Application technology for screening and early auxiliary diagnosis of lung cancer high-risk group and monitoring postoperative treatment effect of lung cancer patient by model and computing power

Abstract

The invention provides an application technology for screening lung cancer high-risk groups, early auxiliary diagnosis and postoperative treatment effect monitoring of lung cancer patients through model and computing power, and belongs to the fields of molecular biotechnology, medical diagnosis and treatment and massive health. The method comprises the following steps: collecting samples of human body liquid, exosomes and the like, detecting the DNA methylation level, miRNA expression quantity, protein level and the quantity of molecular markers such as autoantibodies, exosomes, kits and the like related to the lung cancer by virtue of a modern detection technology, and obtaining individual precision medical treatment and clinical application by virtue of a model and computing power (including a software algorithm, a DeepSeek large model, intelligent agent application and an artificial intelligence method). The problems of lung cancer high-risk group screening, early auxiliary diagnosis, pathological typing, clinical staging, metastatic state, postoperative curative effect evaluation and poor accuracy of national approved kits are solved. The non-destructive detection technology can prevent and treat lung cancer in time, is precise in medical treatment and good in health prevention, and has wide social benefits and popularization and use values. The traditional Chinese medicine composition also has a popularization effect on other non-lung cancers.

Description

Technical Field

[0001] The present invention relates to the combination of the computing power of large models and small models with molecular bioinformatics, preventive medicine, biological detection technology, medical diagnostics, and cancer treatment evaluation for screening high-risk populations, early auxiliary diagnosis, and postoperative treatment monitoring and evaluation of various cancers, belonging to the technical fields of big health, preventive medicine, clinical diagnostics, medical imaging, clinical treatment efficacy evaluation, molecular bioinformatics, and biotechnology. Background Art

[0002] Lung cancer is a major malignant tumor that is prone to occur in people globally. The early clinical symptoms of lung cancer are mostly hidden, making the diagnosis difficult and easily missing the best treatment opportunity. It is reported that approximately 85% of patients are already in the advanced stage when diagnosed, and the malignant tumor has spread and metastasized, with a five-year survival rate of less than 20% and a very poor prognosis. Lung biopsy and pathological examination are the gold standards for diagnosis, but these are all invasive detections, and patients suffer greatly during the pre-operative detection. Therefore, determining which patients should undergo lung biopsy and pathological examination remains a new challenge; Imaging diagnostic detection has the advantage of detecting lesion occupancy of tumor lesions and nodules, but there is a significant qualitative difference from the molecular biological method of the gold standard diagnosis of lung cancer. Additionally, it is affected by the subjective experience of radiologists; The sensitivity and specificity of molecular marker detections such as lung cancer-related kits used clinically are not high. There has been some progress in the research on molecular markers related to DNA methylation levels, protein levels, miRNA expression levels, and immune index levels in lung cancer. However, there are great difficulties in selecting molecular markers and optimizing a group of molecular markers; With the rapid development of large models, big data, artificial intelligence, digitalization, and the innovative application of DeepSeek and intelligent agents, the results of a kit and multiple molecular markers cannot be combined with model and computing power methods for comprehensive and accurate diagnosis and treatment. To overcome the above deficiencies, a series of integrated models and computing power (or DeepSeek large models, intelligent agent applications and innovations, or artificial intelligence methods) are used to empower the calculation of lung cancer-related molecular markers to achieve the screening of high-risk populations for lung cancer, early auxiliary diagnosis, and postoperative treatment monitoring of patients. Such research on lung cancer applications has not been formally reported at home and abroad. Such research can improve accuracy, prevent over-medical treatment, achieve precision medicine, and achieve the purpose of screening high-risk populations, early detection, early auxiliary diagnosis, and guiding the evaluation of postoperative treatment effects through non-invasive detection. Summary of the Invention

[0003] The content of the present invention aims to overcome the deficiencies in the applications in biotechnology, medical diagnosis and treatment, preventive medicine, and big health. By non-invasively or minimally invasively collecting samples from human fluids, molecular markers such as gene methylation levels, miRNA expression levels, protein levels, or immune index levels (including test kits) are detected using biological and medical detection technologies, and molecular markers (including test kits) of any combination or preferred combination among them are detected. Then, the results are calculated using model and computing power application technologies (including the DeepSeek large model, agent application and innovation, or artificial intelligence methods) to improve the accuracy of screening for high-risk lung cancer populations, early auxiliary diagnosis, and postoperative treatment monitoring of patients. Through early screening, early diagnosis, and treatment, the cure rate of lung cancer patients is increased, the best opportunities for early screening, early detection, and not missing early treatment are seized, lung cancer patients are saved from the brink of death, and the quality of life is improved.

[0004] The content of the present invention lies in evaluating the method for determining negative and positive in the test kit for any combination of multiple lung cancer markers approved by the state; expanding the molecular marker donation system to the marker index system of lung imaging features composed of lung cancer imaging indicators, nodules, masses, and lesion occupancies in imaging; achieving the purpose of lung cancer diagnosis and classification recognition through model and computing power empowerment; solving the problems of lung cancer classification recognition, lung cancer pathological typing, and lung cancer clinical staging by establishing and constructing linear or non-linear functions and combining them with the lung cancer classification recognition problem, model, and computing power; pre-using the model and algorithm calculated by the model and computing power (or the table for identifying lung cancer by molecular markers), loading it into a portable device, microcomputer, mobile phone, online background and platform, or proprietary device, etc. through software compilation, and inputting the test results (data) of any individual (person) into the portable device, microcomputer, mobile phone, online background and platform, or proprietary device, etc. through any means, and then outputting the individual calculation results and medical conclusions, etc. after calculation.

[0005] The subject matter of the content of the present invention not only includes an application technology for screening high-risk lung cancer populations, early auxiliary diagnosis, and monitoring the postoperative treatment efficacy of lung cancer patients with model and computing power empowerment for lung cancer, but also completely extends the approach of lung cancer to various other cancers in oncology. For any other cancers, the application technology for screening high-risk populations of other cancers, early auxiliary diagnosis, and monitoring the postoperative treatment efficacy of other cancer patients with model and computing power empowerment is completely similarly promoted. Detailed implementation manners

[0006] Implementation case: By reasonably selecting seven related gene biomarkers (X, Y, Z, W, U, V, N, and M) involved in the occurrence and development of lung cancer, high-risk population A1 for lung cancer, medium-risk population A2 with precancerous lesions, and risk-free healthy population A3 are set. For populations A1 and A2, pathological biopsies and pathological diagnoses are performed in advance, and other methods are used to prove that they are lung cancer or precancerous lesion populations. Healthy population A3 is a healthy population proven by physical examination in a physical examination center. Experiments are conducted on 717 lung cancer patients, 173 precancerous lesion patients, and 92 healthy people using biotechnology and medical tests to obtain experimental result data of the seven related gene molecular biomarkers. Based on a large amount of experimental result data, various event probability databases are established using statistics, such as:

[0007] p(A1)=0.730143, p(A2)=0.176171, p(A3)=0.09368, p(X11|A1)=0.707113, p(Y21|A1)=0.3834, p(Z31|A1)=0.482566, p(W41|A1)=0.783882, p(U51|A1)=0.28020 and other values in the probability database. Now there is a 67-year-old woman with severe pain in the lungs. Her father died of lung cancer. She is willing to undergo lung cancer risk screening (or willing to have lung cancer assisted diagnosis in the hospital). By drawing peripheral blood and through biotechnology and medical tests, the first molecular biomarker is equal to 64, the second molecular biomarker is equal to 18, the third molecular biomarker is equal to 7.2, the fourth molecular biomarker is equal to 2.52, the fifth molecular biomarker is equal to 1.5, the sixth molecular biomarker is equal to 69, and the seventh molecular biomarker is equal to 50. Assume that the actual measured data of the seven molecular markers are represented by X11, Y21, Z31, W41, U51, N61, and M71 respectively, and let the individual test result of this woman be represented by B. Denote the symbol B = X11Y21Z31W41U51N61M71. Using the method of a Bayes model in the model and computing power application technology involved in the present invention,

[0008] p(B|A1)=p(X11|A1)×p(Y21|A1)×p(Z31|A1)×p(W41|A1)×p(U51|A1)p(N61|A1)p(M71|A1)=0.031902. Similarly,

[0009] p(B|A2)=p(X11|A2)×p(Y21|A2)×p(Z31|A2)×p(W41|A2)×p(U51|A2)p(N61|A1)p(M71|A1)=0.000000,

[0010] p(B|A3) = p(X11|A2) × p(Y21|A3) × p(Z31|A3) × p(W41|A3) × p(U51|A3)p(N61|A1)p(M71|A1) = 0.000000,

[0011] P(A1|B) = (p(A1) × p(B|A1)) = 0.999999

[0012] P(A2|B) = (p(A2) × p(B|A2)) = 0.000000

[0013] P(A3|B) = (p(A3) × p(B|A3)) = 0.000000

[0014] Max{0.999999, 0.000000, 0.00000} = 0.999999, and the result belongs to the high-risk group A1 for lung cancer.

[0015] From the perspective of lung cancer risk screening, please ask this woman to go to a top-three hospital immediately for further clinical diagnosis and confirmation; from the perspective of clinical assistant diagnosis, please ask this woman to immediately undergo a pathological biopsy test. The final result is proven by the intraoperative histopathological biopsy result of a certain top-three hospital, indicating that this woman has lung cancer. The result is consistent with that calculated by the model and computing power application technology.

[0016] From the perspective of clinical assistant diagnosis and confirmation of lung cancer and lung cancer risk screening, for high-risk individuals of lung cancer, and considering that this woman currently has severe pain in the lungs and her father died of lung cancer, this patent clinically assisted in diagnosing and confirming that this woman has lung cancer. The final result is that she has lung cancer as shown by the intraoperative histopathological biopsy result of a certain top-three hospital, which is consistent with the result of lung cancer diagnosis calculated by the model and computing power application technology.

Claims

1. An application technology for a model and computing power to empower the screening of high-risk lung cancer populations, early auxiliary diagnosis, and the monitoring of the treatment efficacy of lung cancer patients after surgery, characterized in that, Using existing biological and medical detection technologies, detect the data of any combination of molecular markers in the following basic molecular marker index system; use the following basic large model or small model algorithms to calculate the data of any combination of molecular markers, achieve the following basic target classification and recognition problems, and thereby improve the accuracy of non-invasive lung cancer high-risk population screening, early auxiliary diagnosis, and postoperative regular examination of lung cancer patients for evaluating the therapeutic effect, and realize personalized precision medicine in these three aspects. The basic target classification and recognition problems involved in its characteristics are as follows: The basic target classification and recognition problems are composed of the following two-category, three-category, and four-category and above classification and recognition problems. Its two-category classification (category A and category B) recognition problems: That is, the positive and negative of nodules (or masses or lesions) (malignant and benign), lung cancer (or high risk) and pre-lung cancer lesions (including pulmonary tuberculosis, pneumonia, heavy smoking, high-risk occupational exposure history, chronic obstructive pulmonary disease, diffuse pulmonary fibrosis, lymphadenitis 1, hamartoma, those with a history of malignant tumors or a family history of lung cancer, or lung diseases composed of any two or more of them) (pre-lung cancer lesions can also be called risk categories or medium-risk categories), risk categories (medium-risk categories) respectively with the clinical stage of lung cancer, early-stage lung cancer and risk categories, recurrence and non-recurrence after lung cancer surgery, the improvement or deterioration of the treatment of patients after lung cancer surgery (improvement or deterioration means comparing the development trends of one or any combination of two or more molecular marker indicators for treatment and efficacy evaluation), with lymph node metastasis and without lymph node metastasis, metastasis and non-metastasis before lung cancer surgery, lung cancer and other cancers, small cell lung cancer and adenocarcinoma of the lung, small cell lung cancer and squamous cell carcinoma of the lung, adenocarcinoma of the lung and squamous cell carcinoma of the lung, risk categories (medium-risk categories) respectively with any one of T, N, M in the lung pathological classification, risk categories (medium-risk categories) and no risk (low risk or normal healthy population), etc. Its three-category classification (category A, category B, category C) recognition problems: Category A and category B respectively represent the above-mentioned meanings, and category C can be expanded to distinguish category A and category B, such as category C being a normal healthy person or a normal healthy mild lung disease, etc. (respectively called no risk or low risk). The three-category classification recognition problem is a high-risk category, medium-risk category (risk category), low-risk (or no risk) recognition problem; The three-category recognition problems of category A, category B, and category C include two-category cross-recognition problems of category A and category B, category A and category C, and category B and category C; It also includes that A is with lymph node metastasis, B is without lymph node metastasis, and C is a medium-risk category (risk category); C being a medium-risk category (risk category) and the clinical stage of lung cancer are divided into two categories; The three-category recognition problems of small cell lung cancer, adenocarcinoma of the lung, and squamous cell carcinoma of the lung in pathological classification, etc. Its four-category and above (category A, category B, category C, category D, etc.) recognition problems: Category A, category B, and category C respectively represent the above-mentioned meanings, and category D can be expanded to be a classification that distinguishes category A, category B, and category C; For example, if D is a medium-risk category (risk category), and categories A, B, and C are the recognition problems of small cell lung cancer, adenocarcinoma of the lung, and squamous cell carcinoma of the lung, or if D is a no-risk category, C is a medium-risk category, B is stage 2 and stage 1 of the clinical stage of lung cancer, and A is stage 3 and stage 4 of the clinical stage of lung cancer for the recognition problems of categories A, B, C, and D; Another example is that A, B, and C respectively represent the meanings of high, medium, and low differentiation of lung cancer, and D represents the medium-risk category (risk category) of pre-cancerous lesions for the recognition problems of categories A, B, C, and D. In short, the meanings of categories A, B, C, and D can be arbitrarily changed, and their arbitrarily represented meanings are separated without confusion, and four-category recognition problems can be carried out in combinations of high, medium, and low risks, pathological classification, and lung cancer TNM staging.Classification and recognition problems of four or more categories also include classification and recognition problems of any combination of two categories, any combination of three categories, and any combination of four categories among them. The basic molecular marker index system involved in its characteristics is composed of the following target gene methylation levels, miRNA (microRNA) expression levels (expression levels), protein levels (protein expression levels), or various indicators of autoantibodies, or various indicators of exosomes, or any relevant lung cancer kits, and molecular markers composed of any two or more combinations of molecular markers in the molecular markers, or any one or two or more molecular markers preferably (or newly discovered) by other methods. The above molecular markers constitute the basic molecular marker index system; this basic molecular marker index system also includes using existing biological and medical detection technologies to separately obtain human body fluids such as human blood (plasma, serum, peripheral blood, free tfDNA in human blood, ctDNA free in human tumor blood), urine, lung fluid, exosomes, saliva, bronchoalveolar lavage fluid, pleural effusion, mononuclear cells in spinal fluid, circulating free DNA, circulating tumor DNA, etc., which are all called liquid test samples. For the liquid test samples, quantitative PCR, fluorescence quantitative PCR, quantitative methylation-specific PCR, real-time fluorescence quantitative PCR (qPCR), methylation chips, digital PCR, nucleic acid mass spectrometry, chemiluminescence, ELISA, immunochemiluminescence, chemiluminescence, electrochemiluminescence, next-generation sequencing technology, miRNA expression profile sequencing, fluorescence quantitative PCR detection and analysis after synthesizing cDNA, and detection technologies according to kits (national approved and unapproved, as well as foreign kits), etc., also including measuring instruments such as Abbott and Roche at home and abroad, are used to detect any combined molecular markers of the basic molecular marker index system to obtain detection data (also including data of any one or combination of molecular markers from any other sources, such as data detected by kit methods). Among them, the genes whose methylation levels are respectively referred to are: DNA methylation (one or more base fragments or one or more multiple CpG sites covered by a region) levels of molecular markers such as GLP2R, COPT1, P16, SHOX2, SCT, HOXA7, PTGER, PTEN, GSTP1, SOX1, PTGER4, OTX1, HISTIH3G, RASSFIA, PRTN3, TM, VIM, etc., also including multiple gene DNA methylation blocks, and also including methylation levels measured by kits for any one or combination of molecular markers, as well as newly discovered DNA methylation molecular markers related to lung cancer in the future;Among them, the miRNA expression levels respectively refer to the expression levels of molecular markers such as Let-Ti-5P, CALM13, miR-3654, miR-236, miR-221, miR-378, miR-1291, miR-29c, miR-214-3p, miR-376, miR-197, miR-182, miR-183, miR-1-3p, miR-155, miR-145, miR-21, miR-21-5p, miR-210, miR-320a-3p, miR-210-3p, miR-483, miR-143, miR-126, miR-139, miR-326, miR-101-2, miR-190, miR-944, miR-132-3p, and also include the miRNA expression levels detected by any one or more kits, as well as the expression levels of miRNA molecular markers related to lung cancer newly discovered in the future; among them, the protein level (expression level) or various indicator quantities of autoantibodies or various indicator quantities of exosomes are respectively TK1 (STK1), TNF-a, CA125, CA199, CA153, CEA, HE4, YKL-40, TPS, SCC (SCC-Ag), TSGF, CYFRA21-1, NSE, IL-6, ProGRP, SF, VEGF, TuM-2-PK, CK-18-3A9, M1RNA, P53, PGP9.5, GAGE7, MAGEA1, GBU4-5, CAGE, HSP90b, P53, GAGE7, PGP9.5, GAGE, MAGEA1, P16, SOX2, GBU4-5, etc., the protein levels or the quantities of liquid samples to be tested, the protein levels detected by one or any number of kits, as well as the protein levels related to lung cancer, the quantities of autoantibody indicators, and the quantities of exosome indicators newly discovered in the future; the basic molecular marker index system is composed of the above. The basic large model or small model algorithms involved are characterized in that the classification and recognition algorithms respectively include: random forest algorithm, AdaBoost classification algorithm, decision tree algorithm, logistic regression (Logistic including sparse canonical correlation analysis and logistic regression methods, etc.), Bayesian algorithm, support vector machine algorithm, neural network classification algorithm, machine learning classification algorithm (including hybrid deep learning models, etc.), deep learning classification algorithm, clustering analysis algorithm, discriminant analysis algorithm, multi-task association analysis method (HDS-MTAA), artificial intelligence classification algorithm, K-nearest neighbor algorithm, association rule classification algorithm, kernel-based algorithm, K-means algorithm, fuzzy classification and recognition algorithm, principal component analysis, deep self-reconstruction fusion similarity hashing method (DS-FSH), other classification algorithm recognition algorithms, etc. Comprehensive evaluation is carried out by one algorithm in their classification and recognition algorithms or a combined algorithm of any two or more classification and recognition algorithms. Or for a specifically selected basic target classification and recognition problem, and after detecting the data (or any data from two data sources of other data and detection data) after any combination of a group of specific molecular markers in the basic molecular marker index system, an optimized arbitrary algorithm or a determined algorithm, or any one algorithm in the classification and recognition algorithms is used; all of the above algorithms and their arbitrary combined algorithms constitute the basic large model or small model algorithms.

2. An application technology of membrane type and computing power enabling the screening of high-risk groups for lung cancer, early auxiliary diagnosis, and the monitoring of the treatment efficacy of lung cancer patients after surgery, characterized in that, Expand the above-mentioned basic molecular marker index system in item 1 to imaging devices. Based on chest imaging, the maximum long diameter, volume, density, changes in solid components, margins, infiltration status, vascular growth status, presence or absence of lobulation, spiculation, and pleural indentation sign of pulmonary nodules (or lesions or masses) can be extracted, all of which constitute the marker index system of pulmonary imaging features. The result calculated by using AI technology by adopting any combination of multiple indicators in the marker index system of pulmonary imaging features is consistent (or relatively consistent) with the result obtained by the basic target classification and recognition in item 1 above, so as to achieve precise auxiliary diagnosis of lung cancer. In addition, combine the marker index system of pulmonary imaging features with the basic molecular marker index system in item 1 above, and arbitrarily select a combination of three or more marker indicators from them (preferably or by other methods), and use the similar method in item 1 above (or use artificial intelligence methods or the DeepSeek large model and algorithm) to achieve the purpose of the classification and recognition problem in item 1 above, so as to achieve clinical precise auxiliary diagnosis. Moreover, use the similar method in item 1 above to obtain the screening result of high-risk (lung cancer), and add any one indicator in the marker index system of pulmonary imaging features determined by imaging (or add the direct lineal relatives with a family history of lung cancer), which can achieve precise auxiliary diagnosis of lung cancer. Furthermore, use the similar method in item 1 above to obtain the screening result of high-risk (lung cancer), and add any one indicator in the marker index system of pulmonary imaging features determined by imaging and the situation of direct lineal relatives with a family history of lung cancer, which can achieve precise auxiliary diagnosis of lung cancer.

3. An application technology for a model and computing power to empower the screening of high-risk groups for lung cancer, early auxiliary diagnosis, and the monitoring of the treatment efficacy of lung cancer patients after surgery, characterized in that, For item 1 or item 2 above respectively, expand to construct a linear function or a non-linear function. Use any one or more molecular markers in the basic molecular marker index system (or any one or more indicators in the marker index system of pulmonary imaging features) to depict the variation law from normal people to the formation of lung cancer during the development process (the molecular evolution process law of lung cancer growth and formation), and construct a type of linear function or another type of non-linear function to reflect the variation law. Combine a type of linear function or another type of non-linear function with any one or a combination of two or more algorithms of the basic large model or small model algorithm. By using the similar method in item 1 or 2 above respectively, the purpose of the classification and recognition problem in item 1 or 2 above can be solved, and the purpose of screening high-risk groups of lung cancer or clinical auxiliary diagnosis can be achieved.

4. An application technology for a model and computing power to empower the screening of high-risk lung cancer populations, early auxiliary diagnosis, and the monitoring of the treatment efficacy of lung cancer patients after surgery, characterized in that The pathological classification of lung cancer and the clinical staging TNM of lung cancer. Respectively, use the similar methods of the above 1 (or 2 or 3) to calculate the purposes of respectively identifying small cell lung cancer, squamous cell lung cancer and adenocarcinoma of the lung. This pathological classification of lung cancer falls within the scope required by this clause of this right. Secondly, use one or more functions of a constructed class of linear functions or one or more functions of another class of non-linear functions or one or more functions of the combination of the two classes of functions (or a system of linear equations or a system of non-linear equations). For any one or more molecular markers in the used basic molecular marker index system (or any one or more indicators of the marker index system of lung imaging features) data, establish one or more functions of lung cancer with the changing rules of clinical pathological staging TNM (or one or more functions of the TNM changing rules of small cell lung cancer or squamous cell lung cancer or adenocarcinoma of the lung or non-small cell lung cancer of the lung). Then, for the actual detection data of any lung cancer patient (or small cell lung cancer or squamous cell lung cancer or adenocarcinoma of the lung or non-small cell lung cancer of the lung), substitute the data into the functions of the corresponding lung cancer patient (or small cell lung cancer or squamous cell lung cancer or adenocarcinoma of the lung or non-small cell lung cancer of the lung) respectively, obtain one or more values, and then find the average value of the one or more values, which is the clinical staging TNM of that lung cancer patient (or small cell lung cancer or squamous cell lung cancer or adenocarcinoma of the lung or non-small cell lung cancer of the lung), and also falls within the scope required by this clause of this right. Here, it is marked that the clinical staging TNM includes three classification methods of N, T, M clinical staging. The clinical staging TNM method of this right refers to the algorithms for N, T, M clinical staging respectively. Or developing the clinical staging TNM into a table also falls within the scope required by this clause of this right.

5. An application technology for a model and computing power to empower the screening of high-risk lung cancer populations, early auxiliary diagnosis, and the monitoring of the treatment efficacy of lung cancer patients after surgery, characterized in that Among the molecular markers in the basic molecular marker index system involved in Method 1 (the national-approved kits are regarded as molecular markers), for two or more arbitrarily selected molecular markers that are optimized (or selected or artificially specified), the negative and positive results determined by calculating them using the similar methods of 1 or 2 or 3 above (including those calculated according to the detection methods of the kits, also including those calculated using the sensitivity and specificity analysis methods of the ROC curve, and also including those calculated by other methods), if the sum of the positive quantities is more than the sum of the negative quantities, the final diagnosis is positive; if the sum of the positive quantities is less than the sum of the negative quantities, the final diagnosis is negative; or, for two or more arbitrarily selected molecular markers that are optimized (or selected or artificially specified), and determine (or figure out) the quantity of the molecular markers, for the already determined positive or negative of each molecular marker, respectively stipulate the weight coefficient of each molecular marker according to the contribution size of each molecular marker (or artificially stipulate the weight coefficient of each molecular marker, or use other ideas or other practices or other calculation methods to determine the weight coefficient of each molecular marker respectively), finally assign numerical values to the actually detected (or calculated from the experimental detection data or artificially judged) positive or negative of each molecular marker, and determine (diagnose) positive or negative according to whether the sum of the products of the numerical values of different molecular markers and their corresponding weight coefficients is greater than or equal to a certain fixed number or less than a certain fixed number. All such methods are within the scope required by this clause of this right. It is characterized in that the method for determining (or diagnosing) positive and negative for lung cancer in the entire first paragraph above can be analogously extended to any other cancer. For any multiple molecular biomarkers associated with any other cancer, the method for determining (or diagnosing) negative and positive for any other cancer in a similar manner to the entire first paragraph above also belongs to the scope required by this clause of this right. Here, any other cancer respectively refers to ovarian cancer, breast cancer, colorectal cancer, pancreatic cancer, gastric cancer and other cancers in oncology. It is characterized in that the method for determining (or diagnosing) positive and negative for lung cancer in the entire first paragraph above can be analogously extended to determining (or diagnosing) positive and negative for any kit associated with any cancer respectively. For any multiple molecular biomarkers associated with any one kit, the method for determining (or diagnosing) negative and positive for any one kit in a similar manner to the entire first paragraph above. This method for determining negative and positive within the kit also belongs to the scope required by this clause of this right.

6. An application technology for a model and computing power to empower the screening of high-risk lung cancer populations, early auxiliary diagnosis, and the monitoring of the postoperative treatment efficacy of lung cancer patients, characterized in that, Expand the above basic molecular marker index system to the molecular markers related to lung cancer discovered in the future (including the test kits newly approved by the state in the future). For any combination of one or two or more molecular markers in the basic molecular marker index system and the molecular markers related to lung cancer discovered in the future (including the test kits newly approved by the state in the future), use the similar methods of the above 1, 2, 3, and 4 respectively to achieve the purpose of the classification and recognition problem in the above 1. It belongs to the scope required by this clause of this right. It is characterized in that the content of the first paragraphs of the above 1, 2, 3, 4, 5 and the above 6 is respectively expanded to the DeepSeek large model, agent application and innovation (or artificial intelligence) related to lung cancer; or the DeepSeek large model, agent application and innovation (or artificial intelligence) are respectively expanded to the content of the first paragraphs of the above 1, 2, 3, 4, 5 and the above 6 (or the scenario application of the above lung cancer); or the content of the first paragraphs of the above 1, 2, 3, 4, 5 and the above 6 (or the scenario application of lung cancer) and the DeepSeek large model, agent application and innovation (or artificial intelligence) are respectively combined with each other in corresponding research applications; all of these belong to the scope required by this clause of this right.

7. An application technology of using models and computing power to empower the screening of high-risk lung cancer populations, early auxiliary diagnosis, and the monitoring of the treatment efficacy of lung cancer patients after surgery, characterized in that, For any of the molecular markers described in the above 1 and 6 (such as CA125, CA153, CEA, etc.), take one or any two or more combinations of molecular markers, and according to the corresponding molecular markers, construct one or more functions of a class of linear functions or one or more functions of another class of non-linear functions or one or more functions of the combination of the two classes (the above are collectively referred to as models). For any postoperative lung cancer patient, perform a non-invasive detection experiment on the corresponding molecular marker to obtain the corresponding data (the data can also be measured dynamically and repeatedly), and bring it into the model to dynamically observe (or dynamically evaluate) the development trend of the efficacy of postoperative lung cancer patients (or the development trend of the pathological typing of postoperative lung cancer patients or the development trend of the clinical stage TNM of postoperative lung cancer patients), and then analyze, compare and evaluate the treatment efficacy, progress of postoperative lung cancer patients, and help doctors on how to treat and guide the treatment effect.

8. An application technology for a model and computing power to empower the screening of high-risk lung cancer populations, early auxiliary diagnosis, and the monitoring of the therapeutic efficacy of lung cancer patients after surgery, characterized in that, Among the basic molecular marker donation systems in 1 above and the molecular marker indicators involved in 6 above, take (or preferably take) one or more marker combinations, and based on a large amount of actual data obtained from their experimental detections, respectively study them through similar methods as in 1, 2, 3, 4, 5, 6, 7, and 8 above. In the first mode, the randomly selected (or preferably selected) molecular markers are made into a corresponding data range table. For the data of the molecular markers detected experimentally for any individual, it is compared through the data range table to determine the negativity and positivity of lung cancer. In the second mode, a large amount of detection data of the randomly selected (or preferably selected) molecular marker combinations are calculated into various indices or comprehensive indices. For the data of the molecular markers detected experimentally for any individual, through the corresponding calculations of the various index or comprehensive index data, the negativity and positivity of lung cancer are determined. In the third mode, a large amount of detection data of the randomly selected (or preferably selected) molecular marker combinations are plotted using different calculation methods in software and loaded into a portable device (or microcomputer, mobile phone, online platform, proprietary device, etc.) for calculation. For the data of the molecular markers detected experimentally for any individual, it is input into the portable device (or microcomputer, mobile phone, online platform, proprietary device, etc.) for individual calculation to determine the negativity and positivity of lung cancer.

9. An application technology for a model and computing power to empower the screening of high-risk lung cancer populations, early assisted diagnosis, and the monitoring of the therapeutic efficacy of lung cancer patients after surgery, characterized in that, Previously, respectively use similar methods as in 1, 2, 3, 4, 5, 6, and 7 above to study the specific implementation methods or calculation processes, and then compile the specific implementation methods or calculation processes into software and load them into a portable device (microcomputer, mobile phone, online platform, proprietary device, etc.). For the experimental detection data (including the data processing results of relevant matters) of a patient seeking medical treatment, input them into the portable device (microcomputer, mobile phone, online platform, proprietary device, etc.). After individual calculation, output the result that determines which category the patient seeking medical treatment belongs to among the basic target classification and recognition problems in 1 above.

10. An application technology for a model and computing power to empower the screening of high-risk lung cancer populations, early auxiliary diagnosis, and the monitoring of the postoperative treatment efficacy of lung cancer patients, characterized in that, For any molecular markers (regarding kits as equivalent to molecular markers) associated with any other specific cancers and any combinations of their molecular markers, the application technologies for screening high-risk populations, early auxiliary diagnosis, and monitoring the postoperative treatment efficacy of lung cancer in 1, 2, 3, 4, 5, 6, 7, 8, and 9 above can be respectively extended to any other specific cancers. Any other specific cancers can respectively use similar thinking methods or models and computing power ideas as in 1, 2, 3, 4, 5, 6, 7, 8, and 9 above to respectively solve the application technologies for screening high-risk populations, early auxiliary diagnosis, pathological typing, clinical staging, and evaluating the postoperative treatment efficacy of any other specific cancers, all of which fall within the scope required by this claim clause (1, 2, 3, 4, 5, 6, 7, 8, and 9 above). Here, any other specific cancers respectively refer to various specific cancers included in oncology, such as ovarian cancer, breast cancer, colorectal cancer, pancreatic cancer, gastric cancer, and so on for any other specific cancers.