Traditional Chinese medicine decoction piece full life cycle data management system and method

Abstract

The invention discloses a traditional Chinese medicine decoction piece full life cycle data management system and method, and relates to the field of data processing, and the method comprises the steps: collecting traditional Chinese medicine planting environment data through a sensor, and encrypting and storing the data through a block chain; in the processing link, the moisture evaporation condition is predicted through an LSTM neural network, and an early warning mechanism is set; in the processing process, the process is adjusted through a Raman spectrum; components and pesticide residues of the finished product are ensured to be qualified through mass spectrometric detection; the clinical end collects prescription and curative effect data optimization process; all data are synchronized through a block chain and a cross-chain protocol; through a data lake architecture and a graph neural network model, a planting area and a processing technology are optimized. The system has the advantages that through technologies such as a sensor, a block chain and an AI model, whole-process data tracing and optimization from planting to clinical use are realized, the quality and curative effect of a traditional Chinese medicine product are improved, and transparency and safety of the whole process are ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of data processing, in particular to a traditional Chinese medicine decoction piece whole life cycle data management system and method. BACKGROUND

[0002] As an important part of traditional Chinese medicine, the quality of traditional Chinese medicine decoction pieces directly affects the efficacy and safety of traditional Chinese medicine. With the continuous development of the traditional Chinese medicine industry, how to ensure the quality of traditional Chinese medicine decoction pieces, improve management efficiency, and trace the source of medicinal materials has become a problem that needs to be solved in the industry.

[0003] The current market traditional Chinese medicine decoction piece whole life cycle data management system and method relies on manual or simple monitoring means, lacks systematic data collection and intelligent analysis, and it is difficult to achieve fine management and full traceability in planting, processing, and processing. Especially in the quality control and processing technology of medicinal materials, the traditional method cannot fully utilize the potential of big data and artificial intelligence, and cannot dynamically predict and adjust the possible deviation in the processing process. Therefore, the quality of the product is easily affected by environmental fluctuations or process deviation, and it is difficult to ensure the stability and consistency of each batch of decoction pieces. In addition, the existing traceability system relies on traditional databases or simple tracing methods, information is easy to tamper with or lacks effective real-time monitoring and feedback mechanism, leading to reduced supervision and consumer trust. SUMMARY

[0004] In order to improve the existing system and method, a traditional Chinese medicine decoction piece whole life cycle data management system and method is provided, which realizes full data traceability and optimization from planting to clinical use through sensors, blockchains, AI models and other technologies, improves the quality and efficacy of traditional Chinese medicine products, and ensures the transparency and safety of the whole process.

[0005] To achieve the above purpose, the technical scheme adopted by the present application is: The traditional Chinese medicine decoction piece whole life cycle data management method comprises: Based on the planting link of medicinal materials, the growth environment parameters are collected in real time by sensors, the geographical information data of the planting area is obtained, the initial ingredient content dataset is obtained by near-infrared spectrum detection of medicinal materials, and the planting main body, harvesting time and geographical position data are encrypted and stored by blockchain nodes; Based on the processing link of medicinal materials, the processing environment is monitored in real time by a temperature and humidity sensor, a water evaporation model based on an LSTM neural network is constructed, a water evaporation curve of medicinal materials is dynamically predicted, and a processing deviation early warning mechanism is established; Based on the processing link of medicinal materials, a traditional Chinese medicine chemical component transformation knowledge graph is constructed according to historical traditional Chinese medicine processing data, and the processing parameters are adjusted based on the Raman spectrum detection results; Based on the inspection link of traditional Chinese medicine decoction pieces, the content of effective components and heavy metal pesticide residue indicators in traditional Chinese medicine decoction pieces are detected by high-resolution mass spectrometry, and the detection results are correlated with batch number and processing parameters in time and space; Based on the inspection qualified traditional Chinese medicine decoction pieces, the prescription compatibility data and patient efficacy evaluation are collected at the clinical use end to construct decoction piece clinical test data set, and the processing parameters are optimized and adjusted; The time stamp hash value of the key data of each link in the whole life cycle of traditional Chinese medicine decoction pieces is written into the alliance chain, a unique blockchain digital identification code is generated for each decoction piece batch, and is synchronized to each link through cross-chain protocol; The life cycle data lake architecture is constructed, the planting, processing, processing, quality inspection, clinical link data are integrated, the decoction piece quality prediction model based on graph neural network is constructed, and the planting area optimization scheme and processing technology improvement suggestion are output.

[0006] Preferably, based on the planting link of traditional Chinese medicinal materials, the growth environment parameters are collected in real time by sensors, the geographical information data of planting area is obtained, the initial component content data set is obtained by near infrared spectrum detection of traditional Chinese medicinal materials, and the planting subject, harvesting time and geographical position data are encrypted and stored by blockchain node, which specifically includes: Multiple source sensors are arranged in the planting area of traditional Chinese medicinal materials, and the key parameters of the growth environment are monitored in real time, including temperature, humidity, soil pH value, light intensity, air quality, geographical position information is obtained, including land information, soil quality, climate condition data, and a digital planting area map is formed; The initial data of medicinal material quality is obtained in real time by near infrared spectrum in a non-destructive way, and the main component content is measured; Based on the obtained data of traditional Chinese medicinal material planting link, the data is encrypted and stored by blockchain technology, and uploaded to the blockchain network.

[0007] Preferably, based on the processing link of traditional Chinese medicinal materials, the processing environment is monitored in real time by temperature and humidity sensor, the water evaporation model based on LSTM neural network is constructed, and the water evaporation curve of traditional Chinese medicinal materials is dynamically predicted, and the processing deviation early warning mechanism is established, which specifically includes: The air temperature and humidity environment parameters are detected in real time by the temperature and humidity sensor in the traditional Chinese medicinal material processing site; A water evaporation model based on long short-term memory network is constructed, which is trained by historical processing data to predict the water evaporation curve of traditional Chinese medicinal materials under different environmental conditions; The collected environmental parameter data is constructed into time series data set, and input into the trained water evaporation model to predict the water evaporation of traditional Chinese medicinal materials, and generate real-time water evaporation curve; Based on the prediction results of the water evaporation model, a processing deviation range threshold is set, and when the actual moisture content deviates from the predicted curve by more than the threshold, a deviation warning is triggered.

[0008] Preferably, based on the processing link of traditional Chinese medicinal materials, a traditional Chinese medicine chemical component transformation knowledge graph is constructed according to historical traditional Chinese medicinal material processing data, and the processing parameters are adjusted based on the Raman spectrum detection results, specifically including: Based on the historical processing data, the chemical component transformation knowledge graph is constructed based on the chemical component transformation knowledge graph. The main chemical components of traditional Chinese medicinal materials are taken as nodes, and the possible paths and reaction conditions between the components are taken as edges to construct a traditional Chinese medicine chemical component transformation knowledge graph. Real-time Raman spectrum data of traditional Chinese medicinal material samples in the processing process are collected and matched with the chemical component nodes in the traditional Chinese medicine chemical component transformation knowledge graph to obtain the chemical component content and transformation relationship of the traditional Chinese medicinal material samples. Based on the results of Raman spectrum matching, the transformation rate of each chemical component in the processing process is calculated, and the transformation of key components is particularly concerned. Based on the obtained transformation rate results, the processing technology is adjusted for the processing process deviating from the target interval, including adjusting the frying temperature and cooking time.

[0009] Preferably, based on the inspection link of traditional Chinese herbal piece finished products, the effective component content and heavy metal and pesticide residue index data of traditional Chinese herbal piece finished products are detected by high-resolution mass spectrometry, and the detection results are spatiotemporally associated and mapped with batch number and processing parameters, specifically including: Random samples are taken from each batch of traditional Chinese herbal piece finished products by high-resolution mass spectrometry to obtain the main effective component content, heavy metal element content and pesticide residue content in the traditional Chinese herbal piece. The mass spectrometry detection results are recorded in real time, spatiotemporally associated with the production data of each batch of traditional Chinese herbal pieces, and a spatiotemporal association database is constructed to store batch number, production process, spatiotemporal information and mass spectrometry detection result data.

[0010] Preferably, based on the inspection qualified traditional Chinese herbal pieces, prescription compatibility data and patient efficacy evaluation are collected at the clinical use end to construct a herbal piece clinical test data set for optimizing and adjusting the processing parameters, specifically including: At the clinical use end of traditional Chinese herbal pieces, patient medication information, prescription compatibility data, medication time, dosage and treatment course data are collected to construct a clinical data set. Based on each batch of traditional Chinese herbal pieces in the clinical data set, the processing parameters of the batch are associated to integrate and construct a herbal piece clinical test data set, including patient basic information, prescription compatibility information, patient efficacy evaluation data and traditional Chinese herbal piece processing parameters. Through the random forest algorithm, the basic information of the patient, the prescription matching information and the processing parameters of the traditional Chinese medicine decoction pieces are taken as characteristic variables, and the curative effect evaluation data of the patient is taken as the target variable to model and train the model; Based on the trained random forest model, the correlation rules between the process parameters and the curative effect are obtained by evaluating the importance of each feature, and the processing parameters are optimized.

[0011] Preferably, the time stamp hash value of the key data of each link in the whole life cycle of the traditional Chinese medicine decoction piece is written into the alliance chain, a unique blockchain digital identification code is generated for each decoction piece batch, and is synchronized to each link through a cross-chain protocol, which specifically includes: Based on the data of each link in the whole life cycle of the traditional Chinese medicine decoction piece, a time stamp is generated when the data is collected, and a hash value is generated by hash calculation; A unique hash value is generated for each traditional Chinese medicine decoction piece batch at each link in the life cycle, and the hash value and the link data are created into a block; A unique digital identification code is generated for each batch in the alliance chain as a digital identity, and each block is written into the blockchain in chronological order; Based on the information flow requirements between links, a cross-chain protocol is designed, and the data of each link is synchronized to the blockchain of other related links through the cross-chain protocol.

[0012] Preferably, the life cycle data lake architecture is constructed, the planting, processing, processing, quality inspection, clinical link data are integrated, and a decoction piece quality prediction model based on graph neural network is constructed, and the output planting area optimization scheme and processing technology improvement suggestion specifically includes: The life cycle data lake is constructed based on the planting, processing, processing, quality inspection, clinical link data; According to the links of the life cycle, each link is regarded as a node in the graph, and the nodes are connected through the interlink relationship to form the edges in the graph, and the graph is constructed; The integrated life cycle data is converted into the form of a graph and input into the decoction piece quality prediction model based on graph neural network, the relationship between the nodes is learned through training the graph neural network, and the decoction piece quality is predicted; Based on the decoction piece quality prediction result of the model, the medicinal material quality of different planting areas, the optimal planting area and the processing technology optimization suggestion are obtained.

[0013] Preferably, the blockchain adopts a hierarchical architecture: the data layer is used to store the IPFS hash pointers of the original data of each link; the contract layer is used to deploy the smart contract to realize data upload permission control and automatic generation of audit logs; the application layer is used to provide API interface for the supervision agency to check the data integrity, and supports 1 second of traceability query response.

[0014] Further, a traditional Chinese medicine decoction piece whole life cycle data management system is proposed, comprising: A data acquisition module: the module is used for real-time acquisition of environmental parameters and geographic information data of medicinal material planting links, and initial component data is acquired through near-infrared spectrum detection; A processing environment monitoring module: the module is used for real-time monitoring of the temperature and humidity environment in the medicinal material processing link, predicting the moisture evaporation curve based on the LSTM neural network, and optimizing the processing quality through the deviation early warning mechanism; A processing technology optimization module: the module constructs a chemical component conversion knowledge graph according to historical processing data and Raman spectrum detection results, and adjusts the processing technology in real time to ensure the conversion of effective components of medicinal materials; A sampling data module: the module detects the effective components and pesticide residue indexes of traditional Chinese medicine decoction pieces by high-resolution mass spectrometry, and performs space-time correlation with batch number and processing technology parameters; A clinical test module: the module is used for collecting patient data and efficacy evaluation of the clinical use end, and optimizing the processing technology of traditional Chinese medicine decoction pieces through correlation analysis with processing technology parameters; A blockchain management module: the module generates the timestamp hash value of the data of each link in the whole life cycle of traditional Chinese medicine decoction pieces, and ensures the real-time synchronization and transparent traceability of the data of each link through the blockchain and cross-chain protocol, constructs a graph neural network model through the life cycle data lake, and gives improvement suggestions for the quality prediction of decoction pieces; A processor: the processor is used for processing the calculation process of each formula and the construction calculation process of each model.

[0015] Compared with the prior art, the advantages of the present application are: Through modern technical means, the traditional Chinese medicine industry chain is comprehensively digitized and intelligently upgraded, realizing the whole process traceability and optimization from planting, processing, processing to clinical application. Firstly, the planting environment parameters are collected in real time by using sensors, combined with near-infrared spectrum technology and blockchain encryption storage, to ensure the traceability of the quality and source of medicinal materials. Secondly, through the LSTM neural network model and Raman spectrum analysis, the moisture evaporation and component conversion in the processing process are dynamically monitored to ensure the accuracy of the processing technology. Thirdly, through high-resolution mass spectrometry analysis and space-time data correlation, the quality safety of decoction piece products is effectively guaranteed, and the process is optimized according to the clinical data to improve the efficacy. Finally, with the help of blockchain technology, the data is transparent and tamper-proof, ensuring the real-time synchronization of data in each link and the traceability of supervision, and building an efficient, transparent and safe traditional Chinese medicine decoction piece management system, greatly improving the quality control and optimization level of traditional Chinese medicine products and promoting the intelligent development of traditional Chinese medicine industry. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 The method proposed in the present application is shown in the schematic diagram. Figure 2 The initial ingredient content data set acquisition schematic diagram proposed by the present application; Figure 3 The traditional Chinese medicinal material moisture evaporation curve prediction schematic diagram proposed by the present application; Figure 4 The processing parameter adjustment schematic diagram proposed by the present application; Figure 5 The data space-time correlation mapping schematic diagram proposed by the present application; Figure 6 The processing parameter optimization adjustment schematic diagram proposed by the present application; Figure 7 The blockchain storage schematic diagram proposed by the present application; Figure 8 The medicinal piece quality prediction model schematic diagram proposed by the present application. DETAILED DESCRIPTION

[0017] The following description is used to disclose the present application so that those skilled in the art can implement the present application. The preferred embodiments in the following description are only as examples, and other obvious modifications can be thought of by those skilled in the art.

[0018] The traditional Chinese medicinal piece whole life cycle data management system comprises: The data acquisition module: the module is used for acquiring environmental parameter and geographic information data of traditional Chinese medicinal material planting link in real time, and acquiring initial ingredient data through near-infrared spectrum detection; The processing environment monitoring module: the module is used for monitoring temperature and humidity environment in the processing link of traditional Chinese medicinal materials in real time, predicting moisture evaporation curve based on LSTM neural network, and optimizing processing quality through deviation early warning mechanism; The processing technology optimization module: the module constructs chemical composition conversion knowledge graph according to historical processing data and Raman spectrum detection results, adjusts processing technology in real time, and ensures effective ingredient conversion of traditional Chinese medicinal materials; The sampling data module: the module detects effective ingredients and pesticide residue indexes of traditional Chinese medicinal pieces by high-resolution mass spectrometry, and performs space-time correlation with batch number and processing technology parameters; The clinical test module: the module is used for collecting patient data and curative effect evaluation of clinical use end, and optimizing processing technology of traditional Chinese medicinal pieces through correlation analysis with processing technology parameters; The blockchain management module: the module generates time stamp hash value of data in each link in the whole life cycle of traditional Chinese medicinal pieces, ensures real-time synchronization and transparent traceability of data in each link through blockchain and cross-chain protocol, constructs graph neural network model through life cycle data lake, performs medicinal piece quality prediction, and gives improvement suggestions; The processor: the processor is used for processing calculation process of each formula and construction calculation process of each model.

[0019] Referring to Figure 1 As shown in the Chinese medicine decoction piece whole life cycle data management method, comprising: Step one: based on the planting link of Chinese herbal medicine, the growth environment parameters are collected in real time through the sensor, the geographical information data of the planting area is obtained, the initial ingredient content dataset is obtained through near-infrared spectrum detection of Chinese herbal medicine, and the planting main body, harvesting time and geographical position data are encrypted and stored by the blockchain node; Step two: based on the processing link of Chinese herbal medicine, the processing environment is monitored in real time through the temperature and humidity sensor, the water evaporation model based on LSTM neural network is constructed, the water evaporation curve of Chinese herbal medicine is dynamically predicted, and the processing deviation early warning mechanism is established; Step three: based on the processing link of Chinese herbal medicine, the chemical component transformation knowledge graph of Chinese herbal medicine is constructed according to the historical Chinese herbal medicine processing data, and the processing parameters are adjusted based on the Raman spectrum detection result; Step four: based on the finished Chinese medicine decoction piece sampling link, the effective ingredient content and heavy metal pesticide residue index data in the finished Chinese medicine decoction piece are detected by high-resolution mass spectrometry, and the detection result is time and space correlation mapping with batch number and processing technology parameters; Step five: based on the sampling qualified Chinese medicine decoction piece, the prescription compatibility data and patient efficacy evaluation are collected at the clinical use end, the decoction piece clinical test dataset is constructed, and the processing technology parameters are optimized and adjusted; Step six: the time stamp hash value of the key data in each link in the whole life cycle of Chinese medicine decoction piece is written into the alliance chain, a unique blockchain digital identification code is generated for each decoction piece batch, and is synchronized to each link through cross-chain protocol; Step seven: construct the life cycle data lake architecture, integrate the planting, processing, processing, quality inspection, clinical link data, construct the decoction piece quality prediction model based on graph neural network, and output the planting area optimization scheme and processing technology improvement suggestion.

[0020] Referring to Figure 2 As shown in the Chinese medicine decoction piece whole life cycle data management method, comprising: In the planting area of Chinese herbal medicine, multi-source sensors are arranged to monitor the key parameters of the growth environment in real time, including temperature, humidity, soil pH value, light intensity, air quality, obtain geographical position information including land information, soil quality, climate condition data, and form a digital planting area map; The initial data of the quality of the medicine is obtained in real time by near-infrared spectrum in a non-destructive way, and the main ingredient content is measured; Based on the obtained traditional Chinese medicinal material planting link data, the data is stored by encryption through a blockchain technology and uploaded to a blockchain network.

[0021] Specifically, the environmental data collected by the sensor, geographic information, near-infrared spectrum data and medicinal material related information such as planting subject, harvesting time and geographic location are stored in the blockchain network through an encryption algorithm. The data encryption can adopt symmetric encryption or asymmetric encryption technology to ensure the security of the data. The consensus mechanism of the blockchain ensures the authenticity and consistency of the data. After the data is encrypted, it is uploaded to a decentralized network through a blockchain node. Each data upload forms a new block, which is linked to the previous block through a chain structure, ensuring that the data cannot be tampered with.

[0022] Referring to Figure 3 Based on the processing link of traditional Chinese medicinal materials, the processing environment is monitored in real time through a temperature and humidity sensor, a water evaporation model based on an LSTM neural network is constructed, and a dynamic prediction of the water evaporation curve of traditional Chinese medicinal materials is made. The processing deviation early warning mechanism specifically includes: The temperature and humidity sensor in the processing site of traditional Chinese medicinal materials is used to detect and obtain air temperature and humidity environmental parameters in real time; A water evaporation model based on a long short-term memory network is constructed, which is trained by historical processing data to predict the water evaporation curve of traditional Chinese medicinal materials under different environmental conditions; The collected environmental parameter data is constructed into a time series data set and input into the trained water evaporation model to predict the water evaporation of traditional Chinese medicinal materials and generate a real-time water evaporation curve; Based on the prediction result of the water evaporation model, a processing deviation range threshold is set. When the actual moisture content deviates from the predicted curve by more than the threshold, a deviation warning is triggered.

[0023] Specifically, historical processing data is collected, including temperature, humidity and water evaporation rate under different environmental conditions. These data constitute an input and output time series. The temperature and humidity at each time point are used as input features, and the corresponding water evaporation rate is used as an output label to form a time series data set. The LSTM model is trained using historical processing data to learn the influence of environmental parameters such as temperature and humidity on water evaporation. The LSTM can learn long-term and short-term dependencies through its memory cells, thereby more accurately capturing the dynamic changes in the water evaporation process. Through the trained LSTM model, real-time collected temperature and humidity data is input into the model to dynamically predict the water evaporation of traditional Chinese medicinal materials. Based on this prediction, a curve of the change of moisture content of traditional Chinese medicinal materials over time during processing can be generated. In actual processing, if the deviation of the real-time collected moisture content from the prediction curve exceeds the set threshold, the processing deviation early warning mechanism can be triggered.

[0024] Referring to Figure 4 As shown, based on the processing link of traditional Chinese medicinal materials, a traditional Chinese medicine chemical component transformation knowledge graph is constructed according to historical traditional Chinese medicinal material processing data, and the processing parameters are adjusted based on the Raman spectrum detection results, which specifically includes: Based on the historical processing data, the chemical component transformation knowledge graph is constructed by collecting the chemical component transformation information of traditional Chinese medicinal materials under different processing technologies. The main chemical components of traditional Chinese medicinal materials are taken as nodes, and the possible paths and reaction conditions of component transformation are taken as edges to construct a traditional Chinese medicine chemical component transformation knowledge graph. Real-time collection of Raman spectrum data of traditional Chinese medicinal material samples in the processing process is matched with the chemical component nodes in the traditional Chinese medicine chemical component transformation knowledge graph to obtain the chemical component content and transformation relationship of the traditional Chinese medicinal material samples. Based on the results of Raman spectrum matching, the transformation rate of each chemical component in the processing process is calculated, and the transformation of key components is particularly concerned. Based on the obtained transformation rate results, the processing technology of the transformation rate deviating from the target interval is adjusted, including adjusting the frying temperature and cooking time.

[0025] Specifically, based on historical processing data, the chemical component transformation information of traditional Chinese medicinal materials under different processing technologies is collected, including: Processing technology: such as frying, cooking, drying; Chemical component transformation: under different processing conditions, how the main chemical components of medicinal materials are transformed, for example, some effective components may be decomposed or transformed into other chemical substances at high temperature; Reaction conditions: such as temperature, humidity, time; Through historical data analysis, the transformation path and transformation rate of chemical components under each processing technology are determined; The main chemical components of traditional Chinese medicinal materials are taken as nodes, and the transformation relationship and reaction conditions between components are taken as edges to construct a traditional Chinese medicine chemical component transformation knowledge graph. Each chemical component, such as effective component, main active component, and impurity, is taken as a node in the knowledge graph, and the transformation path and corresponding reaction conditions, such as temperature, time, and humidity, are taken as the attributes of the edge. The weight of the edge is assigned according to experimental data, representing the probability or transformation rate of transformation, and the reaction conditions are taken as one of the conditions of the edge. During the processing, the spectral data of the medicinal material sample is collected in real time by a Raman spectrometer. The Raman spectrum can reveal the main chemical components in the medicinal material and their changes. According to the collected spectral data, the corresponding chemical component nodes in the knowledge graph are matched to obtain the chemical component content and conversion relationship of the current sample. The Raman spectrum reflects the vibration characteristics of the chemical components in the medicinal material sample. By analyzing the Raman spectrum, the content of various chemical components in the medicinal material can be quantitatively obtained. According to the known relationship between the Raman spectrum and the chemical component, the real-time collected spectral data is matched with the chemical component nodes in the knowledge graph to obtain the current chemical component content of the medicinal material; Based on the matching results of the Raman spectrum, the conversion rate of each chemical component in the processing is calculated. According to the conversion rate results, especially the conversion of key components, the processing technology is adjusted to ensure that the component conversion meets the target requirements. The adjustment parameters include: Frying temperature: If the conversion rate of the key component is too high or too low, the frying temperature needs to be adjusted; Cooking time: If some components are not completely converted, the cooking time needs to be extended or other reaction conditions need to be adjusted.

[0026] Referring to Figure 5 As shown, based on the inspection of finished traditional Chinese medicine decoction pieces, the effective component content and heavy metal and pesticide residue index data in the finished traditional Chinese medicine decoction pieces are detected by high-resolution mass spectrometry. The detection results are spatiotemporally associated with batch number and processing parameters, which specifically includes: Randomly select samples from each batch of finished traditional Chinese medicine decoction pieces by high-resolution mass spectrometry to obtain the content of main effective components, heavy metal elements and pesticide residues in the traditional Chinese medicine decoction pieces; The mass spectrometry detection results are recorded in real time, spatiotemporally associated with the production data of each batch of traditional Chinese medicine decoction pieces, and a spatiotemporal association database is constructed to store data including batch number, production process, spatiotemporal information and mass spectrometry detection results.

[0027] Referring to Figure 6 As shown, based on the qualified traditional Chinese medicine decoction pieces, prescription compatibility data and patient efficacy evaluation are collected at the clinical use end to construct a decoction piece clinical test data set for optimizing and adjusting the processing parameters. Specifically includes: At the clinical use end of the traditional Chinese medicine decoction pieces, the patient's medication information, prescription compatibility data, medication time, dosage and treatment course data are collected to construct a clinical data set; Based on each batch of traditional Chinese medicine decoction pieces in the clinical data set, the processing parameter data of the batch is associated to integrate and construct a decoction piece clinical test data set, including patient basic information, prescription compatibility information, patient efficacy evaluation data and traditional Chinese medicine decoction piece processing parameters; Through the random forest algorithm, the basic information of the patient, the prescription compatibility information and the processing parameters of the traditional Chinese medicine decoction pieces are taken as characteristic variables, and the curative effect evaluation data of the patient is taken as the target variable to model and train the model; Based on the trained random forest model, the association rules between the process parameters and the curative effect are obtained by evaluating the importance of each feature, and the processing parameters are optimized.

[0028] Specifically, at the clinical use end of the traditional Chinese medicine decoction pieces, the patient's medication information, prescription compatibility data, medication time, dosage and treatment course data need to be collected, including: Basic information of the patient: such as age, gender, body type; Prescription compatibility data: including the variety and dosage of traditional Chinese medicine decoction pieces used by each patient; Medication time: the daily medication time and medication cycle of the patient; Dosage and treatment course: the dosage of each time and the use time period of the drug; The random forest algorithm is used to model and analyze the relationship between the processing parameters of the traditional Chinese medicine decoction pieces and the curative effect of the patient, the basic information of the patient, the prescription compatibility information and the processing parameters of the traditional Chinese medicine decoction pieces are taken as characteristic variables input into the model, and the curative effect evaluation of the patient is taken as the target variable for training the model; After the model training is completed, the random forest algorithm can evaluate the influence of each characteristic variable on the curative effect and give the characteristic importance, which reflects the contribution degree of each characteristic to the model output. According to the evaluation result, which processing parameters, prescription compatibility information and patient basic information have greater influence on the curative effect; Based on the trained random forest model, the influence of each feature on the curative effect is analyzed, so as to optimize and adjust the processing technology, for example, if the temperature has the greatest influence on the curative effect, the processing temperature can be adjusted; if the time has a significant influence on the curative effect, the processing time can be appropriately extended or shortened.

[0029] Referring to Figure 7 As shown in the figure, the time stamp hash value of the key data in each link of the whole life cycle of the traditional Chinese medicine decoction pieces is written into the alliance chain, a unique blockchain digital identification code is generated for each batch of decoction pieces, and is synchronized to each link through the cross-chain protocol, which specifically includes: Based on the data in each link of the whole life cycle of the traditional Chinese medicine decoction pieces, a time stamp is generated when the data is collected, and a hash value is calculated by hashing the data; A unique hash value is generated for each batch of traditional Chinese medicine decoction pieces at each link of the life cycle, and the hash value and the link data are created into a block; A unique digital identification code is generated for each batch in the alliance chain as a digital identity, and each block is written into the blockchain in chronological order; Based on the information flow requirements between each link, a cross-chain protocol is designed, and through the cross-chain protocol, the data of each link is synchronized to the blockchains of other related links.

[0030] Specifically, each batch of traditional Chinese medicine decoction pieces generates a unique hash value at each link in the life cycle, and the hash value is formed with the link data to form a block, which contains the data and hash value of the link. When calculating the hash value of the block, the hash value of the previous block is included to form a chain structure, ensuring the continuity and tamper resistance of the data, and the formula is: Wherein, is the block generated by the ith link, is the hash value of the previous link, is the timestamp of the current link, is the data of the current link; According to the batch number of traditional Chinese medicine decoction pieces and the block information of all links in the life cycle, a unique digital identification code is generated, and the data of each link and its digital identification code need to be written into the blockchain in chronological order to ensure the order and integrity of the data. Since the whole life cycle of traditional Chinese medicine decoction pieces involves multiple links, there may be different blockchain systems, and the design of cross-chain protocol can ensure that the data of different links can be synchronized between different chains. Whenever a block of a link is generated, the cross-chain protocol will send the hash value and related information of the block to other blockchains that need to be synchronized. The receiving chain will verify the legality of the block and record its data in the corresponding blockchain after receiving the data.

[0031] Referring to Figure 8 , a life cycle data lake architecture is constructed, and planting, processing, processing, quality inspection, and clinical link data are integrated to build a decoction piece quality prediction model based on graph neural network, and output planting area optimization scheme and processing process improvement suggestions, which specifically include: Construct a life cycle data lake based on planting, processing, processing, quality inspection, and clinical link data; According to the links of the life cycle, each link is regarded as a node in the graph, and the nodes are connected through the relationship between the links to form edges in the graph, and the graph is constructed. The integrated life cycle data is converted into a graph form and input into the decoction piece quality prediction model based on graph neural network, the relationship between each node is learned through training the graph neural network, and the decoction piece quality is predicted. Based on the decoction piece quality prediction results of the model, the quality of medicinal materials in different planting areas, the optimal planting area, and the processing process optimization suggestions are obtained.

[0032] Specifically, each link is regarded as a node in the graph, and edges are formed through the relationship between links, such as planting to processing, processing to processing, etc. Different nodes are connected through edges, and the weight of the edge can be determined according to the correlation or causal relationship between links, for example, the planting area and the processing technology may have a direct correlation, and the quality inspection and the clinical effect may have an indirect correlation. Through the graph convolution operation, the feature vector of the node is updated to a new representation, and the information is continuously propagated through multiple layers of GCN. Through the trained graph neural network, the input life cycle graph, the model can predict the quality of each node, i.e. the quality of each link; According to the output of the model, the predicted values of the quality of medicinal materials in different planting areas are compared to select a planting area with higher quality, for example, the planting conditions in a certain area may result in better quality of medicinal materials, and the model can prompt the optimal planting area; According to the output of the model, the predicted values of the quality of medicinal materials in different planting areas are compared to select a planting area with higher quality, for example, the planting conditions in a certain area may result in better quality of medicinal materials, and the model can prompt the optimal planting area.

[0033] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. And the above describes a specific embodiment of the present application. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or may be advantageous.

[0034] Each embodiment in the specification is described in a progressive manner, and the same or similar parts between each embodiment can be referred to each other. Each embodiment focuses on the difference from other embodiments.

[0035] The above is only the preferred embodiment of the present application, and does not limit the present application. Any modification, equivalent replacement, improvement, etc. within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A traditional Chinese medicine decoction piece whole life cycle data management method, characterized in that, The application comprises: Based on the planting link of traditional Chinese medicinal materials, the growth environment parameters are collected in real time by sensors, the geographical information data of the planting area is obtained, the initial ingredient content dataset is obtained by near-infrared spectrum detection of traditional Chinese medicinal materials, and the planting subject, harvesting time and geographical position data are encrypted and stored by a blockchain node; Based on the processing link of traditional Chinese medicinal materials, the processing environment is monitored in real time by a temperature and humidity sensor, a water evaporation model based on an LSTM neural network is constructed, the water evaporation curve of traditional Chinese medicinal materials is dynamically predicted, and a processing deviation early warning mechanism is established; Based on the processing link of traditional Chinese medicinal materials, the processing link of traditional Chinese medicinal materials is constructed according to historical traditional Chinese medicinal material processing data, and the processing parameters are adjusted based on Raman spectrum detection results; Based on the sampling inspection link of traditional Chinese herbal pieces, the effective ingredient content and heavy metal pesticide residue index data of the traditional Chinese herbal pieces are detected by high-resolution mass spectrometry, and the detection results are time and space correlated and mapped with batch numbers and processing parameters; Based on the sampling inspection qualified traditional Chinese herbal pieces, prescription compatibility data and patient efficacy evaluation are collected at the clinical use end, a herbal piece clinical test dataset is constructed, and the processing parameters are optimized and adjusted; The time stamp hash value of the key data in each link of the whole life cycle of traditional Chinese herbal pieces is written into the alliance chain, a unique blockchain digital identification code is generated for each herbal piece batch, and the cross-chain protocol is synchronized to each link; A life cycle data lake architecture is constructed, the planting, processing, processing, quality inspection and clinical link data are integrated, a herbal piece quality prediction model based on a graph neural network is constructed, and a planting area optimization scheme and processing technology improvement suggestion are output.

2. The traditional Chinese medicine decoction piece whole life cycle data management method according to claim 1, characterized in that, The application comprises: A plurality of source sensors are arranged in the planting area of traditional Chinese medicinal materials, the key parameters of the growth environment are monitored in real time, including temperature, humidity, soil pH value, light intensity and air quality, geographical position information is obtained, including land information, soil quality and climate condition data, and a digital planting area map is formed; The initial data of the quality of medicinal materials are obtained in real time by near-infrared spectrum in a non-destructive manner, and the main ingredient content is measured; Based on the obtained data of the planting link of traditional Chinese medicinal materials, the data are encrypted and stored by a blockchain technology, and uploaded to a blockchain network.

3. The traditional Chinese medicine decoction piece whole life cycle data management method according to claim 1, characterized in that, The application comprises: The air temperature and humidity environment parameters are detected in real time by a temperature and humidity sensor in a traditional Chinese medicinal material processing site; A water evaporation model based on a long short-term memory network is constructed, the model is trained by historical processing data, and the water evaporation curve of traditional Chinese medicinal materials under different environmental conditions is predicted; The collected environmental parameter data is constructed into a time series data set and input into the trained moisture evaporation model to predict the moisture evaporation of the traditional Chinese medicinal material and generate a real-time moisture evaporation curve; Based on the prediction result of the moisture evaporation model, a processing deviation range threshold is set, and when it is monitored that the actual moisture content deviates from the prediction curve by more than the threshold, a deviation warning is triggered.

4. The traditional Chinese medicine decoction piece whole life cycle data management method according to claim 1, characterized in that, Based on the processing link of the traditional Chinese medicinal material, a traditional Chinese medicine chemical component transformation knowledge graph is constructed according to historical traditional Chinese medicinal material processing data, and the processing parameters are adjusted based on the Raman spectrum detection result, specifically including: Based on the historical processing data, the chemical component transformation relationship between the components is obtained by collecting the chemical component change information of the traditional Chinese medicinal material under different processing techniques; The main chemical components of the traditional Chinese medicinal material are taken as nodes, and the possible paths and reaction conditions of the component transformation are taken as edges to construct a traditional Chinese medicine chemical component transformation knowledge graph; Real-time Raman spectrum data of the traditional Chinese medicinal material sample in the processing process is collected, matched with the chemical component nodes in the traditional Chinese medicine chemical component transformation knowledge graph, and the chemical component content and transformation relationship of the traditional Chinese medicinal material sample are obtained; Based on the result of the Raman spectrum matching, the transformation rate of each chemical component in the processing process is calculated, and the transformation of the key components is particularly concerned; Based on the obtained transformation rate result, the processing technique of the traditional Chinese medicinal material is adjusted, including adjusting the frying temperature and the cooking time.

5. The traditional Chinese medicine decoction piece whole life cycle data management method according to claim 1, characterized in that, Based on the finished product sampling link of the traditional Chinese herbal pieces, the effective component content and heavy metal and pesticide residue index data of the traditional Chinese herbal pieces are detected by high-resolution mass spectrometry, and the detection result is spatiotemporally associated and mapped with the batch number and processing parameters, specifically including: Random samples of each batch of traditional Chinese herbal pieces are extracted by high-resolution mass spectrometry to obtain the main effective component content, heavy metal element content and pesticide residue content in the traditional Chinese herbal pieces; The mass spectrometry detection result is recorded in real time, spatiotemporally associated with the production data of each batch of traditional Chinese herbal pieces, and a spatiotemporal association database is constructed to store the batch number, production process, spatiotemporal information and mass spectrometry detection result data.

6. The traditional Chinese medicine decoction piece whole life cycle data management method according to claim 1, characterized in that, Based on the sampling qualified traditional Chinese herbal pieces, prescription compatibility data and patient efficacy evaluation are collected at the clinical use end to construct a herbal piece clinical test data set, and the processing parameters are optimized and adjusted, specifically including: At the clinical use end of the traditional Chinese herbal pieces, patient medication information, prescription compatibility data, medication time, dosage and treatment course data are collected to construct a clinical data set; Based on each batch of traditional Chinese herbal pieces in the clinical data set, the processing parameter data of the batch is associated to integrate and construct a herbal piece clinical test data set, including patient basic information, prescription compatibility information, patient efficacy evaluation data and traditional Chinese herbal piece processing parameters; Through a random forest algorithm, patient basic information, prescription compatibility information and traditional Chinese herbal piece processing parameters are taken as characteristic variables, and patient efficacy evaluation data is taken as a target variable for modeling and model training; Based on the trained random forest model, the importance of each feature is evaluated to obtain the association rules of the processing parameters and the efficacy, and the processing parameters are optimized.

7. The traditional Chinese medicine decoction piece whole life cycle data management method according to claim 1, characterized in that, The key data in each link of the whole life cycle of traditional Chinese medicine decoction pieces is timestamped and hashed, and written into the alliance chain, a unique blockchain digital identification code is generated for each decoction piece batch, and is synchronized to each link through a cross-chain protocol, which specifically includes: Based on the data in each link of the whole life cycle of traditional Chinese medicine decoction pieces, a timestamp is generated when the data is collected, and a hash value is calculated by hashing the data; A unique hash value is generated for each batch of traditional Chinese medicine decoction pieces at each link of the life cycle, and the hash value and link data are created into a block; A unique digital identification code is generated for each batch in the alliance chain as a digital identity, and each block is written into the blockchain in chronological order; Based on the information flow between links, a cross-chain protocol is designed, and the data of each link is synchronized to the blockchain of other related links through the cross-chain protocol.

8. The traditional Chinese medicine decoction piece whole life cycle data management method according to claim 1, characterized in that, The construction of the life cycle data lake architecture integrates planting, processing, processing, quality inspection, and clinical link data, and constructs a decoction piece quality prediction model based on graph neural network, and outputs planting area optimization scheme and processing process improvement suggestion, which specifically includes: Based on the planting, processing, processing, quality inspection, and clinical link data, a life cycle data lake is constructed; According to the links of the life cycle, each link is regarded as a node in the graph, and the nodes are connected through the relationship between the links to form the edges in the graph, and the graph is constructed; The integrated life cycle data is converted into a graph form and input into the decoction piece quality prediction model based on graph neural network, the relationship between the nodes is learned through training the graph neural network, and the decoction piece quality is predicted; Based on the decoction piece quality prediction result of the model, the quality of medicinal materials in different planting areas, the optimal planting area, and the processing process optimization suggestion are obtained.

9. The traditional Chinese medicine decoction piece whole life cycle data management method according to claim 1, characterized in that, It also includes a hierarchical architecture of the blockchain: a data layer for storing IPFS hash pointers of raw data in each link; a contract layer for deploying smart contracts to realize data upload permission control and automatic generation of audit logs; an application layer for providing API interfaces for regulatory agencies to check data integrity, supporting 1-second traceability query response.

10. A traditional Chinese medicine decoction piece whole life cycle data management system for implementing the traditional Chinese medicine decoction piece whole life cycle data management method according to any one of claims 1-9, characterized in that, It includes: Data acquisition module: The module is used to collect environmental parameters and geographic information data in the planting link of medicinal materials in real time, and to obtain initial ingredient data through near-infrared spectrum detection; Processing environment monitoring module: The module is used to monitor the temperature and humidity environment in the processing link of traditional Chinese medicinal materials in real time, predict the moisture evaporation curve based on LSTM neural network, and optimize the processing quality through deviation early warning mechanism; Processing technology optimization module: The module constructs a chemical composition transformation knowledge graph according to historical processing data and Raman spectrum detection results, and adjusts the processing technology in real time to ensure the transformation of effective components of traditional Chinese medicinal materials; Sampling data module: The module uses high-resolution mass spectrometry to detect the effective components and pesticide residue indicators of traditional Chinese medicine decoction pieces, and performs spatiotemporal correlation with batch number and processing parameters; Clinical test module: The module is used to collect patient data and efficacy evaluation from the clinical use end, and optimize the processing technology of traditional Chinese medicine decoction pieces through correlation analysis with processing parameters; A blockchain management module: the module generates a timestamp hash value of data at each link in the whole life cycle of traditional Chinese medicine decoction pieces, and ensures real-time synchronization and transparent traceability of the data at each link through a blockchain and a cross-chain protocol, constructs a graph neural network model through a life cycle data lake, performs decoction piece quality prediction, and gives improvement suggestions; A processor: the processor is used for processing the calculation process of each formula and the construction calculation process of each model.

Images