CN-121983325-A - Digital twin construction method for traditional Chinese medicine full-flow diagnosis and treatment of liver cancer

CN121983325ACN 121983325 ACN121983325 ACN 121983325ACN-121983325-A

Abstract

The invention discloses a digital twin construction method for liver cancer traditional Chinese medicine full-flow diagnosis and treatment, which comprises the following steps of firstly collecting multi-mode traditional Chinese medicine diagnosis and treatment data, secondly carrying out data standardization and feature fusion treatment, thirdly carrying out traditional Chinese medicine diagnosis and treatment core element modeling, fourthly carrying out digital twin body initialization construction, fifthly establishing a virtual-real linkage mapping mechanism, sixthly carrying out dynamic simulation and iterative optimization on the diagnosis and treatment process, seventhly carrying out intelligent generation on a full-flow diagnosis and treatment scheme, eighth carrying out model verification and calibration, ninth carrying out diagnosis and treatment effect prediction evaluation, and tenth carrying out continuous update on the digital twin body, and creatively fusing a traditional Chinese medicine three-resolution-one-combination diagnosis and treatment idea and a digital twin technology to construct a digital twin system for liver cancer traditional Chinese medicine full-flow diagnosis and treatment, so as to realize quantification and visualization of traditional Chinese medicine diagnosis and treatment core elements, solve the technical pain points of strong subjectivity and difficult quantification of traditional Chinese medicine diagnosis and treatment, and provide a brand-new technical path for traditional Chinese medicine accurate diagnosis and treatment.

Inventors

SUN KEXIN
LING CHANGQUAN
YANG ZHISHENG

Assignees

上海中医药大学

Dates

Publication Date: 20260505
Application Date: 20260117

Claims (10)

1. The digital twin construction method for the traditional Chinese medicine full-flow diagnosis and treatment of liver cancer is characterized by comprising the following steps of: The method comprises the steps of firstly, acquiring multi-mode traditional Chinese medicine diagnosis and treatment data, and acquiring traditional Chinese medicine four-diagnosis data, western medicine examination data, diagnosis and treatment process data and physique characteristic data of a liver cancer patient; Step two, data standardization and feature fusion processing are carried out, the collected multi-mode data are cleaned, denoised and marked, the Chinese medicine text data are converted by adopting a natural language processing technology, and a unified high-dimensional feature vector set is generated through a feature fusion algorithm; modeling the core elements of traditional Chinese medicine diagnosis and treatment of liver cancer, and constructing a pathogenicity-disease position-disease period identification model, a viscera and meridian mapping model and a vital-qi evil elimination dynamic model based on the diagnosis and treatment idea of 'three differentiation and one combination' of traditional Chinese medicine; initializing and constructing a digital twin body, inputting a quantized result of the core element model constructed in the third step as a core parameter, integrating the core parameter with the high-dimensional feature vector set generated in the second step, and generating a personalized digital twin body initial model of the liver cancer patient by adopting a three-dimensional modeling technology and a finite element analysis method; establishing a virtual-real linkage mapping mechanism, deploying a real-time data transmission channel, and establishing a bidirectional mapping relation of physiological parameters, diagnosis and treatment behaviors and pathological states between a physical diagnosis and treatment process and a digital twin body; step six, dynamic simulation and iterative optimization of the diagnosis and treatment process, namely simulating the diagnosis and treatment intervention effect of the traditional Chinese medicine based on real-time mapping data, and performing iterative optimization on digital twin body model parameters through a reinforcement learning algorithm; Step seven, the whole-flow diagnosis and treatment scheme is intelligently generated, and the personalized and staged traditional Chinese medicine whole-flow diagnosis and treatment scheme for liver cancer is generated by combining the digital twin body simulation result and a traditional Chinese medicine diagnosis and treatment knowledge base; step eight, model verification and calibration, namely calibrating model parameters by adopting an error correction algorithm through comparison and analysis of clinical diagnosis and treatment data and digital twin body simulation data, so as to ensure model accuracy; step nine, predicting and evaluating diagnosis and treatment effects, namely predicting the disease progress, curative effects and adverse reactions of different diagnosis and treatment stages based on the optimized digital twin bodies, and generating a multi-dimensional evaluation report; And ten, continuously updating the digital twin body, accessing the follow-up diagnosis and treatment data, rehabilitation data and follow-up visit data of the patient in real time, and dynamically updating the digital twin body model to realize full-period diagnosis and treatment tracking and optimization.
2. The method for constructing digital twin diagnosis and treatment of liver cancer according to claim 1, wherein in the first step, the four diagnosis data of traditional Chinese medicine include facial diagnosis images, tongue diagnosis images, pulse diagnosis signals and inquiry texts; The western medicine examination data comprise liver imaging data, laboratory examination data and pathology biopsy data; The diagnosis and treatment process data comprise traditional Chinese medicine prescription data, acupuncture point data and treatment time sequence data, wherein the traditional Chinese medicine prescription data comprise a medicine composition, a dosage and a decoction method, the acupuncture point data comprise acupuncture point selection, acupuncture depth and needle retention time, and the treatment time sequence data comprise treatment intervals, treatment course periods and treatment implementation time stamps; the physique characteristic data comprise physique type judging data, genetic background data and life habit data.
3. The digital twin construction method for the traditional Chinese medicine full-flow diagnosis and treatment of the liver cancer is characterized in that a weighted average fusion strategy is adopted by the feature fusion algorithm in the second step, the weight is determined based on influence coefficients of all mode data on the traditional Chinese medicine full-flow diagnosis and treatment of the liver cancer, the influence coefficients are obtained through training of a random forest algorithm, and the standardized processing adopts a Z-score standardized method to map the data to a [0,1] interval in a unified mode.
4. The digital twin construction method for traditional Chinese medicine full-flow diagnosis and treatment of liver cancer according to claim 1 is characterized in that in the third step, the pathogenicity-disease position-disease period identification model is constructed by adopting a deep learning network, input is a standardized multi-mode feature vector, output is a yin-yang cold-heat deficiency-excess pathogenicity judgment result, a six-meridian-viscera disease position positioning result and an early, middle and late disease period division result, and the viscera and meridian mapping model is used for establishing a correlation weight matrix of viscera such as liver, spleen, stomach, kidney and the like and related meridians based on traditional Chinese medicine visceral manifestation theory.
5. The digital twin construction method for traditional Chinese medicine full-flow diagnosis and treatment of liver cancer according to claim 1, wherein in the fourth step, the three-dimensional modeling technology adopts Cinema 4D software to construct a human body basic model, and the morphology of a liver lesion area is accurately matched by combining with the imaging data of a patient; The finite element analysis method adopts ANSYS software to generate a liver tissue grid model to simulate the physiological function and the mechanical property of pathological changes of the liver.
6. The digital twin construction method for traditional Chinese medicine full-flow diagnosis and treatment of liver cancer according to claim 1 is characterized in that in the fifth step, a 5G+ edge computing architecture is adopted for the real-time data transmission channel, the data transmission delay is controlled within 50ms, the bidirectional mapping relation is realized by establishing a data mapping dictionary, and the dictionary comprises a one-to-one correspondence relation between physical entity parameters and digital twin parameters and a conversion formula.
7. The method for constructing digital twin in traditional Chinese medicine full-flow diagnosis and treatment of liver cancer according to claim 1, wherein in the step six, the reinforcement learning algorithm adopts a deep Q network DQN, the diagnosis and treatment effect is promoted to be a reward function, the traditional Chinese medicine dosage adjustment and the acupoint selection optimization are used as action spaces, and diagnosis and treatment intervention simulation parameters of the digital twin are optimized iteratively.
8. The digital twin construction method of full-flow diagnosis and treatment of liver cancer according to claim 1, wherein in the seventh step, the traditional Chinese medicine diagnosis and treatment knowledge base comprises traditional Chinese medicine diagnosis and treatment specifications of liver cancer, clinically verified classical medical records data of liver cancer, traditional Chinese medicine compatibility rules and targeted acupuncture treatment schemes of liver cancer, and the knowledge base is constructed in a knowledge graph form to support dynamic update and association retrieval.
9. The method for constructing digital twin in traditional Chinese medicine full-flow diagnosis and treatment of liver cancer according to claim 1, wherein in the eighth step, the error correction algorithm adopts a least square method to carry out fitting correction on the deviation of physiological indexes, pathological states and clinical actual data of digital twin body simulation, and the corrected model error is controlled within 3%.
10. The method for constructing digital twin in traditional Chinese medicine full-flow diagnosis and treatment for liver cancer according to claim 1, wherein the continuous updating in the step ten adopts an incremental learning algorithm, model parameters are updated based on newly-added data on the premise of not retraining the whole model, and the real-time tracking capability of the digital twin on the patient condition change is ensured.

Description

Digital twin construction method for traditional Chinese medicine full-flow diagnosis and treatment of liver cancer Technical Field The invention belongs to the crossing field of digital twin technology and traditional Chinese medicine diagnosis and treatment technology, and particularly relates to a digital twin construction method for liver cancer traditional Chinese medicine full-flow diagnosis and treatment. Background Liver cancer is one of the high malignant tumors in China, belongs to the categories of liver accumulation, mass and the like in traditional Chinese medicine, and has the symptoms of deficiency and excess, and the pathogenesis core is the excessive phlegm, blood stasis, heat and toxic cementing and condensation of the body resistance and deficiency evil. The traditional Chinese medicine treatment of liver cancer takes strengthening body resistance to eliminate pathogenic factors as a core, emphasizes diagnosis and treatment ideas of three differentiation and one differentiation (differentiation of disease nature, differentiation of disease position and differentiation of disease period, and combination and theory of whole and partial treatment), can participate in all stages of liver cancer treatment in the whole course, and plays roles of inhibiting tumor growth, reducing toxicity, enhancing efficiency and improving life quality. However, the traditional liver cancer traditional Chinese medicine diagnosis and treatment process has a plurality of limitations that firstly the diagnosis and treatment process depends on the experience of doctors, the subjectivity is strong, quantitative evaluation is difficult to realize, secondly the dynamic change of the patient is difficult to track in real time, the diagnosis and treatment scheme is delayed in adjustment, thirdly the core elements of the traditional Chinese medicine meridians, viscera and the like cannot be visually presented, the communication difficulty between doctors and patients is high, and the diagnosis and treatment effect prediction is lack of accurate support. The digital twin technology realizes virtual-real linkage, dynamic simulation and intelligent optimization by constructing a virtual mapping model of a physical entity, and is gradually applied in the medical field, such as virtual patient construction, clinical trial simulation and the like. In the prior art, although related researches such as traditional Chinese medicine visceral manifestation digital twin, whole blood vessel digital people and the like exist, the traditional Chinese medicine visceral manifestation digital twin system is focused on the visualization of single viscera or the diagnosis and treatment assistance of Western medicine diseases, and a digital twin system is not established aiming at the specificity of traditional Chinese medicine whole-flow diagnosis and treatment of liver cancer. Specifically, the prior art does not combine the diagnosis and treatment concept of three-resolution one-with-one of traditional Chinese medicine, cannot realize accurate modeling of the core elements of the diagnosis and treatment of traditional Chinese medicine, lacks an effective combination mechanism of multi-mode traditional Chinese medicine diagnosis and treatment data, is difficult to support generation of a personalized diagnosis and treatment scheme, does not establish a virtual-real linkage and dynamic optimization mechanism of full-flow diagnosis and treatment, and cannot meet the requirements of the whole-course diagnosis and treatment of liver cancer traditional Chinese medicine. Therefore, a construction method for combining diagnosis and treatment characteristics of traditional Chinese medicine with digital twin technology is needed, the pain point problem of traditional liver cancer diagnosis and treatment is solved, and the accuracy and intelligence level of diagnosis and treatment are improved. Based on the method, a digital twin construction method for liver cancer traditional Chinese medicine full-flow diagnosis and treatment is designed. Disclosure of Invention Aiming at the situation, in order to overcome the defects of the prior art, the invention provides a digital twin construction method for traditional Chinese medicine full-flow diagnosis and treatment of liver cancer, which effectively solves the problems proposed by the background technology. In order to achieve the purpose, the invention provides the following technical scheme that the digital twin construction method for the traditional Chinese medicine full-flow diagnosis and treatment of liver cancer comprises the following steps: The method comprises the steps of firstly, acquiring multi-mode traditional Chinese medicine diagnosis and treatment data, and acquiring traditional Chinese medicine four-diagnosis data, western medicine examination data, diagnosis and treatment process data and physique characteristic data of a liver cancer patient; Step two, data standardization and featur