Neurology rehabilitation degree evaluation method and system

Abstract

The invention discloses a neurology rehabilitation degree evaluation method and system, and relates to the technical field of medical rehabilitation evaluation.The method comprises the following specific steps that data collection is conducted, specifically, whole genome sequencing is conducted on a neurology patient, and whole genome data of the patient is obtained, the method comprises the following steps: collecting initial disease severity degree data of a patient, and evaluating the symptom expression of the patient when the patient sees a doctor for the first time by a professional neurology doctor according to a set disease evaluation standard. A key gene screening strategy and a synaptic plasticity index establishment method are applied to construct a multi-dimensional and more accurate neurology rehabilitation degree evaluation system, so that the neural function repair level is deeply known from the root level of nerve cell connection reconstruction; and accurate prediction and real-time evaluation of the rehabilitation process are realized through the personalized rehabilitation trajectory prediction model, so that the real rehabilitation state of the patient can be reflected more comprehensively and meticulously.

Description

Technical Field

[0001] The present invention relates to the technical field of medical rehabilitation evaluation, and in particular to a method and system for evaluating the degree of rehabilitation in neurology. Background Art

[0002] The field of neurological disease rehabilitation is a vital part of medical research, covering a variety of diseases, such as cerebral infarction, cerebral hemorrhage, Parkinson's disease, etc. These diseases are often accompanied by varying degrees of neurological dysfunction, which seriously affects the quality of life of patients. With the continuous advancement of medical technology, the evaluation of the degree of rehabilitation of neurological diseases has received increasing attention. Rehabilitation evaluation not only helps doctors accurately grasp the patient's rehabilitation progress, but also provides a scientific basis for the formulation and adjustment of rehabilitation plans. In recent years, with the rapid development of gene sequencing technology and the in-depth study of neuroscience, the application of individual genetic data and microscopic characteristics of brain neuron synapses in neurological rehabilitation evaluation has gradually attracted attention, providing new perspectives and possibilities for rehabilitation evaluation.

[0003] However, traditional methods for evaluating the degree of rehabilitation in neurology mostly focus on macroscopic neurological performance, such as using neurological deficit rating scales to assess the recovery of patients' motor, sensory, cognitive and other functions. Although these macroscopic evaluations can reflect the patient's rehabilitation status to a certain extent, they often lack in-depth consideration of the root level of neurological function repair. In particular, when using genetic data and synaptic microscopic features for evaluation, traditional methods have obvious limitations. For example, in terms of gene screening, traditional methods mostly rely on general gene screening models and lack precise positioning of neural development and repair mechanisms specific to the rehabilitation of neurological diseases. This leads to inaccurate extraction of key gene information and makes it difficult to fully reflect the genetic characteristics of individual patients in neural repair. In addition, in the setting of synaptic plasticity indicators, traditional methods are often simple and one-sided, making it difficult to fully reflect the degree of reconstruction of connections between nerve cells. These limitations limit the accuracy and effectiveness of traditional methods in evaluating the degree of rehabilitation in neurology.

[0004] In view of the above problems, it is necessary to optimize the existing neurological rehabilitation degree evaluation method and system, integrate individual genetic information and neuronal synaptic microstructure and function-related data through data analysis and indicator establishment, and build a more comprehensive and accurate rehabilitation degree evaluation system. Therefore, it is of great significance to develop a neurological rehabilitation degree evaluation method and system that can comprehensively realize the above characteristics. Summary of the invention

[0005] The purpose of the present invention is to make up for the shortcomings of the prior art and provide a method and system for evaluating the degree of rehabilitation in neurology. It can use innovative data analysis and indicator establishment methods to integrate individual gene-level information and data related to the microstructure and function of neuronal synapses, and construct a more comprehensive and accurate rehabilitation degree evaluation system. By deeply exploring the key gene information closely related to neural development and repair, and creatively establishing quantitative indicators of synaptic plasticity, the present invention can deeply understand the neural function repair situation from the root. At the same time, with the help of artificial intelligence algorithms, a personalized rehabilitation trajectory prediction model is constructed to achieve accurate prediction of the patient's rehabilitation process, real-time evaluation and dynamic adjustment of the rehabilitation plan. Compared with traditional methods, the present invention not only improves the accuracy and effectiveness of rehabilitation evaluation, but also provides more targeted and scientific technical support for the rehabilitation treatment of neurological diseases, which helps to significantly improve the rehabilitation effect and improve the quality of life of patients.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: On the one hand, a method for evaluating the degree of rehabilitation in neurology comprises the following specific steps: Data collection: Whole genome sequencing is performed on patients in the Department of Neurology to obtain the whole genome data of the patients. The initial severity of the disease is collected and the symptoms of the patients at the time of the first visit are evaluated by professional neurologists according to the established disease assessment criteria. At the same time, the basic physical condition data of the patients are collected. In addition, the morphology, number and density of brain neuron synapses are regularly observed and recorded with the help of super-high-resolution microscopy imaging technology combined with nerve cell labeling methods. In combination with in vivo electrophysiological recording technology, the dynamic changes of synaptic transmission efficiency and postsynaptic potential change functional parameters during the rehabilitation process are monitored; Extraction of key gene information and establishment of synaptic plasticity indicators: Import the data obtained by whole genome sequencing into the bioinformatics analysis software platform, perform quality screening, sequence alignment and annotation operations on the original gene sequence data, and use bioinformatics analysis and medical knowledge integration strategies based on the neural development and repair mechanisms unique to the rehabilitation of neurological diseases to deeply explore the gene information closely related to neural development and repair. The microstructural feature data of neuronal synapses collected by super-high-resolution microscopy imaging technology, as well as the synaptic transmission efficiency and postsynaptic potential change parameter data obtained by in vivo electrophysiological recording technology, are used to reflect the degree of connection reconstruction between nerve cells through quantitative calculation. For the synaptic microstructural feature data, image analysis algorithms are used to compare and analyze the image data collected at different time points to calculate the synaptic newborn rate. For the synaptic function parameter data, the data first recorded at the beginning of rehabilitation treatment is used as the baseline. At fixed time intervals, the parameter changes at different time nodes are compared and analyzed, and the quantitative indicators reflecting the changes in synaptic function plasticity, such as the improvement in synaptic transmission efficiency, are obtained through algorithms. Construction of a personalized rehabilitation trajectory prediction model: The extracted key gene information, the established synaptic plasticity quantitative index, and the collected patient's initial disease severity data and basic physical condition data are used as input features, and an artificial intelligence algorithm is used to construct a personalized rehabilitation trajectory prediction model. Through training and learning on a large number of existing neurological rehabilitation case data, the model can predict the patient's expected rehabilitation status indicators at different time points based on the input new patient feature data, forming a personalized rehabilitation trajectory prediction curve; Real-time evaluation of rehabilitation degree: During the patient's rehabilitation process, the patient's current actual rehabilitation status indicators are collected at predetermined time intervals, and the actual rehabilitation status indicators are compared and analyzed with the predicted rehabilitation status indicators of the corresponding time nodes output by the prediction model, and the difference between the two is calculated. The evaluation result of the current patient's rehabilitation degree is determined based on the evaluation criteria; Adjustment of evaluation index weights and optimization of rehabilitation plans: Based on the difference between the actual rehabilitation performance and the predicted trajectory in the real-time evaluation results, the weights of each indicator in the subsequent rehabilitation evaluation are dynamically adjusted. At the same time, based on the adjusted evaluation index weights and the actual rehabilitation progress, the subsequent rehabilitation plan is optimized, and the content, intensity and frequency of rehabilitation training are adjusted to generate a personalized rehabilitation plan that is more in line with the individual patient's current rehabilitation needs.

[0007] Furthermore, in the steps of extracting key gene information and establishing synaptic plasticity indicators, based on the neural development and repair mechanisms unique to the rehabilitation of neurological diseases, the key gene information closely related to neural development and repair is deeply explored. Specifically, with reference to a large number of authoritative clinical research results, basic medical experimental data and professional medical knowledge bases in the field of neurology, according to the role of genes in the neural development and repair links including nerve cell differentiation, neurotransmitter regulation, neuron regeneration and synaptic plasticity regulation, screening rules and algorithms are set to screen out key gene information from the pre-processed whole genome data, and integrate gene sequence characteristics, functional descriptions and related regulatory information to construct a disease-specific key gene data set that is exclusive to individual patients.

[0008] Furthermore, in the steps of extracting key gene information and establishing synaptic plasticity index, screening rules and algorithms are set to screen out key gene information from the pre-processed whole genome data, and the algorithm formula is: ,in, Indicates The comprehensive association score of each gene, It is The weight coefficient of the screening factor, Representative The gene in The correlation degree value under the screening factors, represents the total number of screening factors, The final selected key gene set meets its comprehensive association score Greater than or equal to the preset screening threshold genes were included in this set.

[0009] Furthermore, in the steps of extracting key gene information and establishing synaptic plasticity indicators, an image analysis algorithm is used to compare and analyze image data collected at different time points to calculate the synaptic newborn rate, and the calculation formula is: ,in, Indicated in brain area In time The synaptic birth rate, Represented in brain area within, no. Types of synapses at time The number of newly formed synapses, image data collected by super-resolution microscopy imaging technology, It is Types of synapses in brain regions The importance weight coefficient of Indicated in brain area within, no. The original number of synapses of the synapse type, In this brain region The total number of synapse types classified within the region covers all synapse types that are important for neural function in this area and whose new formation needs to be paid attention to.

[0010] Furthermore, in the steps of extracting key gene information and establishing synaptic plasticity indicators, quantitative indicators reflecting changes in synaptic functional plasticity, such as the degree of improvement in synaptic transmission efficiency, are obtained through an algorithm, and the algorithm formula is: ,in, Indicated in brain area In time The improvement in synaptic transmission efficiency, In the brain area The postsynaptic neuron The electrical activity potential value, The time from the start of rehabilitation to observation Any time variable in the range, Represented in brain area The presynaptic neuron is The electrical activity potential value, represents the initial time point when rehabilitation treatment begins, which is used as the baseline time for calculating the improvement in synaptic transmission efficiency. represents the initial state of synaptic transmission efficiency at the beginning of recovery, is the specific point in the rehabilitation process at which the assessment is currently being made.

[0011] Furthermore, in the step of constructing the personalized rehabilitation trajectory prediction model, an artificial intelligence algorithm is used to construct a personalized rehabilitation trajectory prediction model, and a large amount of existing neurological rehabilitation case data is trained and learned. Specifically, the input feature vector is set to ,in , which correspond to the feature vectors of key gene information, synaptic plasticity quantitative indicators, initial disease severity data, and basic physical condition data, respectively. The hidden layer of the network is represented as , the rehabilitation status indicator vector predicted by the output layer is , for the calculation from the input layer to the first hidden layer ,in, is the activation function, is the corresponding input feature in the first hidden layer and The weight parameter of the interaction, Represents a custom feature fusion operation, is the bias term of the first hidden layer. Hidden layers to The propagation calculation formula of the hidden layer is: ,in, It is The hidden layer corresponds to the previous layer The connection weights of the outputs of neurons, Indicates The hidden layer The output value of a neuron is the result of the previous layer processed by the activation function and serves as the input of the current layer. It is The bias term of the hidden layer is added after the hidden layer. The attention mechanism layer is used to dynamically assign different attention weights to different input features to highlight the feature information that is more critical to the current rehabilitation trajectory prediction. The formula is: ,in, Represents the calculated unnormalized attention score, which is used to measure the importance of each feature dimension in the current rehabilitation trajectory prediction task. Its calculation involves the weight matrix and Output the hidden layer separately and the original input features Perform a linear transformation and add a bias After Activation function processing, is the normalized attention weight, is the input feature vector after attention weighting. After the attention mechanism layer, the weighted feature vector is used The final recovery status index is predicted using the following calculation formula: ,in, is the weight matrix of the output layer, is the bias term of the output layer. In order to train the above network model, the loss function is used to measure the predicted rehabilitation status index Actual recovery status indicators The difference between them is calculated as: ,in, is the total number of predicted time nodes, is the weight coefficient corresponding to each time node, and At the time nodes The predicted rehabilitation status index value and the actual rehabilitation status index value, represents the number of different stages into which the rehabilitation process is divided, It is The number of time nodes included in each rehabilitation stage, and In the Rehabilitation stage time points The predicted and actual rehabilitation status index values, is the variance of the actual rehabilitation status index value calculated at this rehabilitation stage, It is a hyperparameter that balances the two losses. During the training process, an optimization algorithm is used to minimize the above loss function. The weight parameters of the model are updated and adjusted according to the loss function. After multiple iterative training, the performance indicators of the model on the validation set meet the requirements, and a trained personalized rehabilitation trajectory prediction model is obtained.

[0012] Furthermore, in the step of real-time evaluation of the degree of rehabilitation, during the patient's rehabilitation process, the patient's current actual rehabilitation status indicators are collected at predetermined time intervals, including but not limited to the recovery of specific neural functions, changes in related body function indicators, and the latest neuronal synaptic microstructural characteristics and functional parameter data.

[0013] Furthermore, in the step of real-time evaluation of rehabilitation degree, for limb motor function test, the operation is carried out according to standardized movement tasks and scoring criteria, and the sensory function test adopts standard sensory threshold detection means, that is, by using professional equipment such as tactile tester and pain tester, the skin is stimulated step by step from low to high stimulation intensity, and the minimum intensity value that the patient can perceive the stimulation is recorded as the sensory threshold for quantitative evaluation, and the cognitive function evaluation uses a professional scale that has been verified by reliability and validity to ensure that the operation of each neurological function test process is standardized and the results are reliable.

[0014] Furthermore, in the step of real-time evaluation of the degree of rehabilitation, the actual rehabilitation status index is compared and analyzed with the predicted rehabilitation status index of the corresponding time node output by the prediction model, and the difference between the two is calculated, and the calculation formula is: ,in, Refers to the comprehensive rehabilitation difference, It is in The current rehabilitation status indicator value actually collected in the rehabilitation status indicator dimension, Indicates that the prediction model is in the corresponding The predicted value output on the rehabilitation status indicator dimension, The total number of rehabilitation status indicator dimensions included in the comprehensive consideration.

[0015] On the other hand, a neurology rehabilitation degree evaluation system is characterized in that the system includes the following components: a data acquisition module, a key gene and index processing module, a prediction model construction module, a rehabilitation evaluation module and an adjustment and optimization module; The data acquisition module is used to perform whole genome sequencing on neurology patients, obtain whole genome data, and collect the patient's initial disease severity data and basic physical condition data. At the same time, it uses ultra-high resolution microscope imaging technology and in vivo electrophysiological recording technology to collect brain neuron synaptic microstructural characteristics and functional parameter data; The key gene and index processing module uses bioinformatics analysis tools to extract key gene information related to neural development and repair from whole genome data, and analyzes and processes the collected synaptic microscopic data to establish quantitative indicators related to synaptic plasticity; The prediction model building module uses key gene information, synaptic plasticity quantitative indicators, initial disease severity data, and basic physical condition data as inputs, and uses artificial intelligence algorithms to build and train a personalized recovery trajectory prediction model; The rehabilitation evaluation module compares and analyzes the actual rehabilitation status index with the predicted rehabilitation status index output by the prediction model to obtain a rehabilitation degree evaluation result; The adjustment and optimization module dynamically adjusts the evaluation index weights and optimizes the rehabilitation plan according to the actual and predicted differences in the rehabilitation degree evaluation results, thereby generating a new personalized rehabilitation plan.

[0016] Compared with the prior art, the neurological rehabilitation degree evaluation method and system have the following beneficial effects: 1. The present invention integrates individual gene data with neuronal synaptic micro-feature data, and uses key gene screening strategies and synaptic plasticity index establishment methods to construct a multi-dimensional and more accurate neurological rehabilitation degree evaluation system. It not only deeply understands the level of neural function repair from the root level of nerve cell connection reconstruction, but also realizes accurate prediction and real-time evaluation of the rehabilitation process through a personalized rehabilitation trajectory prediction model, thereby being able to more comprehensively and meticulously reflect the patient's true rehabilitation status.

[0017] 2. The present invention realizes dynamic adjustment and optimization of the rehabilitation plan through the synergistic effect of the rehabilitation evaluation module and the adjustment and optimization module. According to the difference between the actual rehabilitation performance and the predicted trajectory in the real-time evaluation results, the weight of each indicator in the subsequent rehabilitation evaluation is dynamically adjusted, and the subsequent rehabilitation plan is optimized based on the adjusted evaluation indicator weights and the actual rehabilitation progress, ensuring that the rehabilitation plan can closely fit the individual patient's current rehabilitation needs, which helps to improve the rehabilitation effect and improve the patient's quality of life.

[0018] Other advantages, objectives and features of the present invention will be set forth in part in the following description and, in part, will be apparent to those skilled in the art based on an examination of the following or may be taught from the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0020] Figure 1 This is a flowchart of a neurological rehabilitation degree evaluation method; Figure 2 This is a flowchart of a neurological rehabilitation degree evaluation system. DETAILED DESCRIPTION

[0021] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation mode, structure, characteristics and effects of the present invention are described in detail below in combination with the accompanying drawings and preferred embodiments.

[0022] Embodiment 1: This embodiment describes in detail a neurological rehabilitation degree evaluation method and system for application in rehabilitation degree evaluation and plan adjustment for patients with brain trauma. Through the present invention, multi-level data are deeply integrated to create a personalized rehabilitation evaluation and plan adjustment system for patients with brain trauma, which helps to improve the scientificity and accuracy of rehabilitation treatment and effectively improve the rehabilitation effect of patients.

[0023] A 35-year-old male patient was admitted to the hospital due to craniocerebral injury caused by a car accident. He was diagnosed with moderate brain trauma, accompanied by skull fracture, cerebral contusion and laceration, and a small amount of intracranial hemorrhage. The patient was in a coma when he was admitted to the hospital. He woke up after emergency treatment, but he had symptoms of neurological damage such as right limb movement disorder, slightly delayed language expression, and memory loss. The patient underwent whole genome sequencing. The sequencing work was performed by a professional medical testing agency in accordance with standard procedures, and high-quality sequence data covering the patient's entire genome was obtained. The neurologist used the Glasgow Coma Scale (GCS) combined with the head The results of CT and other imaging examinations were used to evaluate the initial severity of the patient's condition, and the initial GCS score was determined to be 10 points (moderate brain trauma range). At the same time, the specific location of the patient's brain contusion and laceration, the amount of bleeding and other detailed information related to the condition were recorded as the severity data of the condition. The patient's basic physical condition data, including the patient's age and gender information, were collected to confirm that the patient had a healthy past and no history of underlying diseases. Routine physiological indicators such as blood pressure, blood sugar, and blood lipids were measured, and all indicators were within the normal range. These data were collated and recorded. At the same time, with the help of a super-resolution fluorescence microscope, a specific The fluorescent dyes for marking neuronal synapses are used to mark the neuronal synapses in the frontal lobe (closely related to cognitive and motor functions, and an area that is easily damaged by brain trauma) and parietal lobe (a brain area that controls the sensory and motor functions of the right limbs) of the patient's brain according to the marking method. Every week after the patient's rehabilitation treatment begins, the marked area is imaged and observed using a microscope to obtain high-resolution image data of the neuronal synapses. The morphology (such as measuring synaptic length, width, area and other parameters), number, density and other microstructural feature data of the synapses are extracted and recorded using professional image analysis software. Each collected data is accompanied by detailed metadata information such as imaging time and observation area coordinates for storage. At the same time, in vivo electrophysiological recording technology is used to accurately implant microelectrodes in the corresponding brain areas of the frontal lobe and parietal lobe of the brain (with the help of a neuronavigation system to ensure the accuracy of the implantation position), and functional parameters such as synaptic transmission efficiency and postsynaptic potential changes are recorded in real time. During the rehabilitation process, electrophysiological data are continuously collected at intervals of every 3 days, and it is ensured that the patient's current body position, body temperature and other environmental parameters as well as accurate timestamps are recorded during each collection, and the collected electrophysiological data are stored in a special database in an orderly manner.

[0024] The whole genome sequencing data was imported into a professional bioinformatics analysis platform. After data preprocessing (removal of low-quality sequencing fragments, comparison with the human reference genome, gene annotation, etc.), based on the neural development and repair mechanisms involved in brain trauma rehabilitation, such as nerve cell repair, neural plasticity regulation, and vascular regeneration, and according to the participation of genes in neuronal survival-related signaling pathways, genes that have been reported to be significantly associated with brain trauma rehabilitation, and genes that play a key role in the regulation of synaptic plasticity, the set screening rules and algorithms were used to screen out key gene information from the whole genome data, forming a data set containing about 50 key genes, including BDNF (brain-derived neurotrophic factor gene, which plays an important role in neuronal growth and survival), NRG1 (neuroregulatory protein 1 gene, involved in nerve cell differentiation and synapse formation), and their detailed sequence characteristics, functional descriptions, and regulatory information.

[0025] Based on the synaptic microstructural feature data, an image analysis algorithm based on machine learning was used to compare the synaptic images in the frontal and parietal observation areas at different time points (image data were collected every week). When calculating the synaptic birth rate, the formula , for example, in the frontal lobe, for excitatory synapses ( Indicates the synaptic type, here we take excitatory synapses as an example), and statistics show that in the 4th week ( Week, set rehabilitation start to ) is 50, the original number of synapses The initial statistics are 200, and the importance weight coefficient of excitatory synapses in the frontal lobe area According to the previous research on the neural function and synaptic importance of this area in patients with brain trauma, the value was set to 0.6 (different types of synapses have different weights in different brain regions, determined based on professional knowledge and research). The synaptic regeneration rate of excitatory synapses in the frontal lobe region in the fourth week was calculated to be 0.15. Similarly, the synaptic regeneration rate indicators of different brain regions and different types of synapses at different time points were calculated to fully reflect the changes in synaptic structural plasticity. For synaptic function parameter data, the synaptic transmission efficiency, postsynaptic potential changes and other data recorded for the first time at the beginning of rehabilitation treatment were used as the baseline. At intervals of every 3 days, the formula was used to calculate the synaptic regeneration rate of excitatory synapses in the frontal lobe region in the fourth week. , the magnitude of the increase in synaptic transmission efficiency, for example in the parietal region ( represents the parietal lobe), in the second week ( By integrating the electrophysiological data, the cumulative change of the potential difference between presynaptic and postsynaptic neurons from the beginning of rehabilitation to the second week was compared with the time point corresponding to the beginning of rehabilitation ( The comparison of the cumulative changes in potential difference (representing the time point of rehabilitation start) calculated that the improvement in synaptic transmission efficiency was 2096. Through such calculations, the dynamic changes in synaptic functional plasticity in different brain regions are continuously tracked to provide a basis for subsequent evaluation.

[0026] The extracted 50 key gene information (after normalization, such as mapping numerical features such as gene expression levels to the [0, 1] interval), the calculated quantitative indicators of synaptic plasticity of different types of synapses in each brain region (the synaptic birth rate, the degree of improvement in synaptic transmission efficiency, etc. are also normalized), as well as the patient's initial disease severity data (such as GCS scores, etc., which are standardized), and basic physical condition data (blood pressure, blood sugar, etc. are normalized) are integrated into the input feature vector.

[0027] A personalized rehabilitation trajectory prediction model based on a deep neural network combined with an attention mechanism was used. The network structure was set to contain three hidden layers, with 128, 64, and 32 neurons in each hidden layer, respectively. During the model training process, similar data from 200 patients with brain trauma rehabilitation in the hospital were collected (including the above-mentioned input feature data and actual rehabilitation trajectory data, divided into training set, validation set, and test set in a ratio of 7:2:1). The model was trained using the training set, and a custom loss function was selected (combining the mean square error and the loss function weighted by the importance of the rehabilitation stage, and the relevant weight parameters and hyperparameters were determined by cross-validation). value), and updated the weight parameters with the Adam optimizer. After about 500 iterative trainings, when the loss value of the model on the validation set converged and the prediction accuracy reached the expected requirements, the test set was used for final evaluation to ensure that the model could reliably predict the expected rehabilitation status indicators at different time points (such as predicting once a week within the next 12 weeks) based on the current patient's input feature data, including the degree of recovery of right limb motor function, improvement of language function (predicted by relevant indicators such as language fluency test), and level of cognitive function recovery (predicted by relevant indicators of the simple mental state examination form), etc., to form a personalized rehabilitation trajectory prediction curve.

[0028] During the patient's rehabilitation process, at weekly intervals, professional rehabilitation therapists used the Fuql-Meyer assessment scale to evaluate the patient's right limb motor function and recorded the corresponding scores. Through language communication, question-and-answer, etc., the patient's language expression fluency, accuracy and other language functions were evaluated and quantified. The patient's cognitive function was tested using a simple mental status checklist to obtain the corresponding evaluation results. At the same time, super-resolution fluorescence microscopy imaging technology and in vivo electrophysiological recording technology were used again to collect the latest neuronal synaptic microstructure characteristics and functional parameter data. The collection process followed the previous technical specifications and requirements. The actual collected rehabilitation status indicator data including limb motor function, language function, cognitive function recovery and the latest synaptic microscopic characteristics and functional parameters were matched one by one with the predicted rehabilitation status indicator data output by the prediction model according to the corresponding time node. The comprehensive rehabilitation difference was calculated (according to the formula ,For example Including limb motor function, language function, cognitive function, the rate of new synapses in each brain region, synaptic transmission efficiency and other multi-dimensional indicators), assuming that the calculated value at the 8th week The value is 0.25 (the comprehensive calculation result of the difference between the actual value and the predicted value of the indicators in different dimensions), and compared with the pre-set evaluation criteria (determined by basic clinical experience and statistical analysis of a large amount of previous data, such as setting Determined that recovery met expectations, It is determined that the patient's rehabilitation progress is slightly delayed, etc.), and it is determined that the patient's rehabilitation progress is slightly delayed at this time.

[0029] According to the difference between the actual rehabilitation performance and the predicted trajectory in the real-time evaluation results of the degree of rehabilitation, the analysis found that the patient's right limb motor function recovery and the synaptic regeneration rate index in the frontal lobe area were relatively different from the prediction, indicating that the rehabilitation progress in these two aspects was not ideal. Therefore, we focused on the evaluation of the limb motor function module and the frontal lobe synaptic plasticity-related indicators, increased the weight of related evaluation indicators such as limb strength and joint range of motion in the Fugl-Meyer assessment scale, and increased the weight of the frontal lobe area synaptic regeneration rate index in the overall evaluation (for example, from the original 0.1 to 0.2, and adjusted reasonably according to the degree of difference and the importance of the indicator). According to the adjusted evaluation indicator weights and the actual rehabilitation progress, the rehabilitation plan was optimized, and the rehabilitation training intensity for the right limb was increased. The degree of rehabilitation training was extended, such as extending the rehabilitation training time, increasing the daily limb movement rehabilitation training time from 1.5 hours to 2 hours, increasing the difficulty of strength training, adjusting the use of rehabilitation equipment with higher resistance for training, and at the same time, adjusting the rehabilitation training items, adding neuromuscular electrical stimulation therapy to assist in promoting the recovery of limb muscle function, focusing on stimulating the relevant muscle groups of the right limb, and adding transcranial direct current stimulation (tDCS) therapy to the rehabilitation plan to improve the synaptic transmission efficiency and neuronal plasticity in the frontal lobe area by regulating the excitability of the cerebral cortex, in the hope of improving the patient's cognitive and language functions, generating a personalized rehabilitation plan that is more in line with the patient's current rehabilitation needs, and continuously tracking the patient's subsequent rehabilitation status so as to make dynamic adjustments again according to the new actual situation.

[0030] Embodiment 2: This embodiment describes in detail a specific application of a neurological rehabilitation degree evaluation method and system in the rehabilitation degree evaluation and plan adjustment of Alzheimer's patients. Through the multi-dimensional and personalized approach of the present invention, accurate evaluation of the patient's rehabilitation degree and scientific optimization of the rehabilitation plan are achieved, which helps to delay disease progression and improve the quality of life of patients, and provides a set of innovative and practical technical means for the rehabilitation treatment practice of Alzheimer's disease.

[0031] A 70-year-old female patient recently developed symptoms such as memory loss, cognitive decline, and disorientation. After detailed examination at the hospital, she was diagnosed with mild stage Alzheimer's disease. She also had a history of hypertension, but her blood pressure was controlled at a relatively stable level through medication. A professional sequencing team was arranged to perform whole genome sequencing for the patient, obtain whole genome data, and ensure that the data quality and accuracy met the requirements of subsequent analysis.

[0032] The neurologist assessed the patient's initial disease severity based on the Mini-Mental State Examination (MMSE). The patient's initial MMSE score was 22 points (indicating mild cognitive impairment, consistent with mild Alzheimer's disease). At the same time, the patient's detailed evaluation results in various cognitive dimensions such as orientation, memory, and attention were recorded as part of the disease severity data. In addition, the disease assessment data was further improved by referring to imaging features such as the atrophy of the brain hippocampus shown in the cranial MRI (Alzheimer's disease often involves the hippocampus, causing structural changes).

[0033] The basic physical condition data of the patients were collected. In addition to recording age and gender information, the focus was on the patients' hypertension, including the duration of illness, the types and dosages of antihypertensive drugs taken, etc. Blood pressure was measured regularly, and the blood pressure fluctuation range was between 130-140 / 80-90 mmHg. Other routine physiological indicators such as blood sugar and blood lipids were recorded at the same time. Among them, blood sugar and blood lipid indicators were slightly higher than the critical value in the normal range. These data were sorted into a standardized basic physical condition data set. The hippocampus area of ​​the patient's brain (closely related to memory and cognitive function) was analyzed using a super-resolution fluorescence microscope combined with a specific labeling method for neuronal synapses. After marking the synapses of neurons in the hippocampus (which is the key area of ​​pathological changes in Alzheimer's disease), imaging observations are performed every two weeks to obtain high-resolution synaptic image data. Through professional image analysis software, the morphological characteristics of the synapses (such as parameters such as synaptic size and shape complexity), number, density and other microstructural information are analyzed and recorded in detail. Each acquisition is accompanied by detailed imaging-related metadata (imaging time, observation area positioning, etc.) for subsequent comparative analysis. In vivo electrophysiological recording technology is used simultaneously to accurately implant microelectrodes in the hippocampus (based on neuroanatomical positioning and with the help of high-precision positioning auxiliary equipment to ensure implantation accuracy ), continuously monitor functional parameters such as synaptic transmission efficiency and postsynaptic potential changes, collect electrophysiological data at weekly intervals, record the patient's status at the time of collection (such as quiet resting state, whether there is external interference, etc.) and accurate timestamps, and store the complete electrophysiological data in an orderly manner for subsequent analysis of synaptic functional plasticity changes. After the whole genome sequencing data is imported into the bioinformatics analysis platform for preprocessing, based on the neurodevelopment and repair mechanisms involved in Alzheimer's disease rehabilitation, such as amyloid protein metabolism, tau protein regulation, synaptic plasticity maintenance, and neuroinflammation regulation, with reference to a large number of clinical studies and basic experimental results, based on Based on factors such as the role of genes in the amyloid precursor protein (APP) metabolic pathway, the degree of correlation with tau protein phosphorylation regulation, and the proven close association with improved cognitive function in Alzheimer's disease, customized screening rules and algorithms are used to screen out key gene information from the whole genome data to construct a data set containing about 30 key genes, including APOE (apolipoprotein E gene, which is closely related to the risk of Alzheimer's disease and the progression of the disease), PSEN1 (presenilin 1 gene, involved in APP processing and amyloid protein production), and other genes and their detailed sequence characteristics, functional annotations and regulatory information.

[0034] Based on the synaptic microstructural characteristic data of the hippocampus region, the image analysis algorithm was used to compare the synaptic images at different time points at a two-week interval. Calculate the synaptic rate, for example, in a certain evaluation (set the recovery start to ,exist Weeks), inhibitory synapses in the hippocampus ( Indicates synapse type), and counts the number of new synapses is 30, the original number of synapses The initial statistics were 150. Considering the importance of inhibitory synapses in regulating neural activity and cognitive function in the hippocampus, the weight coefficient According to professional research, it was set to 0.7, and the synaptic regeneration rate of inhibitory synapses in this area in the 6th week was calculated to be 0.14. Similarly, the synaptic regeneration rates of different types of synapses at different time points in the hippocampus region reflect the association between changes in synaptic structural plasticity and cognitive function repair. For the synaptic function parameter data recorded by electrophysiological recording, the baseline data at the beginning of rehabilitation was used as the basis, and the weekly time intervals were calculated using the formula , calculate the improvement in synaptic transmission efficiency, for example, at week 8 ( Weeks) were used to analyze the cumulative changes in the potential difference between pre- and post-synaptic neurons in the hippocampus, and compared with the time at the beginning of rehabilitation ( ) and calculated that the synaptic transmission efficiency increased by 15%. By continuously tracking the changes in these indicators, the impact of synaptic functional plasticity on cognitive function rehabilitation was evaluated.

[0035] The 30 key gene information extracted by integration (normalization was performed to make the relevant data of each gene in an appropriate range), the calculated quantitative indicators of synaptic plasticity of different types of synapses in the hippocampus region (the synaptic birth rate, the improvement of synaptic transmission efficiency, etc. were also normalized), and the patient's initial disease severity data (such as standardized processing such as MMSE scores) and basic physical condition data (normalized values ​​such as blood pressure and blood sugar) were used as input feature vectors.

[0036] A deep neural network combined with an attention mechanism was used to construct a personalized rehabilitation trajectory prediction model. The network architecture was set to include four hidden layers, with 100, 80, 60, and 40 neurons in each hidden layer, respectively. Relevant data of 150 Alzheimer's disease rehabilitation patients in the hospital were collected. The model was trained using a training set, and a custom loss function was selected (the mean square error and the loss function weighted by the importance of the rehabilitation stage were considered comprehensively, and the parameter values ​​were determined by cross-validation and other methods). The Adam optimizer was used to update the model weight parameters. After about 600 iterations of training, the performance indicators of the model on the validation set (such as loss value convergence and prediction accuracy reaching a stable and required level) met the expectations. The test set was then used for final evaluation to ensure that the model could accurately predict the expected rehabilitation status indicators at different time nodes (for example, monthly prediction within the next 6 months) based on the patient's input feature data, such as the degree of improvement in cognitive function (based on changes in MMSE scores, Montreal Cognitive Assessment scale-related indicators, etc.), changes in daily living ability (predicted by indicators related to the Daily Living Activities Ability Scale), etc., and then generate a personalized rehabilitation trajectory prediction curve.

[0037] During the patient's rehabilitation process, professional medical staff used the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment Scale to test and evaluate the patient's cognitive function at strict monthly intervals, recorded the corresponding scores, and used the Daily Living Activities Scale to evaluate the patient's daily living ability, such as self-care ability, activity ability, etc., and quantified the scores. At the same time, super-resolution fluorescence microscopy imaging technology and in vivo electrophysiological recording technology were used again to collect the latest synaptic microstructural characteristics and functional parameter data in the hippocampus region. The collection operation strictly followed the previously established specifications to ensure the consistency and comparability of the data. The actual collected rehabilitation status indicator data including cognitive function, changes in daily living ability, and the latest synaptic microstructural characteristics and functional parameters were matched and compared one by one with the predicted rehabilitation status indicator data output by the prediction model according to the corresponding time node. The comprehensive rehabilitation difference was calculated (using the formula , here Covering multiple dimensions of cognitive function indicators, daily living ability related indicators, the rate of new synapses in the hippocampus, synaptic transmission efficiency, etc.), assuming that the calculated results at the fourth month The value is 0.3 (reflecting the comprehensive difference between the actual value and the predicted value of each dimension), compared with the evaluation criteria previously formulated based on clinical experience and statistical analysis of a large number of case data (for example, setting Determined that recovery met expectations, It is determined that the patient's rehabilitation progress is slightly delayed, etc.), and it is determined that the patient's rehabilitation progress is slightly delayed at this time.

[0038] According to the analysis of the differences between the actual rehabilitation performance and the predicted trajectory in the real-time evaluation results of the degree of rehabilitation, it was found that the differences between the patient's cognitive function recovery and the synaptic transmission efficiency indicators in the hippocampus region and the predictions were relatively prominent, indicating that the rehabilitation process in these two aspects did not meet expectations. Therefore, we focused on the evaluation of cognitive function modules and synaptic plasticity-related indicators (especially synaptic transmission efficiency), increased the weight of memory and attention-related evaluation indicators in the MMSE score, and increased the weight of indicators related to cognitive dimensions such as executive function and language fluency in the Montreal Cognitive Assessment Scale. In addition, we increased the weight of the synaptic transmission efficiency indicator in the hippocampus region in the overall evaluation (for example, from the original 0.12 to 0.2), so that it can better reflect the impact on the progress of rehabilitation.

[0039] According to the adjusted evaluation index weights and the actual progress of rehabilitation, the rehabilitation plan is optimized and adjusted to increase the intensity and frequency of cognitive function training. For example, the duration of special memory training is increased from 30 minutes to 45 minutes every day, and more challenging cognitive stimulation games and training tasks are introduced, such as complex graphic memory, logical reasoning puzzles, etc., to enhance the patient's cognitive ability. At the same time, transcranial magnetic stimulation (TMS) therapy is added to the rehabilitation plan to improve the synaptic function of the hippocampus. By regulating the excitability of the cerebral cortex, it promotes the improvement of synaptic transmission efficiency and neuronal plasticity in the hippocampus region, hoping to better improve the patient's cognitive function. In addition, in view of the patient's history of hypertension and the critically high blood sugar and blood lipids, the dietary guidance is further optimized, the monitoring of blood pressure, blood sugar, and blood lipids is strengthened, and the use of antihypertensive and lipid-lowering drugs is adjusted when necessary to ensure that the overall physical condition is conducive to rehabilitation, generate a personalized rehabilitation plan that is more in line with the patient's current rehabilitation needs, and continue to track the subsequent rehabilitation situation, so as to dynamically adjust the rehabilitation strategy again according to new actual changes and help the patient's rehabilitation to the greatest extent.

[0040] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A method for evaluating the degree of rehabilitation in neurology, characterized in that: The method comprises the following specific steps: Data collection: Whole genome sequencing is performed on patients in the Department of Neurology to obtain the whole genome data of the patients. The initial severity of the disease is collected and the symptoms of the patients at the time of the first visit are evaluated by professional neurologists according to the established disease assessment criteria. At the same time, the basic physical condition data of the patients are collected. In addition, the morphology, number and density of brain neuron synapses are regularly observed and recorded with the help of super-high-resolution microscopy imaging technology combined with nerve cell labeling methods. In combination with in vivo electrophysiological recording technology, the dynamic changes of synaptic transmission efficiency and postsynaptic potential change functional parameters during the rehabilitation process are monitored; Extraction of key gene information and establishment of synaptic plasticity indicators: Import the data obtained by whole genome sequencing into the bioinformatics analysis software platform, perform quality screening, sequence alignment and annotation operations on the original gene sequence data, and use bioinformatics analysis and medical knowledge integration strategies based on the neural development and repair mechanisms unique to the rehabilitation of neurological diseases to deeply explore the gene information closely related to neural development and repair. The microstructural feature data of neuronal synapses collected by super-high-resolution microscopy imaging technology, as well as the synaptic transmission efficiency and postsynaptic potential change parameter data obtained by in vivo electrophysiological recording technology, are used to reflect the degree of connection reconstruction between nerve cells through quantitative calculation. For the synaptic microstructural feature data, image analysis algorithms are used to compare and analyze the image data collected at different time points to calculate the synaptic newborn rate. For the synaptic function parameter data, the data first recorded at the beginning of rehabilitation treatment is used as the baseline. At fixed time intervals, the parameter changes at different time nodes are compared and analyzed, and the quantitative indicators reflecting the changes in synaptic function plasticity, such as the improvement in synaptic transmission efficiency, are obtained through algorithms. Construction of a personalized rehabilitation trajectory prediction model: The extracted key gene information, the established synaptic plasticity quantitative index, and the collected patient's initial disease severity data and basic physical condition data are used as input features, and an artificial intelligence algorithm is used to construct a personalized rehabilitation trajectory prediction model. Through training and learning on a large number of existing neurological rehabilitation case data, the model can predict the patient's expected rehabilitation status indicators at different time points based on the input new patient feature data, forming a personalized rehabilitation trajectory prediction curve; Real-time evaluation of rehabilitation degree: During the patient's rehabilitation process, the patient's current actual rehabilitation status indicators are collected at predetermined time intervals, and the actual rehabilitation status indicators are compared and analyzed with the predicted rehabilitation status indicators of the corresponding time nodes output by the prediction model, and the difference between the two is calculated. The evaluation result of the current patient's rehabilitation degree is determined based on the evaluation criteria; Adjustment of evaluation index weights and optimization of rehabilitation plans: Based on the difference between the actual rehabilitation performance and the predicted trajectory in the real-time evaluation results, the weights of each indicator in the subsequent rehabilitation evaluation are dynamically adjusted. At the same time, based on the adjusted evaluation index weights and the actual rehabilitation progress, the subsequent rehabilitation plan is optimized, and the content, intensity and frequency of rehabilitation training are adjusted to generate a personalized rehabilitation plan that is more in line with the individual patient's current rehabilitation needs.

2. A neurological rehabilitation degree evaluation method according to claim 1, characterized in that: In the steps of extracting key gene information and establishing synaptic plasticity indicators, based on the unique neural development and repair mechanisms of neurological disease rehabilitation, key gene information closely related to neural development and repair is deeply excavated. Specifically, with reference to a large number of authoritative clinical research results, basic medical experimental data and professional medical knowledge bases in the field of neurology, screening rules and algorithms are set according to the role of genes in neural development and repair links including nerve cell differentiation, neurotransmitter regulation, neuron regeneration and synaptic plasticity regulation, key gene information is screened out from the pre-processed whole genome data, and gene sequence characteristics, functional descriptions and related regulatory information are integrated to construct a disease-specific key gene data set that is exclusive to individual patients.

3. A neurological rehabilitation degree evaluation method according to claim 2, characterized in that: In the steps of extracting key gene information and establishing synaptic plasticity index, screening rules and algorithms are set to screen out key gene information from the pre-processed whole genome data. The algorithm formula is: ,in, Indicates The comprehensive association score of each gene, It is The weight coefficient of the screening factor, Representative The gene in The correlation degree value under the screening factors, represents the total number of screening factors, The final selected key gene set meets its comprehensive association score Greater than or equal to the preset screening threshold genes were included in this set.

4. A neurological rehabilitation degree evaluation method according to claim 1, characterized in that: In the steps of extracting key gene information and establishing synaptic plasticity indicators, an image analysis algorithm is used to compare and analyze image data collected at different time points to calculate the synaptic newborn rate, and the calculation formula is: ,in, Indicated in brain area In time The synaptic birth rate, Represented in brain area within, no. Types of synapses at time The number of newly formed synapses, image data collected by super-resolution microscopy imaging technology, It is Types of synapses in brain regions The importance weight coefficient of Indicated in brain area within, no. The original number of synapses of the synapse type, In this brain region The total number of synapse types classified within the region covers all synapse types that are important for neural function in this area and whose new formation needs to be paid attention to.

5. A neurological rehabilitation degree evaluation method according to claim 1, characterized in that: In the steps of extracting key gene information and establishing synaptic plasticity indicators, quantitative indicators such as the improvement of synaptic transmission efficiency that reflect the changes in synaptic functional plasticity are obtained through an algorithm, and the algorithm formula is: ,in, Indicated in brain area In time The improvement in synaptic transmission efficiency In the brain area The postsynaptic neuron The electrical activity potential value, The time from the start of rehabilitation to observation Any time variable in the range, Represented in brain area The presynaptic neuron is The electrical activity potential value, represents the initial time point when rehabilitation treatment begins, which is used as the baseline time for calculating the improvement in synaptic transmission efficiency. represents the initial state of synaptic transmission efficiency at the beginning of recovery, is the specific point in the rehabilitation process at which the assessment is currently being made.

6. A neurological rehabilitation degree evaluation method according to claim 1, characterized in that: In the step of constructing the personalized rehabilitation trajectory prediction model, an artificial intelligence algorithm is used to construct a personalized rehabilitation trajectory prediction model. By training and learning a large amount of existing neurological rehabilitation case data, specifically, the input feature vector is set to ,in , which correspond to the feature vectors of key gene information, synaptic plasticity quantitative indicators, initial disease severity data, and basic physical condition data, respectively. The hidden layer of the network is represented as , the rehabilitation status indicator vector predicted by the output layer is , for the calculation from the input layer to the first hidden layer ,in, is the activation function, is the corresponding input feature in the first hidden layer and The weight parameter of the interaction, Represents a custom feature fusion operation, is the bias term of the first hidden layer. Hidden layers to The propagation calculation formula of the hidden layer is: ,in, It is The hidden layer corresponds to the previous layer The connection weights of the outputs of neurons, Indicates The hidden layer The output value of a neuron is the result of the previous layer processed by the activation function and serves as the input of the current layer. It is The bias term of the hidden layer is added after the hidden layer. The attention mechanism layer is used to dynamically assign different attention weights to different input features to highlight the feature information that is more critical to the current rehabilitation trajectory prediction. The formula is: ,in, Represents the calculated unnormalized attention score, which is used to measure the importance of each feature dimension in the current rehabilitation trajectory prediction task. Its calculation involves the weight matrix and Output the hidden layer separately and the original input features Perform a linear transformation and add a bias After Activation function processing, is the normalized attention weight, is the input feature vector after attention weighting. After the attention mechanism layer, the weighted feature vector is used The final recovery status index is predicted using the following calculation formula: ,in, is the weight matrix of the output layer, is the bias term of the output layer. In order to train the above network model, the loss function is used to measure the predicted rehabilitation status index Actual recovery status indicators The difference between them is calculated as: ,in, is the total number of predicted time nodes, is the weight coefficient corresponding to each time node, and At the time nodes The predicted rehabilitation status index value and the actual rehabilitation status index value, represents the number of different stages into which the rehabilitation process is divided, It is The number of time nodes included in each rehabilitation stage, and In the Rehabilitation stage time points The predicted and actual rehabilitation status index values, is the variance of the actual rehabilitation status index value calculated at this rehabilitation stage, It is a hyperparameter that balances the two losses. During the training process, an optimization algorithm is used to minimize the above loss function. The weight parameters of the model are updated and adjusted according to the loss function. After multiple iterative training, the performance indicators of the model on the validation set meet the requirements, and a trained personalized rehabilitation trajectory prediction model is obtained.

7. A neurological rehabilitation degree evaluation method according to claim 1, characterized in that: In the step of real-time evaluation of the degree of rehabilitation, during the patient's rehabilitation process, the patient's current actual rehabilitation status indicators are collected at predetermined time intervals, including but not limited to the recovery of specific neural functions, changes in related physical function indicators, and the latest neuronal synaptic microstructural characteristics and functional parameter data.

8. A neurological rehabilitation degree evaluation method according to claim 1, characterized in that: In the step of real-time evaluation of rehabilitation degree, for limb motor function test, the operation is carried out according to standardized action tasks and scoring rules. The sensory function test adopts standard sensory threshold detection means, that is, by using professional equipment such as tactile tester and pain tester, the skin is stimulated step by step from low to high stimulation intensity, and the minimum intensity value that the patient can perceive is recorded as the sensory threshold for quantitative evaluation. The cognitive function evaluation uses a professional scale with verified reliability and validity to ensure that the operation of each neurological function test process is standardized and the results are reliable.

9. A neurological rehabilitation degree evaluation method according to claim 1, characterized in that: In the step of real-time evaluation of the degree of rehabilitation, the actual rehabilitation status index is compared and analyzed with the predicted rehabilitation status index of the corresponding time node output by the prediction model, and the difference between the two is calculated. The calculation formula is: ,in, Refers to the comprehensive rehabilitation difference, It is in The current rehabilitation status indicator value actually collected in the rehabilitation status indicator dimension, Indicates that the prediction model is in the corresponding The predicted value output on the rehabilitation status indicator dimension, The total number of rehabilitation status indicator dimensions included in the comprehensive consideration.

10. A neurology rehabilitation degree evaluation system, which is applicable to a neurology rehabilitation degree evaluation method according to any one of claims 1 to 9, characterized in that: The system comprises the following components: a data collection module, a key gene and index processing module, a prediction model building module, a rehabilitation evaluation module and an adjustment and optimization module; The data acquisition module is used to perform whole genome sequencing on neurology patients, obtain whole genome data, and collect the patient's initial disease severity data and basic physical condition data. At the same time, it uses ultra-high resolution microscope imaging technology and in vivo electrophysiological recording technology to collect brain neuron synaptic microstructural characteristics and functional parameter data; The key gene and index processing module uses bioinformatics analysis tools to extract key gene information related to neural development and repair from whole genome data, and analyzes and processes the collected synaptic microscopic data to establish quantitative indicators related to synaptic plasticity; The prediction model building module uses key gene information, synaptic plasticity quantitative indicators, initial disease severity data, and basic physical condition data as inputs, and uses artificial intelligence algorithms to build and train a personalized recovery trajectory prediction model; The rehabilitation evaluation module compares and analyzes the actual rehabilitation status index with the predicted rehabilitation status index output by the prediction model to obtain a rehabilitation degree evaluation result; The adjustment and optimization module dynamically adjusts the evaluation index weights and optimizes the rehabilitation plan according to the actual and predicted differences in the rehabilitation degree evaluation results, thereby generating a new personalized rehabilitation plan.

Images