Dosage regulation and control data storage method and system based on multi-modal diagnosis and treatment data

Abstract

The invention relates to the technical field of medical information, and discloses a dosage regulation and control data storage method and system based on multi-modal diagnosis and treatment data, and the method comprises the steps: obtaining historical diagnosis and treatment data and historical dosage data of a patient; matching the diagnosis result data with health diagnosis result data in the health diagnosis and treatment data, and matching the wearable device data with health wearable device data in the health diagnosis and treatment data; determining whether to adjust the historical dose data to obtain the current dose, and outputting a judgment result needing to be adjusted; and obtaining the adjusted dose data, and storing the diagnosis result data, the diagnosis result similarity, the wearable device data, the wearable data similarity, the historical dose data and the adjusted dose data into a dose adjustment unit in a patient database. According to the invention, multi-modal diagnosis and treatment data can be integrated and applied to dose regulation and control, and a complete medication decision data chain is constructed.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of medical information technology, in particular to a drug dosage regulation data storage method and system based on multi-modal diagnosis and treatment data. BACKGROUND

[0002] Traditional drug dosage adjustment often relies on the clinical experience of doctors and limited examination indicators, such as single-dimensional data such as symptom description and blood test, which is difficult to fully and dynamically reflect the health status of patients. This method not only lacks systematicness and scientificity, but also is prone to individual differences, affecting the treatment effect, and even may cause adverse drug reactions.

[0003] With the development of medical technology, multi-modal diagnosis and treatment data has been gradually applied in clinical practice, including tongue analysis, facial diagnosis, pulse recognition and other data characteristic of traditional Chinese medicine, as well as physiological indicators such as heart rate, blood pressure and sleep quality collected by wearable devices. These multi-source heterogeneous data contain rich patient health information, but how to effectively integrate and apply them to drug dosage regulation still faces many challenges. In the prior art, the use of multi-modal diagnosis and treatment data is scattered, lacking a unified data processing and analysis framework, making it difficult to fully realize the value of these data and provide comprehensive and accurate basis for drug dosage adjustment. At the same time, in terms of data storage, traditional methods usually only record the diagnosis and treatment data and the drug results of patients, ignoring the correlation between data and the key analysis results used for drug dosage adjustment, such as the similarity between diagnosis results and health status, and the matching degree of different modal data. This makes the subsequent disease analysis and treatment plan optimization lack complete data support, which is not conducive to the summary of medical experience and the development of medical technology. In addition, in the management of patient data, there are problems such as scattered information and inconvenience in retrieval, which affects the efficiency and quality of medical services.

[0004] Therefore, it is necessary to provide a drug dosage regulation data storage method and system based on multi-modal diagnosis and treatment data to solve the problems of how to integrate multi-modal diagnosis and treatment data and apply it to drug dosage regulation, lack of unified processing framework, data storage method ignoring key analysis results, and scattered patient data management. SUMMARY

[0005] In view of this, the present application provides a drug dosage regulation data storage method and system based on multi-modal diagnosis and treatment data, aiming to solve the problems of how to integrate multi-modal diagnosis and treatment data and apply it to drug dosage regulation, lack of unified processing framework, data storage method ignoring key analysis results, and scattered patient data management.

[0006] In one aspect, the present application provides a drug dosage regulation data storage method based on multi-modal diagnosis and treatment data, comprising:

[0007] Obtaining basic information and diagnosis and treatment data of a patient, performing information matching in a patient database according to the basic information, and obtaining historical diagnosis and treatment data and historical dosage data of the patient; wherein the diagnosis and treatment data comprises diagnosis result data and wearable device data;

[0008] Matching the diagnosis result data with health diagnosis result data in health diagnosis and treatment data to obtain a diagnosis result similarity, and matching the wearable device data with health wearable device data in the health diagnosis and treatment data to obtain a wearable data similarity;

[0009] Determining whether to adjust the historical dosage data to obtain a current dosage according to the diagnosis result similarity and the wearable data similarity, and outputting a judgment result of the adjustment needed if it is determined that the adjustment is needed;

[0010] Obtaining the adjusted dosage data, and storing the diagnosis result data, the diagnosis result similarity, the wearable device data, the wearable data similarity, the historical dosage data and the adjusted dosage data into a dosage adjustment unit in the patient database.

[0011] Further, when obtaining the basic information and the diagnosis and treatment data of the patient, performing information matching in the patient database according to the basic information, and obtaining the historical diagnosis and treatment data and the historical dosage data of the patient, the method comprises:

[0012] Inputting the basic information of the patient into the patient database for query and matching;

[0013] If the matching is successful, obtaining the historical diagnosis and treatment data and the historical dosage data of the patient;

[0014] If the matching is not successful, establishing the basic information of the patient in the patient database, and inputting the diagnosis and treatment data and the dosage data of the patient.

[0015] Further, the diagnosis result data comprises tongue appearance analysis data, facial diagnosis data, pulse appearance recognition data, interrogation data and corneal information recognition data;

[0016] The wearable device data comprises heart rate data, blood pressure data and sleep quality data.

[0017] Further, when matching the diagnosis result data with health diagnosis result data in health diagnosis and treatment data to obtain a diagnosis result similarity, the method comprises:

[0018] Respectively performing feature extraction on the tongue appearance analysis data, the facial diagnosis data, the pulse appearance recognition data, the interrogation data and the corneal information recognition data, and constructing a multi-dimensional feature vector;

[0019] For tongue appearance analysis data, extract tongue color, tongue fur color and tongue shape feature parameters, and calculate the tongue appearance feature similarity between the current tongue appearance and healthy tongue appearance through cosine similarity algorithm;

[0020] For facial diagnosis data, use convolutional neural network to extract facial skin color, texture and morphological feature points, and use structural similarity index to evaluate the facial state similarity between the current facial features and healthy facial features;

[0021] For pulse recognition data, extract pulse wave frequency domain features, time domain features and nonlinear dynamics features, and calculate the pulse waveform similarity with healthy pulse through dynamic time warping algorithm;

[0022] For inquiry data, construct a TCM symptom library and semantic network, and calculate the semantic similarity between the current symptom description and healthy inquiry description through word vector model;

[0023] For corneal information recognition data, extract corneal transparency, blood vessel morphology and nerve distribution feature parameters, and use fuzzy clustering algorithm to calculate the corneal state similarity with healthy cornea;

[0024] Sum the tongue feature similarity, facial state similarity, pulse waveform similarity, semantic similarity and corneal state similarity to calculate the diagnosis result similarity.

[0025] Further, when the wearable device data is matched with the healthy wearable device data in the health diagnosis and treatment data to obtain the wearable data similarity, it includes:

[0026] For heart rate data, construct a 24-hour dynamic heart rate curve, extract resting heart rate, exercise heart rate variability and heart rate recovery time, and calculate the heart rate data similarity between the current heart rate curve and healthy heart rate curve through Pearson correlation coefficient;

[0027] For blood pressure data, extract systolic blood pressure fluctuation trend, diastolic blood pressure circadian rhythm and pulse pressure difference, and use time series analysis method to evaluate the blood pressure data similarity with healthy blood pressure pattern;

[0028] For sleep quality data, extract sleep latency, sleep efficiency, sleep stage proportion and micro-arousal index based on polysomnogram data, and use support vector machine classifier to calculate the sleep data similarity with healthy sleep structure;

[0029] Sum the heart rate data similarity, blood pressure data similarity and sleep data similarity to obtain the wearable data similarity.

[0030] Further, when the diagnosis result similarity and the wearable data similarity are used to determine whether to adjust the historical drug amount data to obtain the current drug amount, it includes:

[0031] The diagnostic result similarity is compared with a historical diagnostic result similarity to obtain a first comparison result;

[0032] The wearing data similarity is compared with a historical wearing data similarity to obtain a second comparison result;

[0033] According to the first comparison result and the second comparison result, whether to adjust historical dosage data to obtain current dosage is determined.

[0034] Further, when determining whether to adjust the historical dosage data according to the first comparison result and the second comparison result to obtain the current dosage, the method comprises:

[0035] A difference between the diagnostic result similarity and the historical diagnostic result similarity is calculated, and is recorded as a diagnostic similarity difference; and a difference between the wearing data similarity and the historical wearing data similarity is calculated, and is recorded as a wearing similarity difference;

[0036] If the diagnostic similarity difference and the wearing similarity difference are both greater than or equal to zero, it is preliminarily determined that the historical dosage data is not adjusted;

[0037] Otherwise, it is determined that the historical dosage data needs to be adjusted to increase the dosage.

[0038] Further, if the diagnostic similarity difference and the wearing similarity difference are both greater than or equal to zero, the method further comprises:

[0039] A difference threshold is set, and if the diagnostic similarity difference and the wearing similarity difference are both greater than the difference threshold, it is determined that the historical dosage data needs to be adjusted to reduce the dosage;

[0040] Otherwise, it is determined that the historical dosage data is not adjusted.

[0041] Further, when determining whether to adjust the historical dosage data, the method further comprises:

[0042] The determination result of whether to adjust the historical dosage data is output and displayed, and the determination result of increasing the dosage and reducing the dosage is output.

[0043] Compared with the prior art, the present application has the beneficial effects that from the perspective of diagnosis and treatment accuracy, by integrating diagnosis result data and wearable device data, the patient's health status can be comprehensively and stereoscopically reflected. By matching these data with health diagnosis and treatment data to obtain similarity, the deviation degree of the current state and the health state of the patient can be accurately judged, a scientific basis for dose adjustment is provided, medication deviation caused by single data judgment is avoided, and the accuracy and effectiveness of treatment are greatly improved. In terms of data utilization and storage, not only the diagnosis and treatment data of the patient are recorded, but also key information such as diagnosis result similarity and wearable data similarity are stored together to construct a complete medication decision data chain. These data can provide rich historical reference for subsequent disease analysis and treatment plan optimization, facilitate doctors to summarize treatment experience, and also help medical researchers to mine data value and explore more effective medication rules. From the perspective of patient management, according to the matching of patient basic information and historical data, new patients are automatically filed and old patients quickly retrieve information, realizing the orderly management of patient data. At the same time, the judgment result of dose adjustment is clearly output, which facilitates medical staff to timely grasp the patient's medication adjustment demand, improves the efficiency and quality of medical services, and also provides a strong guarantee for the patient's personalized and continuous treatment.

[0044] In another aspect, the present application also provides a dose regulation data storage system based on multi-modal diagnosis and treatment data, comprising:

[0045] The acquisition module is configured to obtain patient basic information and diagnosis and treatment data, match information in the patient database according to the basic information, and obtain patient historical diagnosis and treatment data and historical dose data; wherein the diagnosis and treatment data includes diagnosis result data and wearable device data;

[0046] The matching module is configured to match the diagnosis result data with health diagnosis result data in the health diagnosis and treatment data to obtain diagnosis result similarity, and match the wearable device data with health wearable device data in the health diagnosis and treatment data to obtain wearable data similarity;

[0047] The output display module is configured to determine whether to adjust the historical dose data to obtain the current dose according to the diagnosis result similarity and the wearable data similarity, and if it is determined to adjust, output the judgment result that needs to be adjusted;

[0048] The storage module is configured to obtain the adjusted dose data, and store the diagnosis result data, the diagnosis result similarity, the wearable device data, the wearable data similarity, the historical dose data and the adjusted dose data into the dose adjustment unit in the patient database.

[0049] It can be understood that the dose regulation data storage method and system based on multi-modal diagnosis and treatment data provided by the present application have the same beneficial effects, which will not be repeated here. Attached Figure Description

[0050] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0051] Figure 1 A flowchart illustrating a method for storing drug dosage control data based on multimodal diagnostic and treatment data provided in an embodiment of the present invention;

[0052] Figure 2 This is a functional block diagram of a drug dosage control data storage system based on multimodal diagnostic and treatment data provided in an embodiment of the present invention. Detailed Implementation

[0053] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0054] In some embodiments of this application, see Figure 1 As shown, this embodiment provides a data storage method for drug dosage control based on multimodal diagnosis and treatment data, including the following steps:

[0055] S100. Obtain basic patient information and treatment data; perform information matching in the patient database based on the basic information to obtain the patient's historical treatment data and historical medication data; wherein, the treatment data includes diagnostic result data and wearable device data;

[0056] S200. Match the diagnostic result data with the health diagnostic result data in the health diagnosis data to obtain the diagnostic result similarity; match the wearable device data with the health wearable device data in the health diagnosis data to obtain the wearable data similarity.

[0057] S300. Based on the similarity of the diagnostic results and the similarity of the wearable data, determine whether to adjust the historical medication data to obtain the current medication dosage. If it is determined that adjustment is required, output the determination result that needs to be adjusted.

[0058] S400, obtaining the adjusted dosage data, storing the diagnosis result data, the diagnosis result similarity, the wearable device data, the wearable data similarity, the historical dosage data and the adjusted dosage data into a dosage adjustment unit in the patient database.

[0059] It can be understood that, from the perspective of diagnosis and treatment accuracy, by integrating the diagnosis result data and the wearable device data, the health status of the patient can be comprehensively and stereoscopically reflected. By matching these data with the health diagnosis and treatment data to obtain the similarity, the deviation degree of the current state and the health state of the patient can be accurately judged, a scientific basis for dosage adjustment can be provided, medication deviation caused by single data judgment can be avoided, and the accuracy and effectiveness of treatment can be greatly improved. In terms of data utilization and storage, not only the diagnosis and treatment data of the patient are recorded, but also the diagnosis result similarity, the wearable data similarity and other key information are stored, and a complete medication decision data chain is constructed. These data can provide rich historical reference for subsequent disease analysis and treatment scheme optimization, facilitate doctors to summarize treatment experience, and also help medical researchers to mine data value and explore more effective medication rules. From the perspective of patient management, according to the matching of the historical data based on the basic information of the patient, the new patient is automatically filed, and the information of the old patient is quickly retrieved, so that the ordered management of the patient data is realized. At the same time, the output of the dosage adjustment judgment result is clear, which facilitates medical staff to timely master the dosage adjustment demand of the patient, improves the efficiency and quality of medical services, and also provides a strong guarantee for the individualized and continuous treatment of the patient.

[0060] In some embodiments of the present application, when the patient basic information and diagnosis and treatment data are obtained, the information matching is performed in the patient database according to the basic information, and the historical diagnosis and treatment data and the historical dosage data of the patient are obtained, the method comprises the following steps:

[0061] The basic information of the patient is input into the patient database for query and matching;

[0062] If the matching is successful, the historical diagnosis and treatment data and the historical dosage data of the patient are obtained;

[0063] If the matching is unsuccessful, the basic information of the patient is established in the patient database, and the diagnosis and treatment data and the dosage data of the patient are input.

[0064] It can be understood that, from the perspective of the integrity and continuity of data management, the identity of new and old patients can be quickly identified by querying and matching in the database by taking the basic information of the patient as an index. For old patients, the historical diagnosis and treatment data and the drug dosage data are directly called, which facilitates the doctor to refer to the past treatment, continue the treatment plan or adjust the medication strategy, and ensures the continuity of treatment; for new patients, the initial diagnosis and treatment and drug dosage data are timely recorded to avoid data omission and ensure that the whole course of treatment data of the patient can be properly recorded, and complete data resources are accumulated for subsequent treatment and research. From the aspect of medical service efficiency, the process has high automation degree, and the medical staff does not need to manually judge whether the patient is a first visit, which simplifies the data acquisition process, reduces the information input time, enables the medical staff to devote more energy to the diagnosis and treatment work, and improves the overall medical service efficiency.

[0065] For example, when an old patient with hypertension re-visits, the medical staff inputs the basic information such as the identity card number and name of the patient into the patient database, and then the past blood pressure monitoring data, medication record and diagnosis and treatment suggestion of Ms. Wang are called out, and the doctor quickly formulates a medication adjustment plan according to the current condition; when a new patient visits for the first time, the system automatically creates an electronic file of the new patient and records the blood pressure value, symptom performance and initial medication dosage at the first diagnosis and treatment after finding no matching information in the query, which lays a foundation for subsequent continuous tracking treatment. This way not only guarantees the orderly management of patient data, but also improves the accuracy and efficiency of medical services.

[0066] In some embodiments of the present application, the diagnosis result data includes tongue appearance analysis data, facial diagnosis data, pulse identification data, interrogation data and corneal information identification data;

[0067] The wearable device data includes heart rate data, blood pressure data and sleep quality data.

[0068] In some embodiments of the present application, when the diagnosis result data is matched with the health diagnosis result data in the health diagnosis and treatment data to obtain the diagnosis result similarity, it includes:

[0069] The tongue appearance analysis data, facial diagnosis data, pulse identification data, interrogation data and corneal information identification data are respectively subjected to feature extraction to construct a multi-dimensional feature vector;

[0070] For the tongue appearance analysis data, the color of tongue body, the color of tongue fur and the tongue shape feature parameters are extracted, and the tongue feature similarity of the current tongue appearance and the health tongue appearance is calculated by a cosine similarity algorithm;

[0071] For the facial diagnosis data, the convolutional neural network is used to extract the facial skin color, texture and morphological feature points, and the structural similarity index is used to evaluate the facial state similarity of the current facial features and the health facial features;

[0072] For pulse recognition data, extract pulse wave frequency domain features, time domain features and nonlinear dynamics features, and calculate the similarity of pulse waveform with healthy pulse through dynamic time warping algorithm;

[0073] For inquiry data, build a TCM symptom library and semantic network, and calculate the semantic similarity of current symptom description and healthy inquiry description through word vector model;

[0074] For corneal information recognition data, extract corneal transparency, blood vessel morphology and nerve distribution feature parameters, and calculate the similarity of corneal state with healthy cornea using fuzzy clustering algorithm;

[0075] The diagnostic result similarity is calculated by weighting and summing the tongue feature similarity, facial state similarity, pulse waveform similarity, semantic similarity and corneal state similarity.

[0076] In some embodiments of the present application, when the wearable device data is matched with the healthy wearable device data in the health diagnosis and treatment data to obtain the wearable data similarity, it includes:

[0077] For heart rate data, construct a 24-hour dynamic heart rate curve, extract resting heart rate, exercise heart rate variability and heart rate recovery time, and calculate the heart rate data similarity of the current heart rate curve with the healthy heart rate curve through Pearson correlation coefficient;

[0078] For blood pressure data, extract systolic blood pressure fluctuation trend, diastolic blood pressure circadian rhythm and pulse pressure difference, and evaluate the blood pressure data similarity with healthy blood pressure pattern using time series analysis method;

[0079] For sleep quality data, extract sleep latency, sleep efficiency, sleep stage proportion and micro-arousal index based on polysomnography data, and calculate the sleep data similarity with healthy sleep structure using support vector machine classifier;

[0080] The heart rate data similarity, blood pressure data similarity and sleep data similarity are weighted and summed to obtain the wearable data similarity.

[0081] It can be understood that the present application deeply mines and accurately analyzes the diagnosis and multi-source heterogeneous data of the wearable device, comprehensively uses various algorithms to process different modal data such as tongue appearance, face, and pulse appearance, avoids the limitations of single data or algorithm, and can fully capture the subtle changes of the patient's health status. For example, through advanced algorithms such as cosine similarity and convolutional neural network, the differences between the current data and the health template are quantified, and complex medical features are converted into comparable numerical values to provide scientific and accurate basis for dosage adjustment. From the clinical application value, the weighted sum calculation of the comprehensive similarity fully considers the importance of each modal data in the disease assessment, so that the final diagnosis result is more in line with the actual needs of the clinic, effectively improves the accuracy and reliability of the medication decision, and helps to realize personalized and precise medical treatment.

[0082] For example, the doctor collects his tongue appearance analysis data and finds that the tongue is dark red and the tongue fur is thick and greasy. Through the cosine similarity algorithm, the tongue appearance feature similarity is lower compared with the healthy tongue appearance template. At the same time, the face diagnosis data is analyzed by using the convolutional neural network, and it is found that the face color is pale and the texture is abnormal, and the face state similarity is also lower than the threshold. In terms of wearable device data, the 24-hour dynamic heart rate curve shows that the resting heart rate is high and the variability is reduced, and the heart rate data similarity calculated by the Pearson correlation coefficient is not ideal. The time series analysis of blood pressure data also shows that the systolic blood pressure fluctuation is abnormal. The weighted sum result of the similarity of each modal data is obtained.

[0083] In some embodiments of the present application, when the diagnosis result similarity and the wearable data similarity are determined, the method comprises:

[0084] Comparing the diagnosis result similarity with the historical diagnosis result similarity to obtain a first comparison result;

[0085] Comparing the wearable data similarity with the historical wearable data similarity to obtain a second comparison result;

[0086] According to the first comparison result and the second comparison result, it is judged whether the historical dosage data is adjusted to obtain the current dosage.

[0087] In some embodiments of the present application, when the diagnosis result similarity and the wearable data similarity are determined, the method comprises:

[0088] The difference between the diagnosis result similarity and the historical diagnosis result similarity is calculated, which is recorded as the diagnosis similarity difference; the difference between the wearable data similarity and the historical wearable data similarity is calculated, which is recorded as the wearable similarity difference;

[0089] If the diagnosis similarity difference value and the wearing similarity difference value are both greater than or equal to zero, it is preliminarily judged that the historical dose data is not adjusted;

[0090] Otherwise, it is judged that the historical dose data needs to be adjusted to increase the dose.

[0091] In some embodiments of the present application, if the diagnosis similarity difference value and the wearing similarity difference value are both greater than or equal to zero, the preliminary judgment that the historical dose data is not adjusted includes:

[0092] A difference threshold value is set, and if the diagnosis similarity difference value and the wearing similarity difference value are both greater than the difference threshold value, it is judged that the historical dose data needs to be adjusted to reduce the dose;

[0093] Otherwise, it is judged that the historical dose data is not adjusted.

[0094] It can be understood that, from the scientificity of medical decision-making, the present application quantifies and compares the similarity of current and historical diagnosis results and the similarity of wearing data, and replaces the pure reliance on experience with a data-driven manner, which can more objectively reflect the patient's disease trend. For example, when the diagnosis similarity difference value and the wearing similarity difference value are both non-negative, it indicates that the patient's health state is stable or has improved, and the preliminary judgment of maintaining the dose can avoid unnecessary dose adjustment and reduce the risk caused by frequent dose changes. The introduction of the difference threshold value further refines the judgment, which can not only capture the need to reduce the dose when the disease significantly improves, but also timely increase the dose when the disease fluctuates, achieving the precision and dynamics of dose adjustment. From the perspective of clinical application, this method provides a clear quantitative reference standard for medical staff, improves the consistency and standardization of medication decision-making, and helps to improve the treatment effect and patient safety.

[0095] For example, when the patient is reexamined, the current diagnosis result similarity of the patient is calculated as 0.75, the historical diagnosis result similarity is 0.7, the diagnosis similarity difference value is 0.05; the wearing data similarity is improved from 0.68 to 0.72, and the wearing similarity difference value is 0.04. Since both difference values are greater than zero, it is preliminarily judged that the dose does not need to be adjusted. Further referring to the set difference threshold value 0.03, both difference values exceed the threshold value, indicating that the patient's health state has improved, so it is judged that the dose should be appropriately reduced, and the judgment result of reducing the dose is output for the doctor to refer to.

[0096] In some embodiments of the present application, when it is judged that adjustment is needed, the judgment result of the adjustment needed is outputted, which includes:

[0097] The judgment result of whether the historical dose data needs to be adjusted is outputted and displayed, and the judgment results of increasing the dose and reducing the dose are outputted.

[0098] It can be understood that, from the clinical operation level, the judgment result of whether to adjust, increase or decrease the drug amount is clear, the complex data analysis conclusion is converted into simple and easy-to-understand information, which greatly reduces the time cost and understanding difficulty of medical staff to obtain key information, and combined with their own clinical experience to quickly develop or adjust the treatment plan, which effectively improves the diagnosis and treatment efficiency. From the perspective of medical quality assurance, the standardized output format avoids medication errors caused by ambiguous information transmission, ensures the accuracy and consistency of drug amount adjustment decision, and provides safer and more reliable medical services for patients. At the same time, the clear result display is also convenient for medical staff to discuss cases and summarize experience, and promotes the sharing and inheritance of medical knowledge.

[0099] For example, through analysis of the patient's multi-modal diagnosis and treatment data, it is calculated that the diagnosis result similarity and the wearing data similarity are significantly decreased compared with the historical data, and it is determined that the drug amount needs to be adjusted. Then, the electronic medical record system and the doctor workstation interface are prompted to "adjust the drug amount", and the "increase the drug amount" is marked. The doctor refers to the prompt text and adjusts the patient's drug amount.

[0100] On the other hand, referring to Figure 2 The application also provides a drug amount adjustment data storage system based on multi-modal diagnosis and treatment data, which is used for applying the above-mentioned drug amount adjustment data storage method based on multi-modal diagnosis and treatment data, and includes:

[0101] The acquisition module is configured to obtain patient basic information and diagnosis and treatment data, and to perform information matching in the patient database according to the basic information to obtain historical diagnosis and treatment data and historical drug amount data of the patient; wherein the diagnosis and treatment data includes diagnosis result data and wearing device data;

[0102] The matching module is configured to match the diagnosis result data with the health diagnosis result data in the health diagnosis and treatment data to obtain a diagnosis result similarity, and to match the wearing device data with the health wearing device data in the health diagnosis and treatment data to obtain a wearing data similarity;

[0103] The output display module is configured to determine whether to adjust the historical drug amount data to obtain a current drug amount according to the diagnosis result similarity and the wearing data similarity, and to output the judgment result that needs to be adjusted if it is determined that adjustment is needed;

[0104] The storage module is configured to obtain the adjusted drug amount data, and to store the diagnosis result data, the diagnosis result similarity, the wearing device data, the wearing data similarity, the historical drug amount data and the adjusted drug amount data into a drug amount adjustment unit in the patient database.

[0105] It can be understood that, from the perspective of clinical operation, the judgment result of whether to adjust, increase or reduce the drug amount is clear, which reduces the time cost and understanding difficulty of medical staff in obtaining key information. Doctors do not need to re-interpret the complex calculation process and data, and can directly formulate or adjust the treatment plan according to the output result combined with their own clinical experience, thereby effectively improving the diagnosis and treatment efficiency. From the perspective of medical quality guarantee, the standardized output format avoids medication errors caused by ambiguous information transmission, ensures the accuracy and consistency of the drug amount adjustment decision, and provides safer and more reliable medical services for patients.

[0106] Those skilled in the art will appreciate that embodiments of the application can be provided as methods, systems or computer program products. Accordingly, the application can be embodied in the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the application can be embodied in the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk memory, CD-ROM, optical memory, etc.) embodying computer usable program code.

[0107] The application is described with reference to flowcharts and / or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks. Figure 1 The functions specified in one or more flows and / or blocks.

[0108] These computer program instructions can also be stored in a computer-readable memory that can direct the computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including instruction means, which implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks. Figure 1 The functions specified in one or more flows and / or blocks.

[0109] These computer program instructions can also be loaded into a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable data processing apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable data processing apparatus provide a process for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks. Figure 1steps of the functions specified in the one or more blocks.

[0110] It should be noted that the above-mentioned embodiments are only used to illustrate the technical solutions of the present application, but not limit the present application. Although the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that the technical solutions of the present application can still be modified or equivalent replaced without departing from the spirit and scope of the present application, and any modification or equivalent replacement should be covered in the protection scope of the claims of the present application.

Claims

1. A method for storing drug dosage control data based on multimodal diagnosis and treatment data, characterized in that, include: The system acquires basic patient information and medical data, performs information matching in the patient database based on the basic information, and obtains the patient's historical medical data and historical medication data; wherein, the medical data includes diagnostic result data and wearable device data; The diagnostic result data is matched with the health diagnostic result data in the health diagnosis and treatment data to obtain the diagnostic result similarity. The wearable device data is matched with the health wearable device data in the health diagnosis and treatment data to obtain the wearable data similarity. Based on the similarity of the diagnostic results and the similarity of the wearable data, determine whether to adjust the historical medication data to obtain the current medication dosage. If it is determined that adjustment is required, output the determination result that needs to be adjusted. The adjusted medication dosage data is obtained, and the diagnostic result data, diagnostic result similarity, wearable device data, wearable data similarity, historical medication dosage data, and adjusted medication dosage data are stored in the medication adjustment unit of the patient database.

2. The method for storing drug dosage control data based on multimodal diagnostic and treatment data according to claim 1, characterized in that, The process of obtaining basic patient information and treatment data, and matching information in the patient database based on the basic information to obtain historical treatment data and historical medication data includes: Input the patient's basic information into the patient database for querying and matching; If the match is successful, the patient's historical medical records and historical medication data will be obtained. If a match is not found, the patient's basic information will be created in the patient database, and the patient's diagnosis and treatment data and medication dosage data will be entered.

3. The method for storing drug dosage control data based on multimodal diagnostic and therapeutic data according to claim 1, characterized in that, The diagnostic results data include tongue image analysis data, facial diagnostic data, pulse identification data, consultation data, and corneal information identification data; The wearable device data includes heart rate data, blood pressure data, and sleep quality data.

4. The method for storing drug dosage control data based on multimodal diagnostic and treatment data according to claim 3, characterized in that, When matching the diagnostic result data with the health diagnostic result data in the health diagnosis data to obtain the diagnostic result similarity, the process includes: Feature extraction was performed on tongue image analysis data, facial diagnosis data, pulse identification data, consultation data, and corneal information identification data to construct multi-dimensional feature vectors. For tongue image analysis data, extract tongue color, tongue coating color and tongue shape feature parameters, and calculate the tongue image feature similarity between the current tongue image and the healthy tongue image using the cosine similarity algorithm; For facial diagnostic data, a convolutional neural network is used to extract facial skin color, texture and morphological feature points, and a structural similarity index is used to evaluate the facial state similarity between the current facial features and healthy facial features. For pulse identification data, frequency domain features, time domain features and nonlinear dynamic features of pulse waves are extracted, and the similarity of pulse waveform with healthy pulse is calculated by dynamic time warping algorithm. For consultation data, a TCM symptom lexicon and semantic network are constructed, and the semantic similarity between the current symptom description and the health consultation description is calculated through a word vector model. For corneal information identification data, corneal transparency, vascular morphology and nerve distribution feature parameters are extracted, and fuzzy clustering algorithm is used to calculate the similarity of corneal state with healthy cornea; The similarity of diagnostic results is calculated by weighted summation of similarity in tongue features, facial condition, pulse waveform, semantics, and corneal condition.

5. The method for storing drug dosage control data based on multimodal diagnostic and treatment data according to claim 4, characterized in that, When matching wearable device data with health wearable device data in health diagnosis and treatment data to obtain wearable data similarity, the process includes: For heart rate data, a 24-hour dynamic heart rate curve is constructed, and resting heart rate, exercise heart rate variability and heart rate recovery time are extracted. The similarity of the current heart rate curve and the healthy heart rate curve is calculated by Pearson correlation coefficient. For blood pressure data, the systolic blood pressure fluctuation trend, diastolic blood pressure diurnal rhythm and pulse pressure difference were extracted, and time series analysis was used to assess the similarity of blood pressure data with healthy blood pressure patterns. For sleep quality data, sleep latency, sleep efficiency, proportion of each sleep stage and micro-arousal index are extracted based on polysomnography data. Support vector machine classifier is used to calculate the similarity of sleep data with healthy sleep structure. The wearable data similarity is obtained by weighted summation of the heart rate data similarity, blood pressure data similarity, and sleep data similarity.

6. The method for storing drug dosage control data based on multimodal diagnostic and treatment data according to claim 5, characterized in that, When determining whether to adjust historical medication data to obtain the current medication dosage based on the similarity of the diagnostic results and the similarity of the wearable data, the following steps are included: The similarity of the diagnostic results is compared with the similarity of historical diagnostic results to obtain the first comparison result; The similarity of the wearable data is compared with the similarity of historical wearable data to obtain the second comparison result; Based on the first and second comparison results, determine whether to adjust the historical drug dosage data to obtain the current drug dosage.

7. The method for storing drug dosage control data based on multimodal diagnostic and treatment data according to claim 6, characterized in that, When determining whether to adjust the historical drug dosage data to obtain the current drug dosage based on the first comparison result and the second comparison result, the following steps are included: The difference between the similarity of the diagnostic results and the similarity of historical diagnostic results is calculated and denoted as the diagnostic similarity difference; the difference between the similarity of the wearable data and the similarity of historical wearable data is calculated and denoted as the wearable similarity difference. If both the diagnostic similarity difference and the wearability similarity difference are greater than or equal to zero, it is preliminarily determined that the historical medication data will not be adjusted. Otherwise, it is determined that the historical drug dosage data needs to be adjusted to increase the dosage.

8. The method for storing drug dosage control data based on multimodal diagnostic and treatment data according to claim 7, characterized in that, If both the diagnostic similarity difference and the wearability similarity difference are greater than or equal to zero, then the preliminary judgment is that no adjustment will be made to the historical medication data, including: If both the diagnostic similarity difference and the wearability similarity difference are greater than the difference threshold, it is determined that the historical drug dosage data needs to be adjusted to reduce the drug dosage. Otherwise, it is determined that the historical drug dosage data will not be adjusted.

9. The method for storing drug dosage control data based on multimodal diagnostic and therapeutic data according to claim 8, characterized in that, When determining that an adjustment is needed, outputting the determination result includes: The system outputs and displays the results of whether historical drug dosage data needs to be adjusted, as well as the results of increasing or decreasing the drug dosage.

10. A drug dosage control data storage system based on multimodal diagnostic and therapeutic data, used to apply the drug dosage control data storage method based on multimodal diagnostic and therapeutic data as described in any one of claims 1-9, characterized in that, include: The data acquisition module is configured to acquire basic patient information and medical data, and to perform information matching in the patient database based on the basic information to acquire the patient's historical medical data and historical medication data; wherein, the medical data includes diagnostic result data and wearable device data; The matching module is configured to match the diagnostic result data with the health diagnostic result data in the health diagnosis and treatment data to obtain the diagnostic result similarity, and to match the wearable device data with the health wearable device data in the health diagnosis and treatment data to obtain the wearable data similarity. The output display module is configured to determine whether to adjust the historical medication data to obtain the current medication based on the similarity of the diagnostic results and the similarity of the wearable data; if it is determined that adjustment is required, the determination result of the adjustment is output. The storage module is configured to acquire the adjusted medication data and store the diagnostic result data, diagnostic result similarity, wearable device data, wearable data similarity, historical medication data, and adjusted medication data into the medication adjustment unit in the patient database.

Images