Big data-based personalized rehabilitation scheme processing generation method, system and platform and storage medium

Abstract

The invention discloses a personalized rehabilitation scheme processing generation method, system and platform based on big data and a storage medium. The method comprises the following steps: generating and acquiring first data corresponding to a rehabilitation scheme, and creating a vector database corresponding to the rehabilitation scheme based on the first data; wherein the first data is multi-dimensional feature data corresponding to patient big data; and performing weight processing on the first data by adopting an analytic hierarchy process, and performing matching processing to generate a personalized rehabilitation training scheme corresponding to the patient to be rehabilitated in combination with the sparsity and separability of the vector database, so that the personalized rehabilitation scheme can be constructed and generated. And the service is provided based on the personalized rehabilitation scheme in combination with a simple mode of a wearable vest. Namely, a personalized rehabilitation scheme is optimized by collecting, simulating and analyzing a large amount of patient data and based on the information such as medical history, treatment response and living habits, and service is provided in combination with a simple mode of a wearable vest.

Description

Technical Field

[0001] The present invention belongs to the technical field of personalized rehabilitation plan processing, and specifically relates to a method, system, platform and storage medium for generating personalized rehabilitation plans based on big data. Background Art

[0002] Currently, frozen shoulder (also known as "adhesive capsulitis" or "frozen shoulder") is a disease that affects the range of motion of the shoulder joint, usually accompanied by pain and functional impairment. For the treatment and rehabilitation of this disease, traditional methods often adopt standardized treatment paths. However, the conditions, physical conditions, lifestyles, causes, ages, and physical conditions of each patient are different. Therefore, the responses to treatment may also vary, including differences in the intensity of rehabilitation, number of times per day, duration of each rehabilitation session, and rehabilitation cycle. This can lead to situations of under-rehabilitation and over-rehabilitation in the rehabilitation method, resulting in unsatisfactory rehabilitation effects.

[0003] That is to say, in view of the above problems, currently there is a lack of a method for generating personalized rehabilitation plans based on big data, that is, it is impossible to construct personalized rehabilitation plans.

[0004] Therefore, in view of the current lack of a method for generating personalized rehabilitation plans based on big data, that is, the technical problem and defect that it is impossible to construct personalized rehabilitation plans, it is urgent to design and develop a method, system, platform and storage medium for generating personalized rehabilitation plans based on big data. Summary of the Invention

[0005] In order to overcome the deficiencies and difficulties of the above-mentioned prior art, the purpose of the present invention is to provide a method, system, platform and storage medium for generating personalized rehabilitation plans based on big data, so as to construct and generate personalized rehabilitation plans, and provide services in a simple way by combining with a wearable vest based on the personalized rehabilitation plan.

[0006] The first object of the present invention is to provide a method for generating personalized rehabilitation plans based on big data; the second object of the present invention is to provide a system for generating personalized rehabilitation plans based on big data; the third object of the present invention is to provide a platform for generating personalized rehabilitation plans based on big data; the fourth object of the present invention is to provide a computer-readable storage medium.

[0007] The first object of the present invention is achieved as follows: The method includes the following steps:

[0008] Generate and obtain first data corresponding to the rehabilitation plan, and create a vector database corresponding to the rehabilitation plan based on the first data; wherein, the first data is multi-dimensional feature data corresponding to patient big data;

[0009] Using the analytic hierarchy process, weight the first data, and combine the sparsity and separability of the vector database to generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated through matching processing.

[0010] Furthermore, generating and obtaining the first data corresponding to the rehabilitation plan, and creating a vector database corresponding to the rehabilitation plan based on the first data further includes:

[0011] Establish a first model corresponding to the first data; wherein, the first model is a model of personal multi-feature content; the multi-dimensional feature data includes condition data, physical condition data, lifestyle data, cause of disease data, and age and physical condition data;

[0012] Based on the first model, construct a personalized vector database, a generalized knowledge graph vector database, and a famous doctor's personalized case vector database corresponding to the rehabilitation plan respectively.

[0013] Furthermore, using the analytic hierarchy process, weighting the first data, and combining the sparsity and separability of the vector database to generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated through matching processing further includes:

[0014] Construct a second model corresponding to the first data, wherein the second model is an entity recognition model;

[0015] Generate and obtain a first matrix corresponding to the second model, and optimize the sequence order corresponding to the training part of speech based on the first matrix; the first matrix is a state transition matrix.

[0016] Furthermore, using the analytic hierarchy process, weighting the first data, and combining the sparsity and separability of the vector database to generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated through matching processing further includes:

[0017] Generate and obtain second data corresponding to the rehabilitated patient, and generate corresponding third data based on the second data; wherein, the second data is personalized information data of the rehabilitated patient's main complaint; the third data is similarity data between at least two rehabilitated patients;

[0018] Generate and obtain fourth data corresponding to the rehabilitated patient, and fuse and process the third data based on the fourth data; wherein, the fourth data is symptom weight data and proportional coefficient data.

[0019] Further, the method of using the analytic hierarchy process to process the weights of the first data and combining the sparsity and separability of the vector database to match and generate a personalized rehabilitation training plan for the patient to be rehabilitated further includes:

[0020] Calculating and generating third data corresponding to the rehabilitated patient; wherein, the calculation formula is:

[0021] Similarity C = α 1 * α 2 * λ T * S(12);

[0022] Further, the method of using the analytic hierarchy process to process the weights of the first data and combining the sparsity and separability of the vector database to match and generate a personalized rehabilitation training plan for the patient to be rehabilitated further includes:

[0023] Generating and obtaining fifth data corresponding to the knowledge graph, and calculating and generating corresponding sixth data according to the fifth data; wherein, the fifth data is word data; the sixth data is the similarity data between word vectors; the calculation formula is as follows:

[0024]

[0025] In the formula, w i and w j are two word vectors, n is the dimension of the word vector, w ik and w jk respectively represent the values of the k-th elements in w i and w j ;

[0026] Based on the sixth data, determining and generating seventh data and eighth data corresponding to the sixth data; wherein, the seventh data is the correlation data between word vectors; the eighth data is the prediction probability data of word vectors; the calculation formula of the prediction probability data is as follows:

[0027]

[0028] In the formula, w i is the word vector of the central word, w j is the word vector of the predicted word, and m is the total number of different words in the corpus.

[0029] The second object of the present invention is achieved as follows: The system is used to implement the above-mentioned method for processing and generating a personalized rehabilitation plan based on big data; the system includes:

[0030] A data generation and creation unit for generating and obtaining first data corresponding to a rehabilitation plan, and creating a vector database corresponding to the rehabilitation plan based on the first data; wherein the first data is multi-dimensional feature data corresponding to patient big data.

[0031] A data processing and generation unit for processing the first data with the analytic hierarchy process, and matching and processing to generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated in combination with the sparsity and separability of the vector database.

[0032] Further, the data generation and creation unit further includes:

[0033] A first construction module for establishing a first model corresponding to the first data; wherein the first model is a model of personal multi-feature content; the multi-dimensional feature data includes condition data, physical condition data, lifestyle data, etiology data, and age and physical condition data.

[0034] A second construction module for respectively constructing a personalized vector database, a generalized knowledge graph vector database, and a famous doctor's personalized case vector database corresponding to the rehabilitation plan based on the first model.

[0035] And / or, the data processing and generation unit further includes:

[0036] A third construction module for constructing a second model corresponding to the first data, wherein the second model is an entity recognition model.

[0037] A first processing module for generating and obtaining a first matrix corresponding to the second model, and optimizing the sequence order corresponding to the training part of speech based on the first matrix; the first matrix is a state transition matrix.

[0038] A first generation module for generating and obtaining second data corresponding to the rehabilitated patient, and generating corresponding third data based on the second data; wherein the second data is personalized information data of the rehabilitated patient's main complaint; the third data is similarity data between at least two rehabilitated patients.

[0039] A second generation module for generating and obtaining fourth data corresponding to the rehabilitated patient, and fusing the third data based on the fourth data; wherein the fourth data is symptom weight data and proportional coefficient data.

[0040] A first calculation module for calculating and generating third data corresponding to the rehabilitated patient; wherein the calculation formula is:

[0041] Similarity C =α1 *α 2 *λ T *S(12);

[0042] A second calculation module, configured to generate and obtain fifth data corresponding to the knowledge graph, and calculate and generate corresponding sixth data according to the fifth data; wherein, the fifth data is word data; the sixth data is similarity data between word vectors; the calculation formula is as follows:

[0043]

[0044] In the formula, w i and w j are two word vectors, n is the dimension of the word vector, w ik and w jk respectively represent the values of the k-th elements in w i and w j ;

[0045] A third calculation module, configured to determine and generate seventh data and eighth data corresponding to the sixth data based on the sixth data; wherein, the seventh data is correlation data between word vectors; the eighth data is prediction probability data of word vectors; the calculation formula of the prediction probability data is as follows:

[0046]

[0047] In the formula, w i is the word vector of the central word, w j is the word vector of the predicted word, and m is the total number of different words in the corpus.

[0048] The third object of the present invention is achieved as follows: It includes a processor, a memory, and a control program for a personalized rehabilitation plan processing and generating platform based on big data; wherein, in the processor, the control program for the personalized rehabilitation plan processing and generating platform based on big data is executed, the control program for the personalized rehabilitation plan processing and generating platform based on big data is stored in the memory, and the control program for the personalized rehabilitation plan processing and generating platform based on big data implements the personalized rehabilitation plan processing and generating method based on big data.

[0049] The fourth object of the present invention is achieved as follows: The computer-readable storage medium stores a control program for a personalized rehabilitation plan processing and generating platform based on big data, and the control program for the personalized rehabilitation plan processing and generating platform based on big data implements the personalized rehabilitation plan processing and generating method based on big data.

[0050] The present invention generates and obtains first data corresponding to a rehabilitation plan through a method, and creates a vector database corresponding to the rehabilitation plan based on the first data; wherein the first data is multi-dimensional feature data corresponding to patient big data; the first data is processed with weights by using the analytic hierarchy process, and combined with the sparsity and separability of the vector database, a personalized rehabilitation training plan corresponding to a patient to be rehabilitated is generated through matching processing, a personalized rehabilitation plan can be constructed, and services can be provided based on the personalized rehabilitation plan in combination with a simple method of a wearable vest.

[0051] That is to say, the solution of the present invention optimizes a personalized rehabilitation plan by collecting, simulating and analyzing data of a large number of patients, and based on the above information such as medical history, treatment response, living habits, etc., and provides services in combination with a simple method of a wearable vest. Brief Description of the Drawings

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for description in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0053] Figure 1 It is a schematic flowchart of a method for processing and generating a personalized rehabilitation plan based on big data of the present invention;

[0054] Figure 2 It is a schematic diagram of the system framework of a method for processing and generating a personalized rehabilitation plan based on big data of the present invention;

[0055] Figure 3 It is a schematic flowchart of knowledge graph construction of a method for processing and generating a personalized rehabilitation plan based on big data of the present invention;

[0056] Figure 4 It is a schematic flowchart of entity extraction of a method for processing and generating a personalized rehabilitation plan based on big data of the present invention;

[0057] Figure 5 It is a schematic diagram of the system architecture of a system for processing and generating a personalized rehabilitation plan based on big data of the present invention;

[0058] Figure 6 It is a schematic diagram of the platform architecture of a platform for processing and generating a personalized rehabilitation plan based on big data of the present invention;

[0059] Figure 7 It is a schematic diagram of the architecture of a computer-readable storage medium in an embodiment of the present invention;

[0060] In the figure: word embedding: word embedding; LE-BERT: an algorithm of Bert; CRF: hidden Markov method; B(I)-disease: the disease in the database (input); B(I)-Symptom: the feature in the database (input). Detailed implementation manners

[0061] To better understand the purpose, technical solution and advantages of the present invention, the present invention will be further described below in conjunction with the accompanying drawings and specific implementation manners. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification.

[0062] The present invention can also be implemented or applied through other different specific examples. The details in this specification can also be modified and changed based on different viewpoints and applications without departing from the spirit of the present invention.

[0063] It should be noted that if there are directional indications (such as up, down, left, right, front, back...) involved in the embodiments of the present invention, the directional indications are only used to explain the relative position relationship and movement conditions between components in a specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indications will also change accordingly.

[0064] In addition, if there are descriptions such as "first", "second", etc. involved in the embodiments of the present invention, the descriptions of "first", "second", etc. are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of the features. Secondly, the technical solutions between various embodiments can be combined with each other, but it must be based on the fact that those skilled in the art can implement it. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the protection scope required by the present invention.

[0065] Preferably, a method for processing and generating a personalized rehabilitation plan based on big data according to the present invention is applied to one or more terminals or servers. The terminal is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to a microprocessor, an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP), an embedded device, etc.

[0066] The terminal can be a computing device such as a desktop computer, a notebook, a palm computer, or a cloud server. The terminal can interact with the customer through a keyboard, a mouse, a remote control, a touchpad, or a voice control device, etc.

[0067] The present invention aims to implement a method, a system, a platform, and a storage medium for processing and generating a personalized rehabilitation plan based on big data.

[0068] As Figure 1 shown, it is a flowchart of the method for processing and generating a personalized rehabilitation plan based on big data provided by an embodiment of the present invention.

[0069] In this embodiment, the method for processing and generating a personalized rehabilitation plan based on big data can be applied to a terminal with a display function or a fixed terminal. The terminal is not limited to a personal computer, a smart phone, a tablet computer, a desktop computer or an all-in-one computer equipped with a camera, etc.

[0070] The method for processing and generating a personalized rehabilitation plan based on big data can also be applied to a hardware environment composed of a terminal and a server connected to the terminal through a network. The network includes but is not limited to: a wide area network, a metropolitan area network, or a local area network. The method for processing and generating a personalized rehabilitation plan based on big data in the embodiment of the present invention can be executed by the server, can be executed by the terminal, or can be jointly executed by the server and the terminal.

[0071] For example, for a terminal that needs to process and generate a personalized rehabilitation plan based on big data, the function of processing and generating a personalized rehabilitation plan provided by the method of the present invention can be directly integrated on the terminal, or a client for implementing the method of the present invention can be installed. Again, the method provided by the present invention can also run on a device such as a server in the form of a Software Development Kit (SDK), and provide an interface for the function of processing and generating a personalized rehabilitation plan in the form of an SDK. The terminal or other devices can implement the function of processing and generating a personalized rehabilitation plan through the provided interface. The following further elaborates on the present invention with reference to the accompanying drawings.

[0072] As Figures 1 - 4 shown, the present invention provides a method for processing and generating a personalized rehabilitation plan based on big data. The method includes the following steps:

[0073] S1. Generate and obtain first data corresponding to the rehabilitation plan, and create a vector database corresponding to the rehabilitation plan based on the first data; wherein, the first data is multi-dimensional feature data corresponding to the patient's big data;

[0074] S2. Use the analytic hierarchy process to weight-process the first data, and combine the sparsity and separability of the vector database to match and process to generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated.

[0075] The generating and obtaining of the first data corresponding to the rehabilitation plan, and based on the first data, creating a vector database corresponding to the rehabilitation plan, further includes:

[0076] S11. Establish a first model corresponding to the first data; wherein, the first model is a model of personal multi-feature content; the multi-dimensional feature data includes condition data, physical condition data, lifestyle data, etiology data, and age and physical condition data;

[0077] S12. Based on the first model, respectively construct a personalized vector database, a generalized knowledge graph vector database, and a famous doctor's personalized case vector database corresponding to the rehabilitation plan.

[0078] The using the analytic hierarchy process to weight-process the first data, and combining the sparsity and separability of the vector database to match and process to generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated, further includes:

[0079] S21. Construct a second model corresponding to the first data, wherein the second model is an entity recognition model;

[0080] S22. Generate and obtain a first matrix corresponding to the second model, and based on the first matrix, optimize and process the sequence order corresponding to the training part of speech; the first matrix is a state transition matrix.

[0081] The using the analytic hierarchy process to weight-process the first data, and combining the sparsity and separability of the vector database to match and process to generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated, further includes:

[0082] S23. Generate and obtain second data corresponding to the rehabilitated patient, and based on the second data, generate corresponding third data; wherein, the second data is personalized information data of the rehabilitated patient's main complaint; the third data is similarity data between at least two rehabilitated patients;

[0083] S24. Generate and obtain fourth data corresponding to the rehabilitated patient, and based on the fourth data, fuse and process the third data; wherein, the fourth data is symptom weight data and proportional coefficient data.

[0084] Using the analytic hierarchy process, weighting the first data, and combining the sparsity and separability of the vector database, matching and processing to generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated, further including:

[0085] S25. Calculate and generate the third data corresponding to the rehabilitated patient; where the calculation formula is:

[0086] Similarity C =α 1 *α 2 *λ T *S(12)

[0087] In the formula, α 1 , α 2 : The rehabilitee C 1 The main complaints and the main complaints C of this disease collected by the database 2 Respectively have p and N symptoms. Through the semantic analysis of the large model, it is determined that there are m symptoms with non-zero similarity between the two main complaints. Then the proportionality coefficient is:

[0088]

[0089] λ is a vector, λ = [λ 1 , λ 2 , …, λ m

[0090] Generally, the symptoms with higher occurrence frequencies are less able to reveal the individual differences between rehabilitees. Because the higher the occurrence frequency of a certain symptom, it indicates that to a large extent it is a commonality among most patients. If too high a weight is assigned to this symptom, then the differences between the main complaints of different rehabilitees cannot be distinguished. On the contrary, those symptoms with low occurrence frequencies are more helpful in distinguishing the differences between the main complaints of different rehabilitees. Because according to the idea of small probability events, once an event with a low occurrence probability occurs, the probability of this event revealing the essence of things is greater. The occurrence frequency of each symptom is:

[0091] f 1 , f 2 , …, f N

[0092] Therefore, the symptom weight coefficient designed in this paper is as shown in the formula:

[0093]

[0094] Taking the first m λ j , it constitutes λ T

[0095] P: is a vector {p 1 , p 2 , …, p​m}

[0096] (1). For symptom k (between 1 and N), if patient C 1 and patient C 2 both chief complaints have symptom Sym 1 = 1, according to the probability pi output by the model for similar chief complaint symptoms, where the first m are non-zero.

[0097] Using the analytic hierarchy process, weighting the first data, and combining the sparsity and separability of the vector database, matching and processing to generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated, further including:

[0098] S26. Generate and obtain the fifth data corresponding to the knowledge graph, and calculate and generate the corresponding sixth data according to the fifth data; wherein, the fifth data is word data; the sixth data is the similarity data between word vectors; the calculation formula is as follows:

[0099]

[0100] In the formula, w i and w j are two word vectors, n is the dimension of the word vector, w ik and w jk respectively represent the value of the k-th element in w i and w j ;

[0101] S27. Based on the sixth data, determine and generate the seventh data and the eighth data corresponding to the sixth data; wherein, the seventh data is the correlation data between word vectors; the eighth data is the prediction probability data of word vectors; the calculation formula of the prediction probability data is as follows:

[0102]

[0103] In the formula, w i is the word vector of the central word, w j is the word vector of the predicted word, and m is the total number of different words in the corpus.

[0104] Specifically, in the embodiment of the present invention, a personalized rehabilitation system based on big data is provided, including an agent module (A), a knowledge base (B) constructed by the personalized information, general information of the rehabilitated person and the clinical information of famous doctors, a retrieval and COT generation module (C), a knowledge graph reliability verification module, and a model personalized guidance module (D).

[0105] The solution of the present invention focuses on the big data analysis module. After this module obtains the personalized information of the rehabilitator, through the following big data analysis methods, relatively optimized rehabilitation training plans and hardware rehabilitation intensities are given. Specifically:

[0106] Define a vector database for multi-dimensional features (disease condition, physical condition, lifestyle, cause of disease, age and physical condition) and three types of knowledge bases (personalized, generalized knowledge graph, and personalized cases of famous doctors).

[0107] Based on the modeling method of an individual's multi-feature content (disease condition, physical condition, lifestyle, cause of disease, age and physical condition), the analytic hierarchy process is used to determine the weight calculation method of multi-dimensional features, and issues such as the sparsity, separability, and recommendation efficiency of the vector database are verified.

[0108] Use the GrapRAG technology to match the data of multiple dimensions to the preferences of the rehabilitator from big data. Combine the recommendation methods based on user personal information and collaborative filtering, and adopt a hybrid rehabilitation training plan and rehabilitation intensity recommendation based on multi-dimensional features, which improves the adaptability, scalability and other characteristics of the personalized recommendation model.

[0109] That is to say, the integration of personalized knowledge, general knowledge and the experience of famous doctors forms a knowledge information composition with a wide knowledge coverage, reflecting personalization and absorbing the experience of famous doctors, and is verified for reliability by the knowledge graph, providing reliable personalized guidance suggestions.

[0110] Knowledge graph construction technology

[0111] (1) Knowledge graph query dictionary

[0112] Table 1 Example of entity query dictionary

[0113] Entity Query Dictionary Example Disease {Scapulohumeral periarthritis, Muscle strain} Symptom {Pain, Cramp, Soreness} Treatment {Massage, Hot compress, Moxibustion} Examination {Individual body movements, Doctor's manual examination}

[0114] (2) Entity recognition

[0115] Based on the LEBERT and CRF (Markov) entity recognition models. In LEBERT, a large-scale pre-trained Chinese word vector published by Tencent AILab is used to inject into the bottom layer of BERT, which can help BERT better understand the semantic information and context relationship in the Chinese language. The dimension of this pre-trained word vector is 200, and the total number is 8,824,330. The role of CRF is to solve the sequence labeling problem. In the form of directly connecting the pre-trained language model to the classifier, without considering the part-of-speech order, invalid outputs will be obtained, such as "B-disease I-symptom". Therefore, the solution of the present invention adopts the model of connecting LEBERT and CRF, and uses the state transition matrix of CRF to optimize the correct sequence order.

[0116] The optimized loss function integrates rebalancing weighted sum and Negative Tolerant Regularization (NTR). The Distribution-balanced Loss (DB) first reduces the redundant information of label co-occurrence, which is crucial in multi-label scenarios, and then explicitly assigns lower weights to the "easy-to-classify" negative instances.

[30] 。

[0117] First, to rebalance the weights, in the single-label scenario, the resampling probability of an instance can be weighted; while in the multi-label scenario, if the same strategy is followed, an instance with multiple labels can be oversampled with a probability of Therefore, the rebalanced weights can be normalized using Through the smoothing function, maps r to [α, α + 1], and the Rebalanced-FL (R-FL) loss function is defined as: DB

[0118]

[0119] Then, NTR treats the positive and negative instances of the same label differently. The scale factor λ and the inherent class-specific bias v i are introduced to lower the threshold of the tail class and avoid over-suppression.

[0120]

[0121] Among them, is the positive example, is the negative example. v i can be estimated by minimizing the loss function with the scale factor κ and the class prior at the beginning of training, so:

[0122]

[0123] Finally, DB integrates the rebalancing weighted sum and NTR as:

[0124]

[0125] Personalized knowledge retrieval: After determining the personalized information of the recovered person's main complaints, retrieve the acquaintance degree calculation and processing:

[0126] (1) For the symptom Sym 1 , if two recovered person's main complaints both have the symptom Sym 1 , then Sym 1 ​The similarity is 1; secondly, if the semantic network graph is used to determine that there are similar symptoms of Sym in the main complaints of the recovered patients 1 assuming the number of similar symptoms (including Sym 1 ) is n, then the similarity between Sym 1 and its similar symptoms is defined as 1 / (n - 1) in the solution of the present invention; if there are neither the same symptoms nor similar symptoms in the main complaints of the recovered patients, the similarity between the two main complaints of the recovered patients in Sym 1 is zero.

[0127] (2) Calculate the symptom weight and the proportionality coefficient: The symptom weight and the proportionality coefficient are two indispensable coefficients when fusing the similarity between each individual symptom into the overall similarity of the main complaints of the recovered patients. The weight of a symptom is the magnitude of the contribution of the symptom to the overall similarity, and the value ranges from 0 to 1. Assuming that there are N symptoms in total that appear in the main complaint database of the recovered patients:

[0128] Sym 1 ,Sym 2 ,…,Sym N (5)

[0129] The frequency of occurrence of each symptom is:

[0130] f 1 ,f 2 ,…,f N (6)

[0131] Generally speaking, the higher the frequency of occurrence of a symptom, the less able it is to reveal the individual differences between the recovered patients. Because the higher the frequency of occurrence of a certain symptom, it means that to a large extent it is the commonality of most patients. If too high a weight is assigned to this symptom, then the differences between the main complaints of different recovered patients cannot be distinguished. On the contrary, those symptoms with a low frequency of occurrence are more helpful in distinguishing the differences between the main complaints of different recovered patients. Because according to the idea of small probability events, when an event with a low probability of occurrence occurs, the probability of this event revealing the essence of things is greater. Therefore, the symptom weight coefficient designed in the solution of the present invention is as shown in formula (7):

[0132]

[0133] When measuring the similarity between the main complaints of two recovered patients, the proportionality coefficient refers to the proportion of the number of symptoms with non-zero similarity in the total number of symptoms contained in the main complaints of the two recovered patients respectively. For example, if the main complaint C 1 and the main complaint C 2 have p and q symptoms respectively, and there are m symptoms with non-zero similarity between them, then the proportionality coefficient is:

[0134]

[0135] (3) Similarity of the main complaints of recovered patients: After calculating the similarity, weight, and proportional coefficient of each symptom between the main complaints of two recovered patients, the following parameters represented in vector form can be obtained:

[0136] S = [Sym 1 , Sym 2 , …, Sym m T (9)

[0137] λ = [λ 1 , λ 2 , …, λ m T (10)

[0138] α 1 , α 2 (11)

[0139] The above three variables respectively represent: the similarity of all symptoms; the weight of symptoms; the proportional coefficient of symptoms with non - zero similarity in the main complaints of two recovered patients. Combining formula (9), formula (10), and formula (11), the similarity expression of the main complaints of recovered patients can be obtained as:

[0140] Similarity C = α 1 * α 2 * λ T * S (12)

[0141] Reliability verification technology for the output results of the large model: Given a text of the output results of the large model, it is divided into several text windows, and each window contains a central word and several surrounding words.

[0142] (1) Predict the central word based on the context words in the window, thereby training the vector representation of each word. Given a word, predicting the word in the knowledge graph, the similarity between word vectors is represented by the following formula:

[0143]

[0144] Among them, w i and w j are two word vectors, n is the dimension of the word vector, w ik and w jk respectively represent the values of the k - th element in w i and w j . By calculating the inner product between two word vectors, their similarity can be obtained.

[0145] ​​(2) The matching between these words and the information fusion of the knowledge graph. Finally, it is judged that the greater the similarity between the word vectors, the greater the correlation between these two words. Improve the above results, which are used to represent the prediction probability of word vectors. The formula is as follows:

[0146]

[0147] Among them, w i is the word vector of the central word, w j is the word vector of the predicted word, and m is the total number of different words in the corpus. By calculating the inner product between each word vector and the central word vector, the prediction probability of each word given the central word can be obtained.

[0148] To achieve the above object, the present invention also provides a personalized rehabilitation plan processing and generating system based on big data, and the system is used to implement the above-mentioned personalized rehabilitation plan processing and generating method based on big data; as Figure 5 shown, the system specifically includes:

[0149] A data generation and creation unit, which is used to generate and obtain the first data corresponding to the rehabilitation plan, and based on the first data, create a vector database corresponding to the rehabilitation plan; wherein, the first data is multi-dimensional feature data corresponding to the patient's big data;

[0150] A data processing and generation unit, which is used to weight-process the first data by using the analytic hierarchy process, and combine the sparsity and separability of the vector database to match and process and generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated.

[0151] The data generation and creation unit further includes:

[0152] A first construction module, which is used to establish a first model corresponding to the first data; wherein, the first model is a model of personal multi-feature content; the multi-dimensional feature data includes disease condition data, physical condition data, lifestyle data, etiology data, and age and physical condition data;

[0153] A second construction module, which is used to respectively construct a personalized vector database, a generalized knowledge graph vector database, and a famous doctor's personalized case vector database corresponding to the rehabilitation plan based on the first model;

[0154] And / or, the data processing and generation unit further includes:

[0155] A third construction module, which is used to construct a second model corresponding to the first data, wherein the second model is an entity recognition model;

[0156] The first processing module is used to generate and obtain a first matrix corresponding to the second model, and based on the first matrix, optimize and process the sequence order corresponding to the training part-of-speech; the first matrix is a state transition matrix;

[0157] The first generation module is used to generate and obtain second data corresponding to the rehabilitation patients, and based on the second data, generate corresponding third data; wherein, the second data is the personalized information data of the main complaints of the rehabilitators; the third data is the similarity data between at least two rehabilitators;

[0158] The second generation module is used to generate and obtain fourth data corresponding to the rehabilitation patients, and based on the fourth data, fuse and process the third data; wherein, the fourth data is the symptom weight data and the proportional coefficient data;

[0159] The first calculation module is used to calculate and generate third data corresponding to the rehabilitation patients; wherein, the calculation formula is:

[0160] Similarity C =α 1 *α 2 *λ T *S (12)

[0161] In the formula, α 1 ,α 2 : The main complaints of rehabilitator C 1 and the main complaints C of this disease collected in the database 2 have p and N symptoms respectively. Through the semantic analysis of the large model, it is determined that there are m symptoms with non-zero similarity between the two main complaints. Then the proportional coefficient is:

[0162]

[0163] λ is a vector, λ = [λ 1 ,λ 2 ,…,λ m

[0164] Generally, the symptoms with higher occurrence frequencies are less able to reveal the individual differences between rehabilitators. Because the higher the occurrence frequency of a certain symptom, it indicates that to a large extent it is a commonality of most patients. If too high a weight is assigned to this symptom, then the differences between the main complaints of different rehabilitators cannot be distinguished. On the contrary, those symptoms with lower occurrence frequencies are more helpful in distinguishing the differences between the main complaints of different rehabilitators. Because according to the idea of small probability events, when an event with a low occurrence probability occurs, the probability of this event revealing the essence of things is greater. The occurrence frequency of each symptom is:

[0165] f 1 ,f 2 ,…,f​N

[0166] Therefore, the symptom weight coefficient designed in this paper is as follows:

[0167]

[0168] Take the first m λ j , and it constitutes λ T

[0169] P: is the vector {p 1 , p 2 , …, p m}

[0170] (1). For symptom k (between 1 and N), if the two chief complaints of patient C 1 and patient C 2 both have symptom Sym 1 = 1, according to the probability pi output by the model for the similarity of chief complaint symptoms, where the first m are non-zero;

[0171] The second calculation module is used to generate and obtain the fifth data corresponding to the knowledge graph, and calculate and generate the corresponding sixth data according to the fifth data; wherein, the fifth data is word data; the sixth data is the similarity data between word vectors; the calculation formula is as follows:

[0172]

[0173] In the formula, w i and w j are two word vectors, n is the dimension of the word vector, w ik and w jk respectively represent the values of the kth element in w i and w j ;

[0174] The third calculation module is used to determine and generate the seventh data and the eighth data corresponding to the sixth data based on the sixth data; wherein, the seventh data is the correlation data between word vectors; the eighth data is the prediction probability data of word vectors; the calculation formula of the prediction probability data is as follows:

[0175]

[0176] In the formula, w i is the word vector of the central word, w j is the word vector of the predicted word, and m is the total number of different words in the corpus.

[0177] In the embodiment of the system solution of the present invention, the method steps involved in the processing and generation of a personalized rehabilitation solution based on big data have been elaborated above. That is to say, the functional modules in the system are used to implement the steps or sub-steps in the above method embodiment, which will not be elaborated here.

[0178] To achieve the above object, the present invention further provides a platform for processing and generating a personalized rehabilitation solution based on big data, as Figure 6 shown, including a processor, a memory, and a control program for the platform for processing and generating a personalized rehabilitation solution based on big data; wherein, the processor executes the control program for the platform for processing and generating a personalized rehabilitation solution based on big data, and the control program for the platform for processing and generating a personalized rehabilitation solution based on big data is stored in the memory, and the control program for the platform for processing and generating a personalized rehabilitation solution based on big data realizes the method steps of the processing and generation of the personalized rehabilitation solution based on big data. For example:

[0179] S1. Generate and obtain first data corresponding to the rehabilitation solution, and based on the first data, create a vector database corresponding to the rehabilitation solution; wherein, the first data is multi-dimensional feature data corresponding to patient big data;

[0180] S2. Use the analytic hierarchy process to weight-process the first data, and combine the sparsity and separability of the vector database to match and process and generate a personalized rehabilitation training solution corresponding to the patient to be rehabilitated.

[0181] The specific details of the steps have been elaborated above and will not be elaborated here.

[0182] In the embodiment of the present invention, the built-in processor of the platform for processing and generating a personalized rehabilitation solution based on big data can be composed of integrated circuits. For example, it can be composed of a single packaged integrated circuit, or can be composed of multiple integrated circuits with the same or different functions packaged, including a combination of one or more central processing units (CPU), microprocessors, digital processing chips, graphics processors, and various control chips. The processor uses various interfaces and lines to connect to each component, and by running or executing the programs or units stored in the memory, and calling the data stored in the memory, to execute various functions of the processing and generation of the personalized rehabilitation solution based on big data and process data;

[0183] The memory is used to store program codes and various data, is installed in the platform for processing and generating a personalized rehabilitation solution based on big data, and realizes the high-speed and automatic access of programs or data during operation.

[0184] The memory includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electrically-erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc memories, a magnetic disc memory, a magnetic tape memory, or any other computer-readable medium that can be used to carry or store data.

[0185] To achieve the above object, the present invention further provides a computer-readable storage medium, such as Figure 7 shown, the computer-readable storage medium stores a control program for a personalized rehabilitation plan processing and generation platform based on big data. The control program for the personalized rehabilitation plan processing and generation platform based on big data implements the steps of the personalized rehabilitation plan processing and generation method based on big data, for example:

[0186] S1. Generate and obtain first data corresponding to the rehabilitation plan, and create a vector database corresponding to the rehabilitation plan based on the first data; wherein, the first data is multi-dimensional feature data corresponding to patient big data;

[0187] S2. Use the analytic hierarchy process to weight-process the first data, and match and process to generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated in combination with the sparsity and separability of the vector database.

[0188] The specific details of the steps have been described above and will not be repeated here.

[0189] In the description of the embodiments of the present invention, it should be noted that any process or method description shown in the flowchart or described in other ways herein can be understood as representing a module, segment, or part of code including one or more executable instructions for implementing a specific logical function or process. The scope of the preferred embodiments of the present invention includes additional implementations, where the functions can be executed in a substantially simultaneous manner or in the reverse order according to the involved functions, rather than in the order shown or discussed, which should be understood by those skilled in the technical field to which the embodiments of the present invention belong.

[0190] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as an ordered list of executable instructions for implementing a logical function, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processing module, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in combination with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in combination with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection portion with one or more wirings (electronic device), a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM).

[0191] In addition, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or other appropriate processing as necessary, and then stored in a computer memory.

[0192] In the embodiments of the present invention, to achieve the above object, the present invention further provides a chip system, which includes at least one processor. When the program instructions are executed in the at least one processor, the chip system is caused to execute the steps of the method for generating a personalized rehabilitation plan based on big data, for example:

[0193] S1. Generate and obtain first data corresponding to the rehabilitation plan, and create a vector database corresponding to the rehabilitation plan based on the first data; wherein the first data is multi-dimensional feature data corresponding to the patient's big data.

[0194] S2. Use the analytic hierarchy process to weight-process the first data, and combine the sparsity and separability of the vector database to perform matching processing to generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated.

[0195] The specific details of the steps have been described above and will not be elaborated here.

[0196] Those of ordinary skill in the art can realize that, combining the units and algorithm steps of each example described in the embodiments disclosed in the solution of the present invention, can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraint conditions of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application. Those skilled in the art can clearly understand that, for the convenience and conciseness of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated here.

[0197] The present invention generates and obtains first data corresponding to a rehabilitation plan through a method, and based on the first data, creates a vector database corresponding to the rehabilitation plan; wherein, the first data is multi-dimensional feature data corresponding to patient big data; use the analytic hierarchy process to weight-process the first data, and combine the sparsity and separability of the vector database to perform matching processing to generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated, can construct and generate a personalized rehabilitation plan, and provide services based on the personalized rehabilitation plan in combination with a simple method of a wearable vest.

[0198] That is to say, the solution of the present invention optimizes a personalized rehabilitation plan by collecting, simulating and analyzing data of a large number of patients, and based on the above information such as medical history, treatment response, living habits, etc., and provides services in combination with a simple method of a wearable vest.

[0199] The above-described embodiments only represent several implementation manners of the present invention, and their descriptions are relatively specific and detailed, but should not be construed as limiting the scope of the patent of the present invention. It should be noted that, for those of ordinary skill in the art, without departing from the concept of the present invention, several deformations and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention should be subject to the appended claims.

Claims

1. A method for generating and processing personalized rehabilitation programs based on big data, characterized in that: The method comprises the steps of: Generate and obtain first data corresponding to the rehabilitation program, and create a vector database corresponding to the rehabilitation program based on the first data; wherein the first data is multi-dimensional feature data corresponding to the patient's big data; The first data is processed by weight using the hierarchical analysis method, and the sparsity and separability of the vector database are combined to generate a personalized rehabilitation training program corresponding to the patient to be rehabilitated through matching processing.

2. A method for processing and generating a personalized rehabilitation program based on big data according to claim 1, characterized in that: The step of generating and acquiring first data corresponding to the rehabilitation program, and creating a vector database corresponding to the rehabilitation program based on the first data, further includes: Establishing a first model corresponding to the first data; wherein the first model is a model of multi-feature content of an individual; the multi-dimensional feature data includes disease data, physical condition data, lifestyle data, etiology data, and age and physical condition data; Based on the first model, a personalized vector database, a generalized knowledge graph vector database and a personalized case vector database of famous doctors corresponding to the rehabilitation plan are constructed respectively.

3. The method for generating a personalized rehabilitation plan based on big data according to claim 1, characterized in that: The method of using the hierarchical analysis method to weight the first data and combining the sparsity and separability of the vector database to match and process to generate a personalized rehabilitation training program corresponding to the patient to be rehabilitated also includes: Constructing a second model corresponding to the first data, wherein the second model is an entity recognition model; Generate and obtain a first matrix corresponding to the second model, and based on the first matrix, optimize the sequence order corresponding to the training part of speech; the first matrix is ​​a state transfer matrix.

4. A method for generating a personalized rehabilitation plan based on big data according to claim 1 or 3, characterized in that: The method of using the hierarchical analysis method to weight the first data and combining the sparsity and separability of the vector database to match and process to generate a personalized rehabilitation training program corresponding to the patient to be rehabilitated also includes: Generate and obtain second data corresponding to the recovered patient, and generate corresponding third data based on the second data; wherein the second data is the recovered patient's main complaint personalized information data; the third data is the similarity data between at least two recovered patients; Generate and obtain fourth data corresponding to the rehabilitation patient, and based on the fourth data, fuse and process the third data; wherein the fourth data is symptom weight data and proportional coefficient data.

5. A method for processing and generating a personalized rehabilitation program based on big data according to claim 4, characterized in that: The method of using the hierarchical analysis method to weight the first data and combining the sparsity and separability of the vector database to match and process to generate a personalized rehabilitation training program corresponding to the patient to be rehabilitated also includes: The third data corresponding to the recovered patient is calculated and generated; wherein the calculation formula is: Similarity C =α1*α2*P*λ T (12)。 6. A method for generating a personalized rehabilitation program based on big data according to claim 1 or 3, characterized in that: The method of using the hierarchical analysis method to weight the first data and combining the sparsity and separability of the vector database to match and process to generate a personalized rehabilitation training program corresponding to the patient to be rehabilitated also includes: Generate and obtain fifth data corresponding to the knowledge graph, and calculate and generate corresponding sixth data based on the fifth data; wherein the fifth data is word data; the sixth data is similarity data between word vectors; the calculation formula is as follows: In the formula, w i and w j are two word vectors, n is the dimension of the word vector, w ik and w jk Respectively represent w i and w j The value of the kth element in ; Based on the sixth data, determine and generate seventh data and eighth data corresponding to the sixth data; wherein the seventh data is the correlation data between word vectors; the eighth data is the prediction probability data of the word vectors; the calculation formula of the prediction probability data is as follows: In the formula, w i is the word vector of the central word, w j is the word vector of the predicted word, and m is the total number of different words in the corpus.

7. A system for processing and generating personalized rehabilitation plans based on big data, characterized in that: The system is used to implement a method for generating a personalized rehabilitation program based on big data as described in any one of claims 1 to 6; the system comprises: A data generation and creation unit, used to generate and obtain first data corresponding to the rehabilitation program, and create a vector database corresponding to the rehabilitation program based on the first data; wherein the first data is multi-dimensional feature data corresponding to the patient's big data; The data processing and generating unit is used to adopt the hierarchical analysis method to weight the first data, and combine the sparsity and separability of the vector database to match and process to generate a personalized rehabilitation training plan corresponding to the patient to be rehabilitated.

8. The system for generating and processing individualized rehabilitation plans based on big data according to claim 7, characterized in that: The data generation and creation unit further includes: A first building module is used to establish a first model corresponding to the first data; wherein the first model is a model of multi-feature content of an individual; the multi-dimensional feature data includes disease data, physical condition data, lifestyle data, etiology data and age and physical condition data; A second construction module is used to construct, based on the first model, a personalized vector database, a generalized knowledge graph vector database and a personalized case vector database of famous doctors corresponding to the rehabilitation program; And / or, the data processing and generating unit further includes: A third construction module is used to construct a second model corresponding to the first data, wherein the second model is an entity recognition model; A first processing module is used to generate and obtain a first matrix corresponding to the second model, and optimize the sequence order corresponding to the training part of speech based on the first matrix; the first matrix is ​​a state transfer matrix; The first generating module is used to generate and obtain second data corresponding to the recovered patient, and generate corresponding third data based on the second data; wherein the second data is the personalized information data of the recovered patient's main complaint; and the third data is the similarity data between at least two recovered patients; A second generating module is used to generate and obtain fourth data corresponding to the rehabilitation patient, and fuse and process the third data based on the fourth data; wherein the fourth data is symptom weight data and proportional coefficient data; The first calculation module is used to calculate and generate third data corresponding to the rehabilitation patient; wherein the calculation formula is: Similarity C =α1*α2*λ T *S (12); The second calculation module is used to generate and obtain fifth data corresponding to the knowledge graph, and calculate and generate corresponding sixth data based on the fifth data; wherein the fifth data is word data; the sixth data is similarity data between word vectors; the calculation formula is as follows: In the formula, w i and w j are two word vectors, n is the dimension of the word vector, w ik and w jk Respectively represent w i and w j The value of the kth element in ; The third calculation module is used to determine and generate seventh data and eighth data corresponding to the sixth data based on the sixth data; wherein the seventh data is the correlation data between word vectors; the eighth data is the prediction probability data of the word vectors; the calculation formula of the prediction probability data is as follows: In the formula, w i is the word vector of the central word, w j is the word vector of the predicted word, and m is the total number of different words in the corpus.

9. A platform for processing and generating personalized rehabilitation programs based on big data, characterized in that: It includes a processor, a memory and a control program for a platform for processing and generating personalized rehabilitation programs based on big data; wherein the control program for processing and generating personalized rehabilitation programs based on big data is executed on the processor, the control program for processing and generating personalized rehabilitation programs based on big data is stored in the memory, and the control program for processing and generating personalized rehabilitation programs based on big data implements the method for processing and generating personalized rehabilitation programs based on big data as described in any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a control program for a personalized rehabilitation program processing and generating platform based on big data, and the control program for a personalized rehabilitation program processing and generating platform based on big data implements the personalized rehabilitation program processing and generating method based on big data as described in any one of claims 1 to 6.

Images