSM>9.5kPa→ALT>40U / L"). For example, when LSM and ALT are abnormal at the same time, the risk of liver disease increases non-linearly, while Logistic regression may underestimate the risk due to the linear assumption.

[0098] The probability of first liver disease P1, the probability of second liver disease P2, and the proportion of metabolic abnormalities yc bl After processing, the probability of liver risk is obtained:

[0099] P gz = (δ1*P1+δ2*P2)*(1+yc bl ); where δ1 is the weighting factor of P1 and δ2 is the weighting factor of P2;

[0100] Logistic regression excels at handling linear relationships and its results are easy to interpret, but it may overlook higher-order interactions between indicators (such as the synergistic damage effect of LDL-C and ALT); decision trees excel at capturing nonlinear relationships, but their black-box nature makes it difficult to interpret the results (such as the fact that the basis for splitting a certain leaf node is not intuitive); fusion retains the interpretability of linear models and improves the prediction accuracy in complex scenarios through decision trees.

[0101] Diabetic gut risk indicators were acquired and standardized. These standardized indicators included gut microbiota characteristics, inflammation and oxidative stress markers, and gut function indicators. A trained gut disease decision tree model was used to process the standardized gut risk indicators to obtain a first gut disease probability. Weighted scoring rules stored in a database were then used to process the standardized gut risk indicators to obtain a total risk score, which was mapped to a second gut disease probability. Finally, the first and second gut disease probabilities were weighted and summed to obtain the gut risk probability.

[0102] Decision tree models excel at handling nonlinear relationships and feature interactions, improving the accuracy of risk prediction in complex cases (such as diabetic patients with enteritis); weighted scoring rules are based on guideline weights, and the results can be traced back to specific indicators, meeting the clinical need for "transparent evidence chain" (such as explaining to patients that "high risk is due to abnormal gut microbiota and endotoxins"); after fusion, the algorithmic advantages of decision trees are utilized, while the weighted scoring maintains consistency between the assessment logic and clinical practice.

[0103] The determination of the adjunctive treatment plan includes the following steps: Obtaining vital sign data of the diabetic patient, including height, age, body mass index (BMI), diabetes-related indicators, and current comprehensive physical parameters; comparing the vital sign data of the diabetic patient with the vital sign matching data-adjunctive treatment plan set mapping set stored in the database, determining the vital sign matching data closest to the diabetic patient's vital sign data, and obtaining the adjunctive treatment plan set, which includes multiple historical adjunctive treatment plans, each corresponding to a priority value; selecting the historical adjunctive treatment plan corresponding to the highest priority value in the adjunctive treatment plan set as the adjunctive treatment plan corresponding to the current diabetic patient's vital sign data; after setting the adjunctive treatment period, re-performing in vivo liver stem cell metabolism impact testing and gut microbiota metabolism impact testing: if both in vivo liver stem cell metabolism impact testing and gut microbiota metabolism impact testing are normal, adjunctive treatment is stopped; if either in vivo liver stem cell metabolism impact testing or gut microbiota metabolism impact testing is abnormal, the diabetic patient's vital sign data is re-obtained, the adjunctive treatment plan is re-determined, and the priority value of the current adjunctive treatment plan is updated.

[0104] The database stores a large amount of "signs-treatment-effect" data from historical cases. By matching cosine similarity (e.g., the smaller the angle between the sign vectors, the higher the similarity), the current patient can obtain a treatment plan that has been "verified by similar cases", avoiding empirical decision-making and building adjunctive treatment on the basis of historical success cases, thereby improving the effectiveness of the treatment plan.

[0105] Priority values ​​are calculated based on the treatment effects of historical protocols (such as the magnitude of blood glucose reduction and the rate of improvement in metabolic abnormalities); the better the effect, the higher the priority value. The system automatically selects the most cost-effective protocols. For example, among two protocols for similar symptoms, the protocol with the higher priority value may be selected due to "fewer side effects and longer-lasting efficacy," optimizing treatment safety and cost-effectiveness. A set duration for adjunctive therapy (e.g., 2 weeks) is established to avoid long-term use of unnecessary adjunctive therapies (e.g., excessive use of probiotics may lead to gut microbiota dependence). The criterion for discontinuing medication is "whether metabolic function has recovered," rather than a fixed treatment course, achieving "precise discontinuation" and avoiding overtreatment.

[0106] The priority value can be updated by deducting the corresponding value and updating the priority value when the detection of the impact of liver stem cell metabolism or the impact of abnormal gut microbiota metabolism occurs in vivo.

[0107] An efficacy evaluation system for the combined treatment of diabetes with liver stem cells and gut microbiota, such as Figure 2 As shown, the system includes a pre-treatment quality testing module for testing the quality of liver stem cell suspension and gut microbiota suspension to determine if the quality meets the standards. If the quality does not meet the standards, the liver stem cell suspension and gut microbiota suspension are replaced, and the pre-treatment quality testing is repeated. An in vivo testing module is used to perform in vivo liver stem cell testing and gut microbiota testing at set intervals during treatment, provided the quality of the liver stem cell suspension and gut microbiota suspension meets the standards. If both in vivo liver stem cell testing and gut microbiota testing meet the standards, treatment continues.

[0108] The metabolic impact detection module is used to detect the metabolic impact of liver stem cells and gut microbiota when either of these tests fails to meet the target. If both tests are normal, diabetes-related indicators are tested to obtain a treatment efficacy evaluation coefficient. The risk probability and auxiliary treatment plan determination module is used to identify the risk of abnormal signs, determine the liver risk probability and the gut risk probability, and determine the auxiliary treatment plan when either test is abnormal.

[0109] An electronic device includes: a processor; and a memory storing computer program instructions, which, when executed by the processor, cause the processor to perform the efficacy evaluation method described above for the combined treatment of diabetes with liver stem cells and gut microbiota.

[0110] A computer-readable storage medium for storing a program, which, when executed by a processor, implements the method described above for evaluating the efficacy of combined liver stem cell and gut microbiota therapy for diabetes.

[0111] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0112] This invention is described with reference to flowchart illustrations and / or block diagrams of systems, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0113] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0114] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0115] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0116] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for evaluating the effect of a combination therapy of liver stem cells and intestinal flora on diabetes, characterized in that, Includes the following steps: Pretreatment quality testing was performed on liver stem cell suspensions and gut microbiota suspensions to determine whether the quality met the standards. If the standards are not met, the liver stem cell suspension and gut microbiota suspension will be replaced, and the pre-treatment quality test will be repeated. If the target is met, in vivo liver stem cell testing and gut microbiota testing will be performed according to the set cycle during the treatment process to determine whether the target for in vivo liver stem cell testing and gut microbiota testing has been met. If both the liver stem cell test and the gut microbiota test are within the normal range, then the treatment will continue. If either the in vivo liver stem cell test or the gut microbiota test fails to meet the standard, then the in vivo liver stem cell metabolism impact test and the gut microbiota metabolism impact test will be performed: If the tests for the effects of liver stem cell metabolism and gut microbiota metabolism are both normal, then diabetes-related indicators will be tested to obtain a treatment efficacy evaluation coefficient. If either the detection of the impact of liver stem cell metabolism or the detection of the impact of gut microbiota metabolism is abnormal, then the risk of abnormal signs will be identified to determine the probability of liver risk and the probability of gut risk, and at the same time, the adjuvant treatment plan will be determined. To obtain a treatment efficacy evaluation coefficient, the following steps are included: A change analysis was performed on the currently acquired diabetes-related indicators compared to the previously acquired indicators to obtain the change factors. : ； in, This represents the difference between the i-th diabetes-related indicator that exceeds the set value and its corresponding set value from the previously acquired diabetes-related indicators. To and The difference between the corresponding diabetes index and the corresponding reference value, where n is the total number of diabetes indices that exceed the reference value in the previous diabetes-related indexes. This represents the absolute value of the difference between the j-th diabetes-related indicator that is below the reference value and its corresponding reference value from the previously acquired diabetes-related indicators. To and The absolute value of the difference between the corresponding diabetes index and the corresponding reference value, where m is the total number of diabetes indices below the reference value among the previously obtained diabetes-related indices. After normalizing the currently acquired diabetes-related indicators, a cosine similarity analysis was performed between the normalized indicators and the expected normalized diabetes-related indicators to obtain the similarity factor. ; The system obtains the current comprehensive body parameters and compares them with the comprehensive indicators in the comprehensive indicator-comprehensive risk value mapping table stored in the database using cosine similarity. It then determines the comprehensive indicator that is most similar to the current comprehensive body parameters and obtains the corresponding comprehensive risk value. The current comprehensive body parameters include metabolic syndrome-related indicators, insulin secretion and sensitivity indicators, inflammation and oxidative stress indicators, kidney function indicators, and liver function indicators. Change factor Similarity factor Comprehensive risk value The proportion of people with normal metabolism and probability value The data is input into the treatment efficacy evaluation model to obtain the treatment efficacy evaluation coefficient. The treatment efficacy evaluation model takes the following form: ;in, The proportion of metabolically normal individuals is calculated based on the number of samples with normal liver stem cell metabolism. To input the feature vector of the gut microbiota metabolism impact detection into the trained gradient decision tree model to obtain the final predicted value, a logistic function is used to convert the final predicted value into a probability value.

2. The method for evaluating the efficacy of combined liver stem cell and gut microbiota therapy for diabetes according to claim 1, characterized in that, Pretreatment quality testing of liver stem cell suspensions and gut microbiota suspensions to determine whether the quality meets the standards includes the following steps: Pre-treatment characteristic data of liver stem cells and gut microbiota were obtained. The pre-treatment characteristic data of liver stem cells included liver stem cell activity values. Cell purity percentage telomere length Expression levels of genes related to glucose metabolism Levels of cytokines related to insulin sensitivity Pretreatment gut microbiota characteristics data included the relative abundance of Akkermansia. Bifidobacteria relative abundance Shannon diversity index of microbial communities Simpson dominance index Short-chain fatty acid content and branched-chain amino acid content ; Pre-treatment characteristic parameters of liver stem cells and gut microbiota were obtained from the database. The pre-treatment characteristic parameters of liver stem cells included liver stem cell activity values. Cell purity percentage reference value Telomere length reference value Parameters for the expression levels of genes related to glucose metabolism Reference values ​​of insulin sensitivity-related cytokine levels Pretreatment gut microbiota characteristic data included Akkermansia relative abundance parameters. Bifidobacteria relative abundance reference value Shannon diversity index reference value Simpson dominance index parameter value Reference values ​​for short-chain fatty acid content and the reference value of branched-chain amino acid content ; Based on pre-treatment characteristic data and pre-treatment characteristic parameter data of liver stem cells, stem cell characteristic detection and stem cell metabolic secretion detection were performed to obtain stem cell characteristic evaluation coefficients. and stem cell metabolic secretion assessment coefficient The liver stem cell quality detection coefficient was obtained based on the stem cell characteristic evaluation coefficient and the stem cell metabolic secretion evaluation coefficient. ; If the liver stem cell quality detection coefficient is greater than the liver stem cell quality detection threshold set in the database, then the liver stem cell quality meets the standard. If the liver stem cell quality detection coefficient is not greater than the liver stem cell quality detection threshold set in the database, then the liver stem cell quality is substandard. Based on pre-treatment gut microbiota characteristic data and pre-treatment gut microbiota characteristic parameter data, gut microbiota characteristics and gut microbiota metabolism were detected, and gut microbiota characteristic evaluation coefficients were obtained respectively. and gut microbiota metabolic assessment coefficient The gut microbiota quality detection coefficient was obtained based on the gut microbiota characteristic assessment coefficient and the gut microbiota metabolism assessment coefficient. ; If the gut microbiota quality detection coefficient is greater than the gut microbiota quality detection threshold set in the database, then the gut microbiota quality meets the standard. If the gut microbiota quality detection coefficient is not greater than the gut microbiota quality detection threshold set in the database, then the gut microbiota quality is not up to standard.

3. The method for evaluating the efficacy of combined liver stem cell and gut microbiota therapy for diabetes according to claim 2, characterized in that, Liver stem cell quality testing coefficient The calculation formula is: ； in, for Weighting factors for Weighting factors; Gut microbiota quality detection coefficient : ； in, for Weighting factors for Weighting factors.

4. The method for evaluating the efficacy of combined liver stem cell and gut microbiota therapy for diabetes according to claim 1, characterized in that, Determining whether the in vivo liver stem cell test meets the standards includes the following steps: Based on blood tests, stem cell factor concentration parameters are obtained, including hepatocyte growth factor concentration and insulin-like growth factor-1 concentration. Based on label-based tracing technology, stem cell behavioral parameters are obtained, including the stem cell survival rate. Stem cell migration organ collection Data set on the distribution of stem cells in various organs of the body ; After standardizing the stem cell factor concentration parameters, a weighted summation was performed to obtain the comprehensive coefficient of stem cell factor concentration. Retrieve stem cell behavior parameters stored in the database, and combine these parameters to obtain stem cell behavior evaluation coefficients. ; ； in, The proportion of viable stem cells in the stem cell behavior parameters was determined. This refers to the set of stem cell migration organ parameters in stem cell behavioral parameters. This is a set of data on the distribution of stem cells in various organs of the body, used as a parameter for determining the behavioral parameters of stem cells. Let e ​​be the cosine similarity function, and e be the natural constant. The weighted sum of the stem cell factor concentration comprehensive coefficient and the stem cell behavior assessment coefficient yields the in vivo liver stem cell detection index: If the in vivo liver stem cell detection index is greater than the in vivo liver stem cell detection threshold set in the database, then the in vivo liver stem cell detection meets the standard. If the in vivo liver stem cell detection index is not greater than the in vivo liver stem cell detection threshold set in the database, then the in vivo liver stem cell detection is not up to standard. Determining whether gut microbiota testing meets the standards includes the following steps: The concentration ratio of bile acid metabolites to short-chain fatty acids was obtained, along with the rate of change in the relative abundance of key bacterial species and the trend of changes in the relative content of acetic acid, propionic acid, and butyric acid in short-chain fatty acids. If the ratio of bile acid metabolites to short-chain fatty acids and the rate of change of the relative abundance of key bacterial species are both within the corresponding set range, and the relative content of acetic acid, propionic acid, and butyric acid in short-chain fatty acids all meet the expected trend, then the intestinal flora detection meets the standard; otherwise, the intestinal flora detection does not meet the standard.

5. The method for evaluating the efficacy of combined liver stem cell and gut microbiota therapy for diabetes according to claim 1, characterized in that, The following steps were performed to detect the metabolic effects of liver stem cells in vivo: Obtain data on the impact of stem cell metabolism, including intrinsic parameters of liver stem cells, parameters related to stem cell metabolites, and parameters related to overall metabolism in the body; After normalizing the data on the effects of stem cell metabolism, a feature vector for the detection of the effects of stem cell metabolism was generated. The feature vector for detecting the impact of stem cell metabolism is input into the trained decision tree. Starting from the root node of the decision tree, the tree is traversed downwards step by step according to the node splitting conditions until the leaf node is reached. The number of samples with normal metabolism and the number of samples with abnormal metabolism in the leaf node are counted. The proportions of normal metabolism and abnormal metabolism are calculated based on the number of samples with normal metabolism and the number of samples with abnormal metabolism, respectively. If the proportion of normal metabolism is greater than the proportion of abnormal metabolism, the detection of the metabolic impact of liver stem cells in vivo is normal; otherwise, it is abnormal. The detection of the impact of gut microbiota metabolism includes the following steps: Acquire data on the impact of gut microbiota metabolism, including gut microbiota structural parameters, gut microbiota metabolite-related parameters, and overall metabolic parameters of the body; After normalizing the detection data of the impact of gut microbiota metabolism, the data were organized to form a feature vector of the impact of gut microbiota metabolism. The feature vector of the impact of gut microbiota metabolism on detection is input into the trained gradient decision tree model to obtain the final predicted value; The final predicted value is converted into a probability value through a logic function. If the probability value is less than the set probability threshold, the detection of gut microbiota metabolism is normal; otherwise, it is abnormal.

6. The method for evaluating the efficacy of combined liver stem cell and gut microbiota therapy for diabetes according to claim 5, characterized in that, The steps to obtain the comprehensive indicator-comprehensive risk value mapping table are as follows: Historical comprehensive physical parameters for each treatment cycle are obtained, and cluster analysis is performed on the historical comprehensive physical parameters based on the DBSCAN clustering algorithm to obtain the cluster centers corresponding to each treatment cycle. The historical comprehensive physical parameters corresponding to each cluster center are averaged to obtain the averaged historical comprehensive physical parameters, which are denoted as the comprehensive index. Each cluster center corresponds to one comprehensive index. The probability of anomalies is calculated by counting the number of samples whose historical comprehensive body parameters corresponding to each cluster center are abnormal in the next detection cycle. By mapping each comprehensive indicator to the probability of abnormality, a comprehensive indicator-comprehensive risk value mapping table is obtained for each treatment cycle.

7. The method for evaluating the efficacy of combined liver stem cell and gut microbiota therapy for diabetes according to claim 4, characterized in that, Determining the probability of liver and intestinal risk includes the following steps: Risk indicators for diabetic liver disease were obtained and standardized. The standardized risk indicators for diabetic liver disease included metabolic indicators and liver damage / fibrosis markers. Metabolic indicators included glycated hemoglobin levels. Triglyceride content Low-density lipoprotein cholesterol content Non-esterified fatty acid content Marker parameters for liver damage and fibrosis include alanine aminotransferase levels. Aspartate aminotransferase content and liver stiffness value ; Based on a trained logistic regression model, the standardized risk indicators for diabetic liver disease were processed to obtain the first probability of liver disease. : ； ； ； in, and All are intercepts. , , , , , and All are regression coefficients. and All are relay functions; Based on a trained liver disease decision tree model, the standardized risk indicators for diabetic liver disease are processed to obtain the second probability of liver disease. ; The probability of first liver lesions The probability of second liver lesions and the proportion of metabolic abnormalities After processing, the probability of liver risk is obtained: ； in, for Weighting factors for Weighting factors; Diabetic gut risk indicators were obtained and standardized. The standardized diabetic gut risk indicators included gut microbiota characteristics, inflammation and oxidative stress markers, and gut function indicators. The standardized intestinal risk indicators for diabetes are processed based on the trained intestinal lesion decision tree model to obtain the first intestinal lesion probability. Based on the weighted scoring rules stored in the database, the standardized diabetic intestinal risk indicators are processed to obtain a total risk score, which is then mapped to the probability of second intestinal lesions. The intestinal risk probability is obtained by weighted summation of the probabilities of the first and second intestinal lesions.

8. The method for evaluating the efficacy of combined liver stem cell and gut microbiota therapy for diabetes according to claim 1, characterized in that, Determining an adjunctive therapy plan includes the following steps: Obtain vital signs data of diabetic patients, including height, age, BIM, diabetes-related indicators, and current comprehensive physical parameters; The vital signs data of diabetic patients are compared with the vital signs matching data-auxiliary treatment plan set mapping set stored in the database to determine the vital signs matching data that is closest to the vital signs data of diabetic patients, and to obtain the auxiliary treatment plan set. The auxiliary treatment plan set includes multiple historical auxiliary treatment plans, and each historical auxiliary treatment plan corresponds to a priority value. The historical adjuvant treatment plan with the highest priority value in the set of adjuvant treatment plans will be used as the adjuvant treatment plan corresponding to the current diabetic patient's vital signs data. After setting the duration of adjuvant therapy, the effects of in vivo liver stem cell metabolism and gut microbiota metabolism on the disease were re-tested. If both the in vivo liver stem cell metabolism test and the gut microbiota metabolism test are normal, then adjuvant therapy should be discontinued. If either the detection of the impact of liver stem cell metabolism or the detection of the impact of gut microbiota metabolism are abnormal, the patient's vital signs data will be reacquired, the adjuvant therapy plan will be redefined, and the priority value of the current adjuvant therapy plan will be updated.

9. A system for evaluating the efficacy of combined liver stem cell and gut microbiota therapy for diabetes, used in the efficacy evaluation method for combined liver stem cell and gut microbiota therapy for diabetes as described in any one of claims 1-8, characterized in that, include: The pre-treatment quality testing module is used to perform pre-treatment quality testing on liver stem cell suspension and intestinal flora suspension to determine whether the quality meets the standards. If the quality does not meet the standards, the liver stem cell suspension and intestinal flora suspension are replaced, and the pre-treatment quality testing is performed again. The in vivo detection module is used to perform in vivo liver stem cell and intestinal flora tests according to a set cycle during the treatment process, provided that the quality of the liver stem cell suspension and intestinal flora suspension meets the standards, to determine whether the in vivo liver stem cell and intestinal flora tests meet the standards: if both the in vivo liver stem cell and intestinal flora tests meet the standards, the treatment continues. The metabolic impact detection module is used to detect the metabolic impact of liver stem cells and gut microbiota when either the in vivo liver stem cell detection or the gut microbiota detection fails to meet the standard. If both the in vivo liver stem cell metabolic impact detection and the gut microbiota metabolic impact detection are normal, then diabetes-related indicators are detected to obtain the treatment effect evaluation coefficient. The risk probability and auxiliary treatment module is used to identify the risk of abnormal signs when either the liver stem cell metabolism impact test or the gut microbiota metabolism impact test is abnormal, determine the liver risk probability and the gut risk probability, and determine the auxiliary treatment plan.

Images