Scoliosis auxiliary prescription generation device, method and equipment and storage medium

Abstract

The invention discloses a scoliosis auxiliary prescription generation device, method and equipment and a storage medium, and is applied to the technical field of medical treatment, and the device comprises an input obtaining module which is used for obtaining scoliosis traditional Chinese medicine detection parameters and scoliosis western medicine detection parameters; the prescription generation module is used for performing analysis by utilizing a scoliosis traditional Chinese and western medicine auxiliary prescription generation model based on the scoliosis traditional Chinese medicine detection parameters and the scoliosis western medicine detection parameters to obtain a target scoliosis traditional Chinese and western medicine auxiliary prescription; wherein the scoliosis traditional Chinese and western medicine auxiliary prescription generation model is a model obtained by training a machine learning model based on a scoliosis traditional Chinese and western medicine typing theory knowledge base and a traditional Chinese and western medicine clinical treatment scheme of specific people. Based on the scoliosis traditional Chinese medicine detection parameters and the scoliosis western medicine detection parameters of the patient, the scoliosis traditional Chinese and western medicine auxiliary prescription generation model is utilized for analysis, the traditional Chinese and western medicine auxiliary prescription is intelligently generated, and traditional Chinese and western medicine collaborative treatment is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the medical field, in particular to a scoliosis auxiliary prescription generation device, method, equipment and storage medium. BACKGROUND

[0002] The existing scoliosis classification system such as Lenke and PUMC is mainly suitable for surgical path decision, and the classification method for non-surgical intervention (such as Ponseti, Rigo, three-dimensional classification, etc.) has limited applicability in the child population. The classification principle is highly subjective and has poor reproducibility, and it is difficult to provide standardized staging treatment basis. Traditional Chinese medicine diagnosis is based on constitution, tongue and pulse, but lacks structured and quantitative classification standards, making it difficult to establish a unified and objective diagnostic reference framework between traditional Chinese and Western medicine.

[0003] Therefore, how to improve the accuracy of the scoliosis auxiliary treatment scheme generation is a technical problem that those skilled in the art urgently need to solve. SUMMARY

[0004] Therefore, the purpose of the present application is to provide a scoliosis auxiliary prescription generation device, method, equipment and storage medium, which solves the problem of inaccurate scoliosis auxiliary treatment scheme in the prior art.

[0005] To solve the above technical problems, the present application provides a scoliosis auxiliary prescription generation method, comprising:

[0006] An input acquisition module is configured to acquire traditional Chinese medicine detection parameters and Western medicine detection parameters of scoliosis.

[0007] A prescription generation module is configured to analyze a scoliosis traditional Chinese and Western medicine auxiliary prescription generation model based on the traditional Chinese medicine detection parameters and the Western medicine detection parameters of scoliosis, to obtain a target scoliosis traditional Chinese and Western medicine auxiliary prescription. The scoliosis traditional Chinese and Western medicine auxiliary prescription generation model is a model obtained by training a machine learning model based on a specific population scoliosis traditional Chinese and Western medicine classification theory knowledge base and a specific population traditional Chinese and Western medicine clinical treatment scheme.

[0008] Optionally, the scoliosis auxiliary prescription generation device described above further comprises:

[0009] A prescription extraction module is configured to extract a traditional Chinese medicine prescription, a Western medicine prescription and a traditional Chinese and Western medicine combined prescription based on the specific population scoliosis traditional Chinese and Western medicine classification theory knowledge base and the specific population traditional Chinese and Western medicine clinical treatment scheme.

[0010] The training data generation module generates, based on the Chinese medicine prescription, the Western medicine prescription and the integrated Chinese and Western medicine prescription, a Western medicine feature vector of a first preset dimension corresponding to a Western medicine detection index, a Chinese medicine feature vector of a second preset dimension corresponding to a Chinese medicine detection index, and an auxiliary prescription for scoliosis corresponding to a combination of various feature vectors;

[0011] The model training module trains the machine learning model by taking the Western medicine feature vector of the first preset dimension and the Chinese medicine feature vector of the second preset dimension as inputs and taking the auxiliary prescription for scoliosis as output, to obtain the integrated Chinese and Western medicine auxiliary prescription generation model for scoliosis.

[0012] Optionally, the model training module comprises:

[0013] The alignment unit is configured to determine that an input layer of the machine learning model is the Western medicine feature of the first preset dimension and the Chinese medicine feature of the second preset dimension, and align the Western medicine feature of the first preset dimension and the Chinese medicine feature of the second preset dimension.

[0014] The fusion unit is configured to splice the aligned Western medicine feature and Chinese medicine feature to obtain spliced features, and perform cross-modal feature association on the spliced features through multi-head attention learning to obtain fusion features.

[0015] The training unit is configured to train a classifier of the machine learning model based on the fusion features and the auxiliary prescription for scoliosis corresponding to each fusion feature, to obtain the integrated Chinese and Western medicine auxiliary prescription generation model for scoliosis.

[0016] Optionally, the auxiliary prescription generation device for scoliosis further comprises:

[0017] The integrated Chinese and Western medicine feature dynamic coefficient determination module is configured to determine a dynamic coefficient according to patient age and disease duration.

[0018] The fusion feature adjustment module is configured to adjust each Chinese medicine feature and Western medicine feature in the fusion features through the dynamic coefficient to obtain adjusted fusion features.

[0019] Correspondingly, the training unit comprises:

[0020] The training sub-unit is configured to train a classifier of the machine learning model based on the adjusted fusion features and the auxiliary prescription for scoliosis corresponding to each adjusted fusion feature, to obtain the integrated Chinese and Western medicine auxiliary prescription generation model for scoliosis.

[0021] Optionally, the first preset dimension western medicine feature vector comprises a Cobb angle, a Risser sign, an ATR value, a western medicine classification, and a progression factor, the progression factor being determined based on the Cobb angle, the Risser sign, and an age; and the second preset dimension traditional Chinese medicine feature vector comprises a traditional Chinese medicine classification, a three-dimensional posture assessment, an ankle and gait analysis, a tongue image, an electromyographic signal, a body mass table, and a pulse signal.

[0022] Optionally, the scoliosis auxiliary prescription generation device further comprises:

[0023] a prescription adjustment module configured to adjust the target scoliosis traditional Chinese medicine and western medicine auxiliary prescription to obtain an adjusted traditional Chinese medicine and western medicine auxiliary prescription; wherein the adjustment content comprises an implementation result of a prescription category item, a composition, a dosage, and an usage of a traditional Chinese medicine prescription in a traditional Chinese medicine treatment item, and a brace treatment scheme content and a physical therapy training scheme.

[0024] Optionally, the scoliosis auxiliary prescription generation device further comprises:

[0025] a model dynamic adjustment module configured to train the scoliosis traditional Chinese medicine and western medicine auxiliary prescription generation model based on the adjusted traditional Chinese medicine and western medicine auxiliary prescription to obtain an adjusted scoliosis traditional Chinese medicine and western medicine auxiliary prescription generation model.

[0026] The embodiment of the present application also provides a scoliosis auxiliary prescription generation method, comprising:

[0027] obtaining a scoliosis traditional Chinese medicine detection parameter and a scoliosis western medicine detection parameter;

[0028] based on the scoliosis traditional Chinese medicine detection parameter and the scoliosis western medicine detection parameter, analyzing by using a scoliosis traditional Chinese medicine and western medicine auxiliary prescription generation model to obtain a target scoliosis traditional Chinese medicine and western medicine auxiliary prescription; wherein the scoliosis traditional Chinese medicine and western medicine auxiliary prescription generation model is a model obtained by training a machine learning model based on a specific population scoliosis traditional Chinese medicine and western medicine classification theoretical knowledge base and a specific population traditional Chinese medicine and western medicine clinical treatment scheme.

[0029] The embodiment of the present application also provides a scoliosis auxiliary prescription generation device, comprising:

[0030] a memory configured to store a computer program;

[0031] a processor configured to execute the computer program to realize the steps of the method corresponding to the above-mentioned scoliosis auxiliary prescription generation device.

[0032] The embodiment of the present application also provides a computer readable storage medium (namely a storage medium) which has a computer program stored thereon, and the computer program is executed by a processor to implement the steps of the method corresponding to the scoliosis auxiliary prescription generation device.

[0033] The embodiment of the present application also provides a computer program product comprising computer programs / instructions, and the computer programs / instructions are executed by a processor to implement the steps of the method corresponding to the scoliosis auxiliary prescription generation device.

[0034] It can be seen that the scoliosis auxiliary prescription generation device provided by the present application can obtain the traditional Chinese medicine detection parameters and the western medicine detection parameters of scoliosis through the input acquisition module; and the prescription generation module is used for analyzing the traditional Chinese medicine detection parameters and the western medicine detection parameters of scoliosis based on a scoliosis traditional Chinese and western medicine auxiliary prescription generation model to obtain a target scoliosis traditional Chinese and western medicine auxiliary prescription; wherein the scoliosis traditional Chinese and western medicine auxiliary prescription generation model is a model obtained by training a machine learning model based on a specific population scoliosis traditional Chinese and western medicine typing theory knowledge base and a specific population traditional Chinese and western medicine clinical treatment scheme. Compared with the current dependence on artificial experience and the need for doctors to learn and manually adjust the treatment scheme, the scoliosis auxiliary prescription generation device provided by the present application can analyze the input traditional Chinese medicine detection parameters and western medicine detection parameters of scoliosis of a patient based on the scoliosis traditional Chinese and western medicine auxiliary prescription generation model, intelligently generate a target scoliosis traditional Chinese and western medicine auxiliary prescription, and thus realize the collaborative treatment of traditional Chinese medicine and western medicine and improve the comprehensiveness and accuracy of the scoliosis auxiliary prescription generation.

[0035] In addition, the present application also provides a scoliosis auxiliary prescription generation method, equipment and storage medium, which also have the above beneficial effects. BRIEF DESCRIPTION OF DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor based on the provided drawings.

[0037] Figure 1 A structural schematic diagram of a scoliosis auxiliary prescription generation device provided by the embodiment of the present application;

[0038] Figure 2 A schematic diagram of feature fusion provided by the embodiment of the present application;

[0039] Figure 3 A flowchart of a scoliosis auxiliary prescription generation method provided by the embodiment of the present application;

[0040] Figure 4 and Figure 5 A schematic diagram of a scoliosis TCM auxiliary prescription provided by the embodiment of the present application;

[0041] Figure 6 A structural schematic diagram of a scoliosis auxiliary prescription generation device provided by the embodiment of the present application. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0043] Some nouns appearing in the description of the embodiments of the present application or belong to the following explanations:

[0044] Scoliosis: scoliosis;

[0045] Cobb angle: international standard method for measuring scoliosis angle by X-ray film;

[0046] Risser sign: one of the indicators for judging bone maturity, which can evaluate the potential growth of the spine;

[0047] Physiotherapeutic Scoliosis Specific Exercises, PSSE: specific exercise therapy;

[0048] Special Inpatient Rehabilitation, SIR: special inpatient rehabilitation;

[0049] Lenke classification: Lenke et al. proposed a Lenke classification system based on three-dimensional factors of coronal plane, sagittal plane and axial position of the spine;

[0050] PUMC classification: PUMC classification system for idiopathic scoliosis;

[0051] Rigo classification: developed by Dr. Rigo for Rigo Chenau brace design and treatment related;

[0052] Ponseti classification: Ponseti (based on Schulthess's study) proposed a classification method that distinguishes four main types of scoliosis: thoracic, lumbar, thoracolumbar, and S-shaped;

[0053] Angle of Trunk Rotation, ATR: the degree of trunk rotation, is an important indicator for evaluating scoliosis.

[0054] Currently, there are several problems in the non-surgical treatment of scoliosis in children and adolescents, such as inconsistent diagnostic and treatment standards, reliance on individual experience for treatment plan development, fragmentation of Chinese and Western evaluation systems, and low level of intelligentization, which seriously affect the clinical application and replication of the advantages of integrated traditional Chinese and Western medicine. The specific technical bottlenecks include:

[0055] ① Lack of standardized classification system for conservative treatment.

[0056] Existing scoliosis classification systems such as Lenke and PUMC are mainly suitable for surgical path decision-making. The classification methods for non-surgical intervention (such as Ponseti, Rigo, and three-dimensional classification) have limited applicability in the pediatric population. The classification principles are highly subjective and have poor reproducibility, making it difficult to provide standardized staging and treatment guidelines. Traditional Chinese medicine diagnosis is based on constitution, tongue, and pulse, but lacks structured and quantitative classification standards, making it difficult to establish a unified and objective diagnostic reference framework between traditional Chinese and Western medicine.

[0057] ② Lack of integration mechanism for core diagnosis and treatment information between traditional Chinese and Western medicine.

[0058] Traditional Chinese medicine emphasizes syndrome differentiation, balance of organs, and overall regulation of muscles, bones, and blood; Western medicine relies on standardized indicators such as imaging, scales, and biomechanical evaluation. There are significant differences in evaluation dimensions and data expression methods between the two. In clinical practice, there is a lack of systematic integration mechanism based on multi-modal data (such as Cobb angle, Risser sign, posture three-dimensional parameters, constitution score, tongue and pulse image, etc.), making it difficult to achieve mutual recognition and joint intervention between traditional Chinese and Western medicine data, hindering the scientific construction of collaborative treatment paths between traditional Chinese and Western medicine.

[0059] ③ Existing programs lack intelligent and personalized capabilities.

[0060] The current treatment process highly depends on individual experience or fixed rule templates, making it difficult to dynamically adapt to individual complexity factors such as different growth stages, physical differences, deformity types, biomechanical states, and comorbidities. At the same time, the system lacks effective mechanisms to deposit expert experience or knowledge graphs, making it difficult to support intelligent recommendation and adjustment of personalized programs.

[0061] ④ Lack of comprehensive intelligent diagnosis and treatment platform for scoliosis.

[0062] The existing intelligent systems are mostly focused on screening of scoliosis, posture analysis or single rehabilitation intervention, and there is no comprehensive platform that can systematically integrate traditional Chinese medicine (such as traditional Chinese medicine, massage, constitution identification), western medicine (such as brace, rehabilitation training, functional assessment) intervention modules, and support whole-process data collection, fusion analysis, intelligent recommendation and efficacy closed-loop feedback, which limits the application landing for a wide range of people and multi-level institutions.

[0063] Therefore, the core technical problem to be solved by the present application is to construct a children and adolescents (specific population) scoliosis diagnosis and treatment scheme generation device integrating core indicators of traditional Chinese and western medicine, supporting individual difference adaptation, and having standardization and intelligentization characteristics. The scientificity, efficiency and clinical replicability of diagnosis and treatment are improved, and the landing and promotion of the traditional Chinese and western medicine combined diagnosis and treatment mode in the children population with spinal diseases are promoted.

[0064] Please refer to Figure 1 , Figure 1 The structure diagram of a scoliosis auxiliary prescription generation device provided by the embodiment of the present application can include:

[0065] The input acquisition module 100 is configured to acquire traditional Chinese medicine detection parameters of scoliosis and western medicine detection parameters of scoliosis.

[0066] Each step in the embodiment can be executed by a specified electronic device, which can be a server, a portable terminal or other forms. The embodiment does not limit specific traditional Chinese medicine detection parameters of scoliosis. For example, the western medicine detection parameters of scoliosis in the embodiment can include a combination of any information such as Cobb angle, Risser sign, coronal imbalance, three-dimensional posture assessment, foot and gait analysis, body mass table, and electromyographic signal; the traditional Chinese medicine detection parameters of scoliosis in the embodiment can be a combination of information such as constitution, syndrome differentiation type, tongue image, electromyographic signal, body mass table, and pulse signal. The traditional Chinese medicine detection parameters of scoliosis and the western medicine detection parameters of scoliosis in the embodiment can specifically include patient basic information and scoliosis detection parameters; the basic information covers medical record information such as age, gender, height, and weight; the scoliosis disease classification information includes traditional Chinese medicine classification (such as spleen and stomach disorder type, muscle deficiency type, and blockage of meridians type, and their subdivided description and classification), western medicine classification (such as single neck bending, single chest bending, chest and waist bending, single waist bending, neck and chest double bending, chest and waist double bending, and neck, chest and waist multiple bending according to X-ray film diagnosis, and left convex or right convex direction of each part), and scoliosis detection parameters including Cobb angle, Risser sign, coronal imbalance, three-dimensional posture assessment, foot and gait analysis, tongue image, pulse signal, body mass table, electromyographic signal, ATR and other multi-source information detected by X-ray film imaging examination and scoliosis measuring instrument.

[0067] The prescription generation module 200 is configured to analyze the scoliosis TCM detection parameters and the scoliosis Western medicine detection parameters by using a scoliosis TCM and Western medicine auxiliary prescription generation model to obtain a target scoliosis TCM and Western medicine auxiliary prescription. The scoliosis TCM and Western medicine auxiliary prescription generation model is a model obtained by training a machine learning model based on a specific population scoliosis TCM and Western medicine typing theory knowledge base and a specific population TCM and Western medicine clinical treatment scheme.

[0068] The scoliosis TCM and Western medicine auxiliary prescription generation model in this embodiment can automatically issue a TCM and Western medicine combined diagnosis and treatment prescription suitable for the individual condition of a patient based on a specific population (for example, children and adolescents) scoliosis TCM and Western medicine typing theoretical knowledge base, covering TCM prescriptions, TCM manipulation and guidance suggestions, Western medicine exercise rehabilitation programs, brace fitting suggestions, and the like. When reasoning, the embodiment can combine the patient's basic information (for example, age, gender, height, weight, and medical record information) to reason. The embodiment is not limited to a specific machine learning model. For example, the machine learning model in the embodiment can be a random forest model. Alternatively, the machine learning model in the embodiment can also be a reinforcement learning model. The embodiment is not limited to a specific population. For example, the specific population in the embodiment can be adolescents, for example, corresponding to an age of 14-18 years old. Alternatively, the specific population in the embodiment can be children, for example, corresponding to an age of 0-14 years old. Alternatively, the specific population in the embodiment can be children and adolescents. In the embodiment, the TCM and Western medicine theories of children and adolescents scoliosis etiology, pathogenesis, and syndrome differentiation are integrated, the TCM syndrome differentiation based on the holistic concept and treatment based on syndrome differentiation, and the Western medicine classification based on imaging and clinical signs are sorted out. Scoliosis is divided into pre-disease stage and disease stage (stable stage, progression stage, and mature stage), and a "scoliosis TCM and Western medicine combined treatment staging" is innovatively constructed. A unified and systematic specific population scoliosis TCM and Western medicine typing theoretical knowledge base is formed to treat by stages based on TCM and TCM combined new therapy, serving as the basis for subsequent TCM and Western medicine diagnosis and treatment prescription recommendation technical solutions. Based on the above theoretical knowledge base, the embodiment can combine TCM and Western medicine clinical treatment experience to extract commonly used prescriptions, guidance, and manipulation treatment methods for TCM syndrome differentiation and treatment, Western medicine brace correction and rehabilitation training treatment methods, and TCM and Western medicine combined diet and psychological adjustment and SIR treatment methods to form a TCM and Western medicine clinical treatment program. That is, the scoliosis TCM and Western medicine auxiliary prescription in the embodiment can be any combination of the eight treatment categories (TCM, guidance, manipulation, brace treatment, core training, exercise therapy, diet and psychological adjustment, and SIR).TCM includes TCM oral medicine: syndrome differentiation and treatment (spleen and stomach disorder type, muscle deficiency type, collaterals obstruction type, spleen and kidney deficiency type), external treatment of spleen and kidney deficiency type, acupoint sticking (dorsal spine acupoint), moxibustion (lateral bending curve local), daoyin is the skill of TCM, including Yijinjing, Baduanjin, Qubingyanian 20 postures, Wufengxi, health qigong, etc. Adopting spine health daoyin, including starting posture (Baduanjin·Shuangshou Tiantili Sanjiao), Jiugui Buma Dao (Yijinjing·Jiugui Buma Dao), Hengdan Jiangemo Chi (Yijinjing·Hengdan Jiangemo Chi), ending posture (Baohuan Gong·Jiajizi Shi); TCM includes TCM manipulation, including point pressing activation, pulling and stretching, etc. General manipulation and daoyin are together, that is, manipulation + daoyin; Western medicine includes spine lateral bending correction tools; core training targets trunk training, including plank, cat stretch, side support, bird posture; exercise therapy includes physiological logic curve correction, daily activity guidance (ADL), three-dimensional strengthening correction treatment; diet includes calcium, D3 (vitamin D3), protein; SIR: inpatient rehabilitation treatment, western medicine includes brace treatment, exercise therapy. According to the clinical type of patient lateral bending, individualized correction plan is made, course: according to patient condition, 2-6 hours / day. At the same time, it involves TCM and western medicine, including exercise therapy, SIR. Among them, TCM diet can include: before puberty, diet should focus on invigorating the spleen and stomach, benefiting qi and tonifying blood; after puberty, diet should focus on tonifying liver and kidney, strengthening bones and muscles. Huangjing Mixian (ingredients, method omitted), [function] tonifying qi, nourishing yin, and invigorating the spleen. This diet is mainly used for pre-development teenagers, especially suitable for people with weak spleen and stomach, body fatigue. Kidney Yang Deficiency: Patients with kidney yang deficiency may experience symptoms such as fatigue, aversion to cold, weakness of waist and knees, memory loss, and clear urine. Such patients can eat more warm foods, such as mutton, lychee, longan, leek, walnut, and eucommia. Duzhong Waist Flower (ingredients, method omitted), [function] tonifying liver and kidney, strengthening bones and muscles. This diet is mainly used for post-development teenagers, especially suitable for people with kidney yang deficiency. People with yin deficiency and excessive fire should not eat it. Diet management can include: balanced nutrition: children and adolescents should maintain a balanced diet and intake enough nutrients such as protein, carbohydrates, fats, vitamins, and minerals. These nutrients are important for maintaining normal physiological functions, promoting bone development and repair. Light and digestible: choose light and digestible food, avoid spicy, greasy, cold and other irritating food, these food is easy to stimulate the gastrointestinal tract, affect the digestive and absorptive function, and then aggravate the condition of scoliosis. Eat more calcium-rich foods: calcium is an important component of bones and teeth, and plays an important role in maintaining the normal shape of the spine. Children and adolescents should eat more calcium-rich foods, such as milk, soy products, and seafood. Adequate vitamin D supplementation: Vitamin D helps the absorption and utilization of calcium, and is important for the prevention and treatment of scoliosis. Vitamin D can be supplemented by consuming foods rich in vitamin D (such as fish liver oil, egg yolk, etc.).

[0069] It can be understood that, for the characteristics of the prescription recommendation scene of TCM and Western medicine diagnosis and treatment of scoliosis (multimodal feature fusion, sample may exist class imbalance, and consideration of model interpretability and prediction accuracy), the present application can select random forest (Random Forest) as the core machine learning classification algorithm. Core parameter settings: number of decision trees (n_estimators): 200 (determined by 5-fold cross-validation, balancing accuracy and computational efficiency); maximum tree depth (max_depth): 10 (limiting overfitting, adapting to medium sample size); minimum sample size for node splitting (min_samples_split): 10; minimum sample size for leaf node (min_samples_leaf): 5; random feature subset size (max_features): √17≈4 (randomly select 4 features each time). Output layer design: adopt the "one-to-many" (One-vs-Rest) strategy to build a multi-label classifier, train a binary classification sub-model for each of the 8 treatment categories (TCM, daoyin, manipulation, brace treatment, core training, exercise therapy, diet and psychological adjustment, SIR), and finally output an 8-dimensional (can also be other dimensions, each dimension corresponds to one kind of auxiliary treatment method) binary vector (1 represents recommending this category, 0 represents not recommending).

[0070] It needs to be further explained that, based on any of the above embodiments, the scoliosis auxiliary prescription generation device can further comprise:

[0071] A prescription extraction module is configured to extract TCM prescriptions, Western medicine prescriptions, and integrated TCM and Western medicine prescriptions based on the specific population scoliosis TCM and Western medicine typing theory knowledge base and the TCM and Western medicine clinical treatment scheme.

[0072] A training data generation module is configured to generate Western medicine feature vectors of a first preset dimension corresponding to Western medicine detection indicators, and TCM feature vectors of a second preset dimension corresponding to TCM detection indicators, and various feature vector combinations corresponding to scoliosis auxiliary prescriptions based on the TCM prescriptions, the Western medicine prescriptions, and the integrated TCM and Western medicine prescriptions.

[0073] A model training module is configured to train the machine learning model by taking the Western medicine feature vectors of the first preset dimension and the TCM feature vectors of the second preset dimension as inputs and taking the scoliosis auxiliary prescriptions as outputs, to obtain the scoliosis TCM and Western medicine auxiliary prescription generation model.

[0074] The TCM content in the TCM prescription in this embodiment can include three-dimensional posture evaluation, ankle and gait analysis, body mass table, etc. For ease of understanding, please refer to Table 1 and Table 2, Table 1 is a kind of TCM quantitative index schematic table provided by the embodiment of the application, and Table 2 is another kind of TCM quantitative index schematic table, which is the TCM index content included in the TCM prescription and the TCM and Western medicine prescription.

[0075] Table 1 is a kind of TCM quantitative index schematic table

[0076]

[0077] Table 2 is another kind of TCM quantitative index schematic table

[0078]

[0079] In this embodiment, when the scoliosis is divided into the pre-disease stage and the disease stage (stable period, progressive period and mature period), the quantification method can be: stable period: Cobb angle 10-20° and change <5° in 6 months, Risser sign 0-1 level, TCM typing quantitative index <30 points; progressive period: Cobb angle ≥10° and change ≥5° in 6 months, Risser sign 0-3 level, TCM typing quantitative index 30-60 points; mature period: Cobb angle ≥10° and Risser sign 5 level, TCM typing quantitative index >60 points. In this embodiment, the scoliosis auxiliary prescription can combine the clinical treatment experience of TCM and Western medicine, extract the commonly used prescriptions, daoyin and manual therapy methods of TCM syndrome differentiation and treatment, Western medicine brace correction, rehabilitation training treatment methods, and diet and psychological adjustment, SIR treatment methods of TCM and Western medicine combination as the scoliosis auxiliary prescription. Based on the TCM prescription, Western medicine prescription and TCM and Western medicine combination prescription, the first preset dimension of Western medicine characteristic vector corresponding to the Western medicine detection index and the second preset dimension of TCM characteristic vector corresponding to the TCM detection index are generated, which means that the TCM and Western medicine parameters related to scoliosis are extracted from the TCM prescription, Western medicine prescription and TCM and Western medicine combination prescription, for example, 4-dimensional Western medicine characteristic vector M_western (Cobb angle, Risser sign, ATR value, Western medicine typing, progressive factor); 17-dimensional TCM characteristic vector M_chinese (TCM typing, three-dimensional posture evaluation, ankle and gait analysis, tongue image, electromyographic signal, body mass table, pulse signal); the current TCM and Western medicine taste can also be other dimensions. Through learning the relationship between Western medicine, TCM characteristics and scoliosis auxiliary prescription, a model that can give TCM and Western medicine auxiliary prescription for scoliosis is obtained.

[0080] It needs to be further explained that based on the above-mentioned scoliosis auxiliary prescription generation device, the model training module can include: an alignment unit configured to determine that an input layer of the machine learning model is the first preset dimension of western medicine features and the second preset dimension of traditional Chinese medicine features, and align the first preset dimension of western medicine features and the second preset dimension of traditional Chinese medicine features; a fusion unit configured to splice the aligned western medicine features and traditional Chinese medicine features to obtain spliced features, and perform cross-modal feature association on the spliced features through multi-head attention learning to obtain fusion features; and a training unit configured to train a classifier of the machine learning model based on the fusion features and the scoliosis auxiliary prescriptions corresponding to each fusion feature to obtain the scoliosis traditional Chinese and western medicine auxiliary prescription generation model.

[0081] The embodiment does not limit the specific values of the first preset dimension and the second preset dimension. For example, the first preset dimension in the embodiment can be 4, 6, etc., and the second preset dimension can be 17, 20, etc. For example, the western medicine feature vector of the first preset dimension includes a Cobb angle, a Risser sign, an ATR value, a western medicine classification, and a progression factor, and the progression factor is determined based on the Cobb angle, the Risser sign, and the age; and the traditional Chinese medicine feature vector of the second preset dimension includes a traditional Chinese medicine classification, a three-dimensional posture evaluation, an ankle and gait analysis, a tongue image, an electromyographic signal, a body mass table, and a pulse signal. In the western medicine vector, the progression factor calculation method is as follows: according to the input data, the progression risk related factor of scoliosis is calculated. The calculation model of the progression factor calculation method is as follows: ; wherein the disease course coefficient is 1.2 for less than 6 months, 1.0 for 6-12 months, and 0.8 for more than 12 months; when the coronal imbalance occurs, the result is multiplied by a correction coefficient of 1.3. In the embodiment, the alignment refers to the dimension of the smaller one approaching the dimension of the larger one. For example, the 4-dimensional western medicine feature vector M_western is mapped through a fully connected layer + residual mapping to expand to 17 dimensions, matching the traditional Chinese medicine feature dimension: ; wherein FC is a fully connected layer, Repeat is a dimension replication padding to ensure that the original features are retained after dimension upgrading, M_western is the original western medicine feature vector, is the aligned western medicine feature dimension. In the embodiment, the fusion layer based on the machine learning model learns the association weight of the western medicine and traditional Chinese medicine features to realize cross-modal feature association, and the method is as follows: the upgraded western medicine feature is spliced with the traditional Chinese medicine feature M_chinese, and the multi-head attention (Multi-Head Attention) learning cross-modal feature association is performed to output a 17-dimensional fusion vector M_fusion; the core formula is as follows: In this embodiment, the machine learning model is trained with the fused features as input, so that an output scoliosis auxiliary prescription can be given. For ease of understanding, please refer to Figure 2 , Figure 2 A feature fusion diagram is provided for an embodiment of the present application. When processing Western medicine features, the embodiment needs to eliminate dimensional differences, and processes traditional Chinese medicine feature vectors to convert traditional Chinese medicine indicators into computable variables. For example, Min-Max normalization is used to process Western medicine features, and the normalization formula is: . Wherein, x represents Western medicine features, Min is the minimum value, and Max is the maximum value. Through normalization processing, Western medicine features such as Cobb angle, Risser sign, and ATR value, and traditional Chinese medicine features such as traditional Chinese medicine classification, three-dimensional posture assessment, and body mass table quantification value, are converted to the [0, 1] range and can be directly used for feature fusion and training of the model. Traditional Chinese medicine features are mostly qualitative descriptions. Through structured quantization rules (quantization indicator illustration in Table 2), traditional Chinese medicine classification is converted into quantization scores, and through feature extraction methods (quantization indicator illustration in Table 1), it is converted into numerical variables, which are also normalized by Min-Max.

[0082] It should be further explained that based on any of the above embodiments, the scoliosis auxiliary prescription generation device described above can further include:

[0083] A traditional Chinese medicine and Western medicine feature dynamic coefficient determination module for determining a dynamic coefficient according to patient age and disease duration;

[0084] A fused feature adjustment module for adjusting each of the traditional Chinese medicine features and the Western medicine features in the fused features by the dynamic coefficient to obtain adjusted fused features;

[0085] Correspondingly, the training unit includes a training sub-unit for training a classifier of the machine learning model based on the adjusted fused features and the scoliosis auxiliary prescription corresponding to each of the adjusted fused features, to obtain the traditional Chinese medicine and Western medicine auxiliary prescription generation model.

[0086] This embodiment can adjust the size of the feature value based on the dynamic coefficient, thereby providing the accuracy of the fused feature determination. In this embodiment, after fusion, the Western medicine and traditional Chinese medicine features can be fine-tuned again by the dynamic coefficient a (the range can be 0.3-0.7, which is automatically adapted according to patient age and disease duration (disease onset time)). ; wherein, , are the separated Western medicine and traditional Chinese medicine feature weights in the attention mechanism.

[0087] It needs to be further explained that based on any of the above embodiments, the prescription adjustment module can be further included for adjusting the target scoliosis TCM- Western medicine auxiliary prescription to obtain an adjusted TCM- Western medicine auxiliary prescription; wherein the adjustment content includes prescription category item implementation results, composition, dosage and usage of TCM prescriptions in TCM treatment items, and brace treatment scheme content and exercise therapy training scheme. This embodiment can adjust the target scoliosis TCM- Western medicine auxiliary prescription to improve the accuracy of the prescription.

[0088] It needs to be further explained that based on any of the above embodiments, the prescription adjustment module can be further included for adjusting the target scoliosis TCM- Western medicine auxiliary prescription to obtain an adjusted TCM- Western medicine auxiliary prescription; wherein the adjustment content includes prescription category item implementation results, composition, dosage and usage of TCM prescriptions in TCM treatment items, and brace treatment scheme content and exercise therapy training scheme. This embodiment can adjust the target scoliosis TCM- Western medicine auxiliary prescription to improve the accuracy of the prescription.

[0089] It needs to be explained that the order of the modules and units in the above scoliosis auxiliary prescription generation device can be changed without affecting the logic.

[0090] The scoliosis auxiliary prescription generation device provided by the embodiment of the present application can include: an input acquisition module 100 for acquiring scoliosis TCM detection parameters and scoliosis Western medicine detection parameters; a prescription generation module 200 for analyzing the target scoliosis TCM- Western medicine auxiliary prescription by using a scoliosis TCM- Western medicine auxiliary prescription generation model based on the scoliosis TCM detection parameters and the scoliosis Western medicine detection parameters; wherein the scoliosis TCM- Western medicine auxiliary prescription generation model is a model obtained by training a machine learning model based on a specific population scoliosis TCM- Western medicine typing theory knowledge base and a specific population TCM- Western medicine clinical treatment scheme. Compared with the current dependence on artificial experience, the doctor needs to learn and manually adjust the treatment scheme, the present application can analyze the input patient's scoliosis TCM detection parameters and scoliosis Western medicine detection parameters based on the scoliosis TCM- Western medicine auxiliary prescription generation model, intelligently generate the target scoliosis TCM- Western medicine auxiliary prescription, and realize the TCM- Western medicine collaborative treatment.

[0091] Please refer to Figure 3 , Figure 3A flowchart of a scoliosis auxiliary prescription generation method provided by an embodiment of the present application can include:

[0092] S101, obtain a scoliosis TCM detection parameter and a scoliosis Western medicine detection parameter.

[0093] The execution subject of this embodiment is an electronic device. The scoliosis TCM detection parameter and the scoliosis Western medicine detection parameter in this embodiment can be: Western medicine imaging parameters (Cobb angle, Risser sign, coronal imbalance), three-dimensional posture assessment, foot and ankle and gait analysis, body mass table, electromyographic signal, TCM information (constitution, syndrome differentiation type, etc.), and a progressive factor in the Western medicine detection parameter.

[0094] S102, based on the scoliosis TCM detection parameter and the scoliosis Western medicine detection parameter, analyze by using a scoliosis TCM and Western medicine auxiliary prescription generation model to obtain a target scoliosis TCM and Western medicine auxiliary prescription; wherein the scoliosis TCM and Western medicine auxiliary prescription generation model is a model obtained by training a machine learning model based on a specific population scoliosis TCM and Western medicine typing theory knowledge base and a specific population TCM and Western medicine clinical treatment scheme.

[0095] The machine learning model in this embodiment can be a random forest model, at this time, the random forest model training process can include: data set processing: stratified sampling is performed on 720 data, which is divided into a training set and a test set in a ratio of 8:2, the training set is used for model parameter learning and hyperparameter tuning, and the test set is used for independent evaluation of the performance of the final model. The training set is divided into 5 subsets by stratified 5-fold cross-validation, and cross-training is performed. 4 folds are used as the training set, and the remaining 1 fold is used as the validation set. This is repeated 5 times, and the F1 average of the 5 times is taken as the model performance index. Adjust the hyperparameter value combination, repeat the cross-training step, and select the parameter combination with the highest model performance. Model basic parameter setting: determine the basic parameter combination based on 5-fold cross-validation, and the parameter search range and final value are as follows: decision tree number, search range [50, 100, 150, 200, 250, 300], optimal value 200; maximum tree depth, search range [5, 10, 15, 20, None], optimal value 10; minimum sample size for splitting, search range [2, 5, 10], optimal value 10; minimum sample size for leaves, search range [1, 5, 10], optimal value 5; random feature subset size, search range , optimal value Training iteration procedure: trees 1-50, initial random weight training, bootstrap sampling for each tree, random subset size of 4 for features, Gini coefficient as splitting criterion; trees 51-150, dynamic weight adjustment introduced, 1.5 times sampling weight for misclassified samples; trees 151-200, fixed parameter stable training, feature importance calculated every 10 trees. Multi-label classification implementation: One-vs-Rest strategy adopted to build 8 binary classification sub-models, each sub-model targeting 1 treatment category, sigmoid as output layer activation function, threshold determined by F1 maximization on validation set. Dynamic optimization mechanism: model update triggered every 100 new cases included, incremental learning used to update decision tree weights, 70% of historical model knowledge weights preserved. Prescriptions adjusted manually by doctors marked as high confidence samples, 1.5 times weight assigned in next training. The overall procedure can include: data set processing step: stratified sampling is performed on the medical data set (including the specific population scoliosis traditional Chinese medicine classification theory knowledge base and the specific population traditional Chinese and Western medicine clinical treatment scheme) to divide it into a training set and a test set; cross-validation step: the training set is divided into K subsets according to the stratified K-fold cross-validation method, K-1 subsets are used as cross-validation sets, and the remaining 1 subset is used as a validation set for rotating training and validation; hyperparameter optimization step: adjustment is performed within a predefined hyperparameter search space, and the optimal hyperparameter combination is selected based on the average performance indicators of cross-validation; model training step: based on the selected hyperparameter combination, samples are extracted from the training set using bootstrap sampling, and multiple decision trees are constructed based on the Gini coefficient as the splitting criterion to form a random forest model; multi-label classification step: a one-vs-rest strategy is used to build a binary classification sub-model for each treatment category, and a prediction probability is output through an activation function; dynamic update step: when new case data is included, the model weight is updated in an incremental learning manner, and high confidence samples manually labeled are given higher learning weights. The embodiment takes the scoliosis traditional Chinese medicine detection parameters and the scoliosis western medicine detection parameters as inputs, inputs them into the scoliosis traditional Chinese and western medicine auxiliary prescription generation model, and can obtain a target scoliosis traditional Chinese and western medicine auxiliary prescription. The machine learning model in the embodiment can be a random forest model. For ease of understanding, please refer to Figure 4 and Figure 5 , Figure 4 and Figure 5 together constitute a scoliosis traditional Chinese and western medicine auxiliary prescription, Figure 4 and Figure 5 are a schematic diagram of a scoliosis traditional Chinese and western medicine auxiliary prescription provided by the embodiment. The scoliosis auxiliary prescription generation method in the embodiment corresponds to the scoliosis auxiliary prescription generation device described above, so for specific details, please refer to the description of the scoliosis auxiliary prescription generation device.

[0096] The application fuses western medicine imaging parameters (such as Cobb angle, Risser sign), western medicine classification and traditional Chinese medicine constitution and syndrome information (such as spleen and stomach disorder type, muscle deficiency type and the like), three-dimensional posture evaluation, foot and ankle and gait analysis, and multi-source information such as muscle electrical signals to form a multi-dimensional evaluation system. Therefore, the adaptive traditional Chinese and western medicine auxiliary prescription can be generated according to the specific classification and individual differences (such as the degree of lateral bending, growth potential and constitution characteristics) of the patient, the limitations of the traditional single classification or simple superposition treatment means are avoided, the matching degree of the scheme and the actual condition of the patient is improved, and the generated treatment scheme is more comprehensive, scientific and personalized.

[0097] The synergism of traditional Chinese and western medicine is enhanced. The western medicine quantitative data and traditional Chinese medicine syndrome results are associated through the multi-modal biological index fusion model, instead of simply splicing the two treatment methods. Therefore, the synergy of traditional Chinese medicine holistic conditioning (such as traditional Chinese medicine prescription and manual guiding) and western medicine precise intervention (such as brace treatment and exercise rehabilitation) can be realized, the comprehensive advantages of “overall adjustment + local adjustment” are fully exerted, and the disconnection between traditional Chinese and western medicine diagnosis and treatment is solved.

[0098] The subjectivity can be reduced and the consistency of the scheme can be improved. Since the prescription is generated based on the model, the excessive dependence on the individual experience of the doctor is reduced, the scheme differences caused by the experience differences of different doctors are reduced, the consistency of the diagnosis and treatment schemes between different medical institutions and different doctors is improved, and the clinical promotion and standardized treatment are more favorable.

[0099] More comprehensive diagnosis and treatment contents are covered, and the individual needs are met. Since the prescription recommendation of the application covers multiple aspects such as traditional Chinese medicine prescription, traditional Chinese medicine manual guiding, western medicine exercise rehabilitation, brace adaptation, diet and psychological adjustment, and the doctor can adjust the details according to the special conditions (such as drug allergy and treatment tolerance) of the patient, the systematized and full-dimensional diagnosis and treatment scheme can be provided for the children and adolescents with scoliosis, instead of a single exercise or traditional Chinese medicine prescription, the individual needs and actual medical conditions of the patient are considered, and the feasibility and individualization degree of the scheme are improved.

[0100] A scoliosis auxiliary prescription generation device provided by an embodiment of the application is described below. The scoliosis auxiliary prescription generation device described below can be correspondingly referred to the scoliosis auxiliary prescription generation method described above.

[0101] Please refer to Figure 6 , Figure 6 The structural schematic diagram of the scoliosis auxiliary prescription generation device provided by an embodiment of the application can include:

[0102] The memory 10 is used to store the computer program.

[0103] The processor 20 is used to execute the computer program to realize the scoliosis auxiliary prescription generation method described above.

[0104] The memory 10, the processor 20, and the communication interface 30 communicate with each other through the communication bus 40.

[0105] In the embodiment of the present application, the memory 10 stores one or more programs, which can include program codes including computer operation instructions. In the embodiment of the present application, the memory 10 can store programs for implementing the following functions:

[0106] Obtain the Chinese medicine detection parameters of scoliosis and the western medicine detection parameters of scoliosis.

[0107] Based on the Chinese medicine detection parameters of scoliosis and the western medicine detection parameters of scoliosis, analyze by using a Chinese and western medicine assisted prescription generation model to obtain a target Chinese and western medicine assisted prescription of scoliosis. The Chinese and western medicine assisted prescription generation model is a model obtained by training a machine learning model based on a specific population Chinese and western medicine typing theory knowledge base and a specific population Chinese and western medicine clinical treatment scheme.

[0108] In a possible implementation, the memory 10 can include a program storage area and a data storage area. The program storage area can store an operating system and at least one application required by a function, etc. The data storage area can store data created during use.

[0109] In addition, the memory 10 can include a read-only memory and a random access memory, and provide instructions and data for the processor. A part of the memory can also include an NVRAM. The memory stores an operating system and operation instructions, executable modules or data structures, or subsets thereof, or an extended set thereof, wherein the operation instructions can include various operation instructions for implementing various operations. The operating system can include various system programs for implementing various basic tasks and processing hardware-based tasks.

[0110] The processor 20 can be a central processing unit (CPU), an application specific integrated circuit, a digital signal processor, a field programmable gate array, or other programmable logic device. The processor 20 can be a microprocessor or any conventional processor, etc. The processor 20 can call programs stored in the memory 10.

[0111] The communication interface 30 can be an interface of a communication module, used for connecting with other devices or systems.

[0112] Of course, it should be noted that, Figure 6 The structure shown does not constitute a limitation on the scoliosis assisted prescription generation device in the embodiment of the present application. In actual application, the scoliosis assisted prescription generation device can include more or fewer components than those shown. Figure 6More or less components are shown, or certain components are combined.

[0113] The computer readable storage medium provided by the embodiments of the present application is described below, and the computer readable storage medium described below can be referred to in combination with the scoliosis auxiliary prescription generation method described above.

[0114] The present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps of the scoliosis auxiliary prescription generation method described above.

[0115] The computer readable storage medium can include a U disk, a mobile hard disk, a read-only memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0116] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts of each embodiment can be referred to each other. For the device disclosed by the embodiments, since it corresponds to the method disclosed by the embodiments, the description is relatively simple, and the related parts can be referred to the method part.

[0117] The skilled person can further realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized by electronic hardware, computer software or a combination of the two. In order to clearly show the interchangeability of hardware and software, the composition and steps of each example have been described in the above description. Whether the functions are realized by hardware or software depends on the specific application and design constraints of the technical solution. The skilled person can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0118] Finally, it should be noted that in this paper, relationships such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations. Moreover, the terms "include", "contain" or any other variant are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device.

[0119] The method, device, equipment and storage medium for generating a scoliosis auxiliary prescription are described in detail above, the principle and implementation manner of the present application are described by applying specific examples in this paper, and the above example is only used to help understand the method and core idea of the present application; meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation manner and application range will be changed, and the above description should not be understood as the limitation of the present application.

Claims

1. A scoliosis-assisted prescription generation device, characterized in that, include: The input acquisition module is used to acquire traditional Chinese medicine detection parameters and Western medicine detection parameters for scoliosis. The prescription generation module is used to analyze the scoliosis TCM detection parameters and the scoliosis Western medicine detection parameters using a scoliosis TCM and Western medicine auxiliary prescription generation model to obtain a target scoliosis TCM and Western medicine auxiliary prescription; wherein, the scoliosis TCM and Western medicine auxiliary prescription generation model is a model obtained by training a machine learning model based on a TCM and Western medicine classification theory knowledge base for scoliosis in a specific population and TCM and Western medicine clinical treatment plans for a specific population.

2. The scoliosis-assisted prescription generation device according to claim 1, characterized in that, Also includes: The prescription extraction module is used to extract traditional Chinese medicine prescriptions, Western medicine prescriptions, and integrated traditional Chinese and Western medicine prescriptions based on the theoretical knowledge base of the traditional Chinese and Western medicine classification of scoliosis in the specific population and the clinical treatment plan of traditional Chinese and Western medicine in the specific population. The training data generation module generates, based on the traditional Chinese medicine prescription, the Western medicine prescription, and the integrated traditional Chinese and Western medicine prescription, a first preset dimension of Western medicine feature vector corresponding to the Western medicine detection indicators, a second preset dimension of traditional Chinese medicine feature vector corresponding to the traditional Chinese medicine detection indicators, and scoliosis auxiliary prescriptions corresponding to various combinations of feature vectors. The model training module takes the Western medicine feature vector of the first preset dimension and the traditional Chinese medicine feature vector of the second preset dimension as input, and the scoliosis auxiliary prescription as output to train the machine learning model, thereby obtaining the scoliosis traditional Chinese and Western medicine auxiliary prescription generation model.

3. The scoliosis-assisted prescription generation device according to claim 2, characterized in that, The model training module includes: An alignment unit is used to determine that the input layer of the machine learning model consists of Western medicine features of the first preset dimension and traditional Chinese medicine features of the second preset dimension, and to align the Western medicine features of the first preset dimension and the traditional Chinese medicine features of the second preset dimension. The fusion unit is used to splice the aligned Western medicine features and traditional Chinese medicine features to obtain spliced ​​features, and to perform cross-modal feature association on the spliced ​​features through multi-head attention learning to obtain fused features; The training unit trains the classifier of the machine learning model based on the fusion features and the corresponding scoliosis auxiliary prescriptions to obtain the scoliosis traditional Chinese and Western medicine auxiliary prescription generation model.

4. The scoliosis-assisted prescription generation device according to claim 3, characterized in that, Also includes: The module for determining dynamic coefficients of traditional Chinese and Western medicine characteristics is used to determine dynamic coefficients based on the patient's age and disease course. The fusion feature adjustment module is used to adjust each of the traditional Chinese medicine features and the Western medicine features in the fusion feature through the dynamic coefficient to obtain the adjusted fusion feature; Accordingly, the training unit includes: The training subunit is used to train the classifier of the machine learning model based on the adjusted fusion features and the scoliosis auxiliary prescriptions corresponding to each of the adjusted fusion features, so as to obtain the scoliosis traditional Chinese and Western medicine auxiliary prescription generation model.

5. The scoliosis-assisted prescription generation device according to claim 2, characterized in that, The first preset dimension of Western medicine feature vectors includes Cobb angle, Risser sign, ATR value, Western medicine classification, and progressive factor, wherein the progressive factor is determined based on Cobb angle, Risser sign, and age; the second preset dimension of traditional Chinese medicine feature vectors includes traditional Chinese medicine classification, three-dimensional posture assessment, foot and ankle and gait analysis, tongue image, electromyography signal, body mass scale, and pulse signal.

6. The scoliosis-assisted prescription generation device according to any one of claims 1 to 5, characterized in that, Also includes: The prescription adjustment module is used to adjust the traditional Chinese and Western medicine auxiliary prescription for the target scoliosis to obtain the adjusted traditional Chinese and Western medicine auxiliary prescription. The adjustment includes the implementation results of the prescription category items, the composition, dosage and usage of the traditional Chinese medicine prescription in the traditional Chinese medicine treatment items, as well as the content of the brace treatment plan and the exercise therapy training plan.

7. The scoliosis-assisted prescription generation device according to claim 6, characterized in that, Also includes: The model dynamic adjustment module is used to train the scoliosis trait medicine auxiliary prescription generation model based on the adjusted trait medicine auxiliary prescription to obtain the adjusted scoliosis trait medicine auxiliary prescription generation model.

8. A method for generating an auxiliary prescription for scoliosis, characterized in that, include: Obtain TCM and Western medicine test parameters for scoliosis; Based on the TCM detection parameters and the Western medicine detection parameters of scoliosis, the TCM and Western medicine auxiliary prescription generation model for scoliosis is used for analysis to obtain the target TCM and Western medicine auxiliary prescription for scoliosis; wherein, the TCM and Western medicine auxiliary prescription generation model for scoliosis is a model obtained by training a machine learning model based on the theoretical knowledge base of TCM and Western medicine classification of scoliosis in a specific population and the TCM and Western medicine clinical treatment plan for a specific population.

9. A scoliosis-assisted prescription generation device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the steps of the method corresponding to the scoliosis assistive prescription generation device as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the method corresponding to the scoliosis assistive prescription generation device as described in any one of claims 1 to 7.

Images