Multi-source data fusion-based traditional Chinese medicine lung cancer prevention and treatment system construction method and system

Abstract

The invention relates to a method and a system for constructing a traditional Chinese medicine lung cancer prevention and treatment system based on multi-source data fusion, belongs to the technical field of big data, integrates multi-source information such as global evidence-based literature, real world clinical data and expert decision support, constructs an AI-driven three-layer cognitive model, develops a clinical transformation platform, and provides a traditional Chinese medicine lung cancer prevention and treatment system. The method comprises the following steps: screening dominant treatment stages through an intelligent algorithm, optimizing a diagnosis and treatment scheme, establishing a quantitative correlation model of pathogenesis key elements and tumor metabolic activities, further constructing a syndrome-metabolism knowledge graph, forming a syndrome differentiation validity verification mechanism and a prescription linkage database, and finally designing a result transformation engine. New drug research and development and nosocomial preparation conversion are driven through an evidence chain, and evaluation standards are formulated in cooperation with traditional Chinese medicine syndrome improvement and western medicine solid tumor curative effect indexes. According to the method, the bottleneck problems of data splitting, dialectical staticizing and conversion in traditional research are solved, and full-chain innovation from cognitive modeling to clinical landing is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of big data, and in particular to a traditional Chinese medicine lung cancer prevention and treatment system construction method and system based on multi-source data fusion. BACKGROUND

[0002] Current research on traditional Chinese medicine for preventing and treating lung cancer faces the challenge of insufficient integration of multi-source data. Traditional methods are difficult to systematically integrate ancient literature, clinical practice and modern scientific research results, resulting in a lack of dynamic quantitative standards for syndrome differentiation rules. Various types of evidence are scattered in independent databases, literature evidence is disconnected from real-world diagnosis and treatment, and expert experience has not formed a structured decision support, which restricts the scientific construction of the prevention and treatment system.

[0003] Existing technologies lack intelligent analysis capabilities for the evolution rules of lung cancer syndromes. Clinical syndrome differentiation mainly relies on static classification standards, which cannot associate patient physiological indicators with traditional Chinese medicine pathogenesis transformation paths in real time, making it difficult to accurately match treatment plans to different treatment stages. At the same time, the correlation mechanism between key elements of pathogenesis (such as deficiency, stasis, and toxicity) and tumor metabolic activity has not yet established a quantitative model, hindering the optimization of individualized intervention strategies.

[0004] There are multiple bottlenecks in the transformation of traditional Chinese medicine achievements. The association between syndrome type and prescription is lack of metabolomics support, and new drug research and development is limited by broken evidence chain; the evaluation of curative effect mainly depends on subjective symptom score, and there is no synergistic verification system with the standard of solid tumor curative effect in western medicine. In addition, there is a lack of objective biomarker screening mechanism in the transformation process of hospital preparations, which restricts the clinical popularization and application of prevention and treatment programs. SUMMARY

[0005] To achieve the above-mentioned purpose, the present application provides the following technical scheme:

[0006] According to the first aspect of the present application, a traditional Chinese medicine lung cancer prevention and treatment system construction method based on multi-source data fusion is claimed,

[0007] which comprises:

[0008] S1, acquiring multi-source data of a traditional Chinese medicine lung cancer prevention and treatment system, and constructing an AI-driven three-layer cognitive model based on the multi-source data;

[0009] S2, developing a clinical transformation platform based on the three-layer cognitive model, performing treatment stage screening, scheme optimization and pathogenesis quantification;

[0010] S3, establishing a syndrome-metabolism knowledge graph, outputting a syndrome validity analysis standard and a metabolism-syndrome-treatment method-prescription linkage database;

[0011] S4, constructing an achievement transformation engine, driving new drug research and development and hospital preparation transformation based on an evidence chain, and generating a curative effect evaluation standard.

[0012] Further, the multi-source data of S1 also includes:

[0013] Collecting global lung cancer TCM treatment literature evidence through evidence-based medical evaluation system, and establishing structured retrieval strategy;

[0014] The structured retrieval strategy includes standardized term library of lung cancer TCM disease name, intervention means, advantage population, and efficacy index;

[0015] Real-time collection of real-world clinical data through the wisdom cloud platform of the alliance of 100 TCM hospitals nationwide, covering four diagnostic information, syndrome differentiation types, treatment plans and efficacy evaluation data;

[0016] Obtaining decision support data through expert questionnaire and patient preference research;

[0017] S1 based on the multi-source data to build an AI-driven three-layer cognitive model, further includes:

[0018] Building a three-layer cognitive model based on deep neural networks, including:

[0019] TCM cognitive model, analyzing classic medical cases and theoretical literature, extracting four diagnostic participation and eight syndromes, zangfu, weiqiyingxue syndrome differentiation rules;

[0020] Dynamic syndrome differentiation model, correlating real-time physiological data of patients with TCM syndrome evolution path, generating individualized syndrome differentiation atlas;

[0021] Prognosis prediction model, combining genomic data and TCM syndrome characteristics, identifying macro-micro biomarker combinations of high recurrence risk groups.

[0022] Further, S2 also includes:

[0023] Intelligent screening of class advantage treatment stages through random forest algorithm, including prevention and recurrence consolidation treatment, radiotherapy and chemotherapy adjuvant therapy, drug off-target cutoff treatment, maintenance period dominant treatment, and advanced simple treatment;

[0024] Based on reinforcement learning, dynamically optimize each stage of diagnosis and treatment plan, build cancer turbidity-metabolic entropy coupling mathematical model, and quantify the critical state relationship between virtual, depressed, toxic, phlegm, turbidity, and stasis pathogenesis and cancer, metastasis, and cachexia.

[0025] Further, S3 also includes:

[0026] Establishing syndrome type-metabolite-herbal medicine association rules;

[0027] Developing syndrome differentiation validity verification process, triggering expert review when the matching degree of new enrolled patient data and knowledge graph is less than 70%.

[0028] Further, S4 also includes:

[0029] Screening the monarch drug in Qingjin Peitufang based on the characteristics of "drug off-target effect";

[0030] When determining the efficacy standard, the tumor shrinkage rate is required to be synchronous with the improvement of syndrome score.

[0031] Further, the multi-source data of the S1 further comprises:

[0032] A standardized terminology library is established, containing lung cancer TCM disease name, Chinese medicine dosage form, and western medicine treatment means terminology, a literature is graded and labeled according to evidence-based medicine level by using a Boolean logic combination retrieval strategy in the database

[0033] Real world data, four diagnostic information is collected through a hospital alliance cloud platform, including tongue image RGB value, digital features of pulse waveform, syndrome differentiation type and curative effect data;

[0034] Expert and patient data, the core pathogenesis weight is determined by multiple rounds of expert questionnaire by using the Delphi method, and the treatment preference of patients is collected through a mobile terminal APP.

[0035] Further, the AI-driven three-layer cognitive model based on the multi-source data of the S1 further comprises:

[0036] The TCM cognitive model analyzes the prescription-syndrome relationship in ancient books, and constructs a syndrome differentiation rule decision tree;

[0037] The dynamic syndrome differentiation model accesses a wearable device to monitor the respiratory rate and blood oxygen saturation of a patient in real time, and triggers re-evaluation of the syndrome type when the data is abnormal;

[0038] The prognosis prediction model integrates the CTC detection result and the dark degree of tongue image, and divides the low / middle / high recurrence risk population.

[0039] Further, the S2 further comprises:

[0040] In the treatment stage screening, the best stage is automatically matched according to the survival period extension, the quality of life improvement or the toxicity control index of the treatment target;

[0041] In the scheme optimization, a new scheme is recommended based on the similarity of historical effective cases, the prescription composition is greater than or equal to 80% coincidence degree, and the curative effect is CR.

[0042] The pathogenesis quantification detects the serum metabolite concentration by metabolomics, including the low amino acid spectrum associated with deficiency syndrome, and calculates the turbid environment metabolic disorder index.

[0043] According to the second aspect of the present application, the present application claims to protect a TCM lung cancer prevention and treatment system based on multi-source data fusion, comprising:

[0044] One or more processors;

[0045] a memory having stored thereon one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method for constructing a traditional Chinese medicine lung cancer prevention and treatment system based on multi-source data fusion.

[0046] The application relates to a method and system for constructing a traditional Chinese medicine lung cancer prevention and treatment system based on multi-source data fusion, and belongs to the technical field of big data, wherein global evidence-based literature, real-world clinical data, expert decision support and other multi-source information are integrated, an AI-driven three-layer cognitive model is constructed, a clinical transformation platform is developed, an advantage treatment stage is screened through an intelligent algorithm, a diagnosis and treatment scheme is optimized, a quantitative correlation model of key elements of a disease and tumor metabolic activity is established, a syndrome-metabolism knowledge graph is further constructed, a syndrome validity verification mechanism and a prescription-drug linkage database are formed, and finally, a result transformation engine is designed, new drug research and development and in-hospital preparation transformation are driven through an evidence chain, and a standard for evaluating the efficacy of a TCM syndrome and a Western medicine solid tumor is established. The application solves the problems of data fragmentation, static syndrome differentiation and transformation bottlenecks in traditional research, and realizes full-chain innovation from cognitive modeling to clinical landing. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 A work flow diagram of a method for constructing a traditional Chinese medicine lung cancer prevention and treatment system based on multi-source data fusion as claimed in the embodiments of the application;

[0048] Figure 2 A second work flow diagram of a method for constructing a traditional Chinese medicine lung cancer prevention and treatment system based on multi-source data fusion as claimed in the embodiments of the application;

[0049] Figure 3 A third work flow diagram of a method for constructing a traditional Chinese medicine lung cancer prevention and treatment system based on multi-source data fusion as claimed in the embodiments of the application;

[0050] Figure 4 A structure module diagram of a system for constructing a traditional Chinese medicine lung cancer prevention and treatment system based on multi-source data fusion as claimed in the embodiments of the application. DETAILED DESCRIPTION

[0051] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the application.

[0052] The terms "first", "second", "third" in the present application are only for descriptive purpose, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second", "third" can explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "a plurality of" is at least two, such as two, three, etc., unless otherwise explicitly and specifically limited. All directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present application are only used to explain the relative position relationship, movement condition, etc. between components in a certain posture (as shown in the drawings), and if the certain posture changes, the directional indications also change accordingly. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but can optionally include steps or units not listed, or can optionally include other steps or units inherent to the process, method, product or device.

[0053] Reference herein to "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment can be included in at least one embodiment of the application. The appearance of the phrase in various places in the specification does not necessarily all refer to the same embodiment, nor is it necessarily mutually exclusive of other embodiments. It is explicitly and implicitly understood that the embodiments described herein can be combined with other embodiments.

[0054] According to a first aspect of the present application, a method for constructing a traditional Chinese medicine system for preventing and treating lung cancer based on multi-source data fusion is claimed, referring to Figure 1 , comprising:

[0055] S1, acquiring multi-source data of the traditional Chinese medicine system for preventing and treating lung cancer, and constructing an AI-driven three-layer cognitive model based on the multi-source data;

[0056] S2, developing a clinical transformation platform based on the three-layer cognitive model, performing treatment stage screening, scheme optimization and pathogenesis quantification;

[0057] S3, establishing a syndrome-metabolic knowledge graph, outputting a syndrome validity analysis standard and a metabolic-syndrome-treatment method-herbal medicine linkage database;

[0058] S4, constructing a result transformation engine, driving new drug research and development and in-hospital preparation transformation based on evidence chain, and generating an efficacy evaluation standard.

[0059] Further, the multi-source data of S1 further comprises:

[0060] To collect the evidence of lung cancer treatment with traditional Chinese medicine (TCM) from the global literature through the evidence-based medicine evaluation system, and to establish a structured retrieval strategy.

[0061] The structured retrieval strategy includes a standardized term library for TCM disease names, intervention methods, target populations, and efficacy indicators.

[0062] Real-world clinical data are collected in real time through the smart cloud platform of a nationwide alliance of 100 TCM hospitals, covering four diagnostic information, syndrome differentiation types, treatment plans, and efficacy evaluation data.

[0063] Decision support data are obtained through expert questionnaires and patient preference surveys.

[0064] In this embodiment, real-world data collection specifically includes:

[0065] Intelligent four-diagnosis equipment (tongue analysis instrument, pulse diagnosis instrument) is deployed to collect 17-dimensional quantitative characteristics of tongue color / texture / thickness and pulse frequency / rhythm. A dynamic medical record structured template is constructed to forcibly associate the "theory-method-prescription-drug-effect" data chain. Blockchain technology is used to achieve multi-center data authentication and privacy protection.

[0066] Tongue image collection: High-resolution camera (24 million pixels) is used to capture tongue RGB values, and color space conversion is used to analyze the color difference between yellowish fur (b>15) and whitish fur (b<5). Medical record structuring: Mandatory fields include: main symptom severity classification (coughing blood is classified as blood in sputum / full mouth of fresh blood), prescription composition (accurate to single herb dosage), and efficacy evaluation time window (14±2 days after treatment). Data security: Blockchain nodes are set up on local servers in each hospital, and zero-knowledge proof technology is used to verify data authenticity.

[0067] The S1 AI-driven three-layer cognitive model based on the multi-source data further includes:

[0068] The three-layer cognitive model is constructed based on deep neural networks, including:

[0069] TCM cognitive model: Analyzing classic medical records and theoretical literature, extracting four-diagnosis combined with eight principles, zang-fu organs, wei-qi-yin-xue syndrome differentiation rules;

[0070] Dynamic syndrome differentiation model: Correlating real-time physiological data of patients with TCM syndrome evolution path to generate individualized syndrome differentiation atlas;

[0071] Prognosis prediction model: Fusing genomic data and TCM syndrome characteristics to identify macro-micro biomarker combinations for high-risk recurrence groups.

[0072] In this embodiment, the traditional Chinese medicine cognitive model is constructed by using a BERT-BiLSTM hybrid neural network to analyze the semantics of ancient books and extract the pathogenesis evolution rules under the treatment principle of "eliminating turbidity and promoting the original"; a graph convolution network (GCN) is used to establish a three-dimensional mapping relationship among symptoms, syndromes and pathogenesis, and identify the conversion threshold of phlegm-heat obstructing lung syndrome to lung-spleen qi deficiency syndrome.

[0073] The dynamic syndrome differentiation model includes: establishing a quantitative index system of "deficiency in origin and excess in superficiality" pathogenesis; "deficiency in origin" dimension: serum albumin level + CD4+ / CD8+ ratio + qi deficiency scale score; "excess in superficiality" dimension: circulating tumor cell (CTC) count + IL-6 level + blood stasis syndrome score; predicting the syndrome evolution path through an LSTM time series model.

[0074] Ancient book analysis: first, perform word segmentation on classics such as "Synopsis of Prescriptions of the Golden Chamber", and extract high-frequency symptom-drug combinations (such as "coughing and gasping" appearing more than 50 times is marked as a strong association); pathogenesis evolution: establish early warning indicators for the conversion of phlegm-heat obstructing lung syndrome to lung-spleen qi deficiency syndrome (such as a decrease of more than 10% in CD4+ cells for three consecutive tests).

[0075] Further, the S2 further comprises:

[0076] Intelligently screen the class advantage treatment stage through a random forest algorithm, including prevention and recurrence consolidation treatment, radiotherapy and chemotherapy adjuvant treatment, drug off-target cutoff treatment, maintenance period dominant treatment, and late simple treatment.

[0077] Based on reinforcement learning, dynamically optimize each stage of diagnosis and treatment, construct a cancer turbidity-metabolic entropy coupling mathematical model, and quantify the critical state relationship between deficiency, stagnation, toxin, phlegm, turbidity, and stasis pathogenesis and cancer, metastasis, and cachexia.

[0078] Further, the S3 further comprises:

[0079] Establish the correlation rules of syndrome type-metabolite-prescription;

[0080] Develop a verification process for syndrome differentiation effectiveness, and trigger expert review when the matching degree of new patient data with the knowledge graph is less than 70%.

[0081] In this embodiment, the method for constructing the syndrome-metabolic knowledge graph is:

[0082] Node definition: syndrome type (lung-spleen qi deficiency), treatment method (cultivating the earth to generate gold), prescription (Sijunzi Decoction), metabolite (tryptophan); edge relationship: including "inhibition / activation", "biotransformation", and "clinical significant effect" three types of weighted edges; and applying a GAT network to predict unverified "prescription-metabolite" associations.

[0083] Node relationship definition:

[0084] "suppression" relationship: Sijunzi Decoction → reduce IL-6 expression (evidence level A); "activation" relationship: Huangqi → enhance CD8+ cell activity (evidence level B); new association prediction: when a certain prescription (such as Xiaoyan Decoction) and a metabolite (lactic acid) appear in > 30% of effective cases at the same time, the "regulation of glycolysis" edge relationship is automatically established.

[0085] Further, the S4 further comprises:

[0086] Based on the "drug off-target effect" feature, the monarch drug in Qingjin Peitui Decoction is screened;

[0087] When determining the efficacy standard, it is required that the shrinkage rate of solid tumors and the syndrome score improve synchronously.

[0088] Further, the multi-source data of the S1 further comprises:

[0089] A standardized terminology library is established, including the TCM disease name, Chinese medicine dosage form, and western medicine treatment method terminology. The database uses a Boolean logic combination retrieval strategy and classifies literature according to evidence-based medicine levels.

[0090] Real-world data is collected through a hospital alliance cloud platform, including tongue image RGB values, digital features of pulse waveform, syndrome differentiation types, and efficacy data.

[0091] Expert and patient data: use the Delphi method to determine the core pathogenesis weight through multiple rounds of expert questionnaires, and collect patient treatment preferences through a mobile APP.

[0092] Further, with reference to Figure 2 , the AI-driven three-layer cognitive model based on the multi-source data of the S1 further comprises:

[0093] TCM cognitive model analyzes the prescription-syndrome relationship in ancient literature and constructs a syndrome differentiation rule decision tree.

[0094] The dynamic syndrome differentiation model accesses wearable devices to monitor patients' respiratory rate and blood oxygen saturation in real time, and triggers re-evaluation of the syndrome type when the data is abnormal.

[0095] The prognosis prediction model integrates CTC detection results and tongue image purple degree to divide low / medium / high recurrence risk groups.

[0096] Further, with reference to Figure 3 , the S2 further comprises:

[0097] During the treatment stage screening, the best stage is automatically matched according to the survival period extension, quality of life improvement, or toxicity control indicators of the treatment target.

[0098] When optimizing the scheme, based on the similarity of historical effective cases, the prescription composition is ≥80% coincidence and the efficacy is CR, and a new scheme is recommended.

[0099] Pathogenesis quantification detects serum metabolite concentration by metabolomics, including low amino acid spectrum associated with deficiency syndrome, and calculates turbidity environment metabolic disorder index.

[0100] Among them, the construction and application of TCM evidence system for preventing and treating lung cancer are carried out in the embodiments of the present application.

[0101] Firstly, the literature evidence database is constructed for the implementation of the multi-source evidence collection system;

[0102] The term standardization is implemented, and the term mapping table is established, as shown in Table 1.

[0103] Table 1 Term mapping table

[0104] TCM terminology Standardized coding Western medicine corresponding terms Pulmonary accumulation LC_TCM01 Lung cancer Xiao Yan granules HERB_XY01 Institutional preparation batch number Z2023001

[0105] Hierarchical retrieval:

[0106] Summaries library (UpToDate): search "Traditional Chinese Medicine for NSCLC Resistance", and extract the guideline recommended scheme;

[0107] PubMed search formula: ("TCM" OR "Chinese herbal medicine") AND ("NSCLC" OR "lung cancer") AND ("drug resistance" OR "TKI failure");

[0108] Evidence evaluation: two researchers independently evaluate the authenticity of the literature, and the third senior researcher arbitrates the disputed literature, and the Kappa value is 0.92.

[0109] The real world data collection includes as shown in Table 2.

[0110] Table 2 Intelligent device deployment table

[0111]

[0112] The structured template is as follows:

[0113] markdown

[0114] ## Electronic medical record (example)

[0115] Patient ID: LC2023-0471

[0116] -**Main symptoms**: cough (Ⅲ degree: more than 10 times during the day);

[0117] - Tongue: greasy fur index 0.62 (normal <0.4);

[0118] - Prescription: Xiaoyan Decoction (Huangqi 30g -> IGF-1R activator);

[0119] - Efficacy: CTC 5 -> 3 per 3.5 ml at day 14 (40% decrease);

[0120] Blockchain storage: Each hospital node generates a data hash value (SHA-256 algorithm), and the consortium chain synchronizes consensus every 10 minutes, with a data tampering rate of 0.01%.

[0121] Afterwards, for the construction and validation of the AI model, the TCM cognitive model training was performed, data preprocessing was executed, and ancient books were digitized. A total of 1,920 copies of "Clinical Guide to Medical Cases" were scanned, with an OCR recognition accuracy of 98.7%, and key relationship extraction charts and codes were obtained.

[0122] Model validation was completed, and in 320 new patients, the model's syndrome differentiation accuracy was 89.25% consistent with the expert group's diagnosis.

[0123] A dynamic syndrome differentiation model was implemented, and a monitoring-intervention closed loop was established, as shown in Table 3.

[0124] Table 3 Abnormal Response Table

[0125] Physiological indicators Abnormal threshold System response Blood oxygen saturation <93% for 2 h Increase the dosage of Huangqi to 45g + start oxygen therapy CD4+ / CD8+ <0.8 twice in succession Injection of thymalfasin + adjust the proportion of tonifying drugs

[0126] Case Tracking:

[0127] Patient Zhang (ID LC2024-112) on the 15th day of treatment:

[0128] Shortness of breath episodes increased from 8 to 12 times per day, exceeding the threshold by 20%;

[0129] The system triggered an early warning and pushed a new plan: Sijunzi Decoction with a 20% increase + acupoint sticking at the Danzhong point;

[0130] After 7 days of intervention, the symptoms decreased to 9 times per day, with a decrease of 25%.

[0131] Running the clinical transformation platform, the metabolic entropy model was applied, and the metabolite-disease mechanism mapping was performed, as shown in Table 4.

[0132] Table 4 Metabolite-Disease Mechanism Mapping Table

[0133] TCM pathogenesis Metabolic markers Detection methods Deficiency syndrome Total serum essential amino acids LC-MS / MS Blood stasis syndrome [CAT] TXB2 / PGI2 ratio ELISA

[0134] Critical state management defines an entropy threshold, H≥5.8 as the "tumor" critical state, with a sensitivity of 82.3% and a specificity of 76.5%; Clinical case: Patient Li's entropy value is 6.1→ the system pushes the cutoff treatment plan, Xingpi Huatuo Decoction + CTC clearance therapy→ 3 weeks later, the entropy value drops to 4.9, and PET-CT shows no new metastatic lesions.

[0135] Treatment stage optimization builds a multi-objective decision engine:

[0136] markdown

[0137] ###Optimization decision examples

[0138] Input parameters:

[0139] - EGFR 19del mutation (+);

[0140] - Syndrome type: phlegm-heat obstructing lung syndrome (score 58 points);

[0141] - Metabolic entropy: 4.5;

[0142] Output results:

[0143] **Preferred stage**: Targeted drug off-target cutoff treatment, expected value 0.82;

[0144] - Expected OS: 12.8 months;

[0145] - Recommended regimen: Qingjin Peituo Decoction + Osimertinib increment;

[0146] **Alternative stage**: Radiotherapy and chemotherapy synergistic detoxification, expected value 0.73;

[0147] Efficacy comparison: In the optimization group (n=102), the median PFS was 8.3 months vs. 6.1 months in the experience group (n=98) (HR=0.69, p=0.003).

[0148] For knowledge base and achievement transformation, build syndrome-metabolism knowledge graph, the key relationship in the graph construction process is lung and spleen qi deficiency Abnormal tryptophan metabolism, OR=4.2, p<0.001; Four Gentlemen Decoction ↓ IL-6 expression, dose-effect R 2 =0.76.

[0149] The core index collection specification is shown in Table 5:

[0150] Table 5 Core Index Collection Specification

[0151] Indicators Measurement methods Time window Syndrome score Cough frequency x degree coefficient Treatment day 14±2 CTC count 7.5ml EDTA anticoagulated blood Morning 9:00±30min

[0152] The implementation of new drug research and development Qingjin Peituo Decoction optimization path is shown in Table 6:

[0153] Table 6 New drug development optimization path table

[0154] R&D stage Composition adjustment Key findings Basic formula Huangqi 20g + Shashen 10g Disease control rate 41.2% Optimized formula Huangqi 30g + Shashen 15g DCR↑ to 58.3% (p=0.02)

[0155] Mechanism verification: 30g Huangqi increased IGF-1R phosphorylation level by 3.1 times (Western blot)

[0156] System terminal deployment of doctor decision interface:

[0157] markdown

[0158] [Patient LC2024-215] Metabolic entropy dashboard:

[0159] - Current value: 5.7 (critical threshold 5.8);

[0160] - Warning prompt: **0.1 away from turbidity tumor critical state;

[0161] - Recommended intervention: Xingspi Huatong Decoction 200ml bid x 7 days

[0162] This embodiment has been verified by 3 national TCM clinical research bases, which proves that:

[0163] Differential diagnosis accuracy: Model diagnosis and expert group Kappa value 0.86 (better than traditional method 0.62);

[0164] Conversion efficiency: New drug development cycle shortened by 40% (36 months → 22 months);

[0165] Clinical benefit: Median OS of advanced lung cancer prolonged by 4.2 months (15.1 vs 10.9 months, p=0.008); All experimental data are from registered clinical research (ChiCTR2300078123), in line with the CONSORT statement.

[0166] According to the second embodiment of the present application, the present application claims to protect a TCM lung cancer prevention and treatment system based on multi-source data fusion, which refers to Figure 4 , comprising:

[0167] One or more processors;

[0168] A memory having one or more programs stored thereon, when the one or more programs are executed by the one or more processors, so that the one or more processors implement the one kind of TCM lung cancer prevention and treatment system based on multi-source data fusion.

[0169] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other manners. For example, the division of the units is only a logical function division, and there can be another division manner for the actual implementation, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between different units, or the among different units, can be indirect couplings or communication connections through some interfaces, devices or units, and can be in electrical, mechanical or other forms.

[0170] In addition, each function unit in the various embodiments of the present application can be integrated into a processing unit, or each unit can exist alone physically, or two or more units can be integrated into one unit. The integrated unit can be implemented in the form of hardware, or in the form of a software functional unit. The above is only an embodiment of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation using the content of the present application specification and drawings, or direct or indirect application in other related technical fields, are also included in the patent protection scope of the present application.

[0171] The specific embodiments of the application are described in detail above, but they are only examples. The present application is not limited to the specific embodiments described above. Any equivalent modification or substitution made by those skilled in the art to the present application is also within the scope of the present application. Therefore, any equivalent transformation and modification, improvement, etc. made without departing from the spirit and principle range of the present application should be included in the scope of the present application.

Claims

1. A method for constructing a traditional Chinese medicine system for preventing and treating lung cancer based on multi-source data fusion, characterized in that, Comprise: S1, acquire multi-source data of traditional Chinese medicine for preventing and treating lung cancer, and construct an AI-driven three-layer cognitive model based on the multi-source data; S2, develop a clinical transformation platform based on the three-layer cognitive model, perform treatment stage screening, scheme optimization, and pathogenesis quantification; S3, establish a syndrome-metabolism knowledge graph, output syndrome validity analysis standards, and a metabolism-syndrome-treatment method-herbal medicine linkage database; S4, construct a result transformation engine, drive new drug research and development and hospital preparation transformation based on evidence chains, and generate efficacy evaluation standards.

2. The method according to claim 1, wherein, The multi-source data of S1 further comprises: Collect global lung cancer traditional Chinese medicine treatment literature evidence through an evidence-based medicine evaluation system, and establish a structured retrieval strategy; The structured retrieval strategy includes a standardized terminology library of lung cancer traditional Chinese medicine disease names, intervention methods, target populations, and efficacy indicators; Real-time collection of real-world clinical data through a smart cloud platform of a nationwide alliance of 100 traditional Chinese medicine hospitals, covering four diagnostic information, syndrome typing, treatment schemes, and efficacy evaluation data; Obtain decision support data through expert questionnaires and patient preference surveys; The AI-driven three-layer cognitive model based on the multi-source data of S1 further comprises: Construct a three-layer cognitive model based on a deep neural network, including: A traditional Chinese medicine cognitive model that analyzes classic medical cases and theoretical literature, extracts four diagnostic information, and identifies syndrome differentiation rules of eight principles, zang-fu organs, wei, qi, ying, and blood; A dynamic syndrome differentiation model that correlates real-time physiological data of patients with the evolution path of traditional Chinese medicine syndromes, and generates individualized syndrome differentiation graphs; A prognosis prediction model that integrates genomic data and traditional Chinese medicine syndrome characteristics to identify macro-micro biomarker combinations for high recurrence risk populations.

3. The method according to claim 1, wherein, S2 further comprises: Intelligent screening of treatment stages through a random forest algorithm, including prevention and recurrence consolidation treatment, radiotherapy and chemotherapy adjuvant treatment, drug off-target cutoff treatment, maintenance period dominant treatment, and advanced stage simple treatment; Dynamic optimization of diagnosis and treatment schemes at each stage based on reinforcement learning, construction of cancer turbidity-metabolic entropy coupling mathematical models, and quantification of the critical state relationship between virtual, depressed, toxic, phlegm, turbid, and blood stasis pathogenesis and cancer, metastasis, and dysbolism.

4. The method according to claim 1, wherein, S3 further comprises: Establishing syndrome-metabolite-herbal medicine correlation rules; Developing a syndrome validity verification process, and triggering expert review when the matching degree of new group patient data with the knowledge graph is less than 70%.

5. The method according to claim 1, wherein, S4 further comprises: Screening the main drug in Qingjin Peitu Decoction based on the characteristics of "drug off-target effect"; The efficacy standard requires synchronous improvement of solid tumor shrinkage rate and syndrome score when determining the efficacy standard.

6. The method according to claim 2, wherein, The multi-source data of S1 further comprises: Establishing a standardized terminology library containing lung cancer traditional Chinese medicine disease names, traditional Chinese medicine dosage forms, and western medicine treatment methods, using Boolean logic combination retrieval strategies in the database, and grading literature according to evidence-based medicine levels Real-world data, including tongue image RGB values, digital features of pulse waveform, syndrome typing, and efficacy data, are collected through a hospital alliance cloud platform; Expert and patient data, using the Delphi method to determine the core pathogenesis weight through multiple rounds of expert questionnaires, and collecting patient treatment preferences through a mobile APP.

7. The method according to claim 2, wherein, The AI-driven three-layer cognitive model based on the multi-source data of S1 further comprises: TCM cognitive model analyzes the relationship between prescription and syndrome in ancient literature, and builds a decision tree of syndrome differentiation rules; The dynamic syndrome differentiation model accesses wearable devices to monitor the patient's respiratory rate and blood oxygen saturation in real time, and triggers a re-evaluation of the syndrome when the data is abnormal; The prognosis prediction model integrates the CTC detection results and the dark degree of tongue appearance to divide the low / medium / high recurrence risk population.

8. The method according to claim 3, characterized in that, The S2 further comprises: During the screening of the treatment stage, the best stage is automatically matched according to the indicators of life extension, quality of life improvement, or toxicity control of the treatment target; When optimizing the scheme, a new scheme is recommended based on the similarity of historical effective cases, with a prescription composition of ≥80% coincidence and a CR efficacy; Pathogenesis quantification detects serum metabolite concentration through metabolomics, including low amino acid spectrum associated with deficiency syndrome, and calculates the metabolic disorder index in turbid environment.

9. A traditional Chinese medicine system for preventing and treating lung cancer based on multi-source data fusion, characterized in that, It comprises: One or more processors; A memory having one or more programs stored thereon, when the one or more programs are executed by the one or more processors, so that the one or more processors implement a TCM lung cancer prevention and treatment system construction method based on multi-source data fusion according to any one of claims 1-8.

Images