Cerebral stroke layered rehabilitation quantification method and device

Abstract

The invention relates to a cerebral apoplexy layered rehabilitation quantification method and device, and belongs to the technical field of rehabilitation quantification, historical data of a cerebral apoplexy patient is collected, and the historical data of the cerebral apoplexy patient is input into a cerebral apoplexy intelligent layering engine for layering to determine a layering label; multi-modal data and a basic weight distribution table of a stroke patient are obtained, and a dynamic weight distributor adjusts the weight of the multi-modal data of the patient according to the condition of the multi-modal data of the patient and weight adjustment logic; determining a hierarchical correction coefficient according to the hierarchical label, determining a baseline severity coefficient according to the NIHSS baseline score, and determining a standardization score of the multi-modal data; and determining a rehabilitation progress index according to the adjusted weight of the multi-modal data of the patient, the standardized score, the hierarchical correction coefficient and the baseline severity coefficient. The cerebral apoplexy rehabilitation quantification method provided by the invention has the advantages of medical precise layering, dynamic anti-interference weight distribution and verifiable expansion, and solves the problem that an evaluation result is disjointed from clinical practice.

Description

Technical Field

[0001] This invention belongs to the field of quantitative technology of stroke rehabilitation, and particularly relates to a method and device for quantitative stratified rehabilitation of stroke. Background Technology

[0002] Problems with existing quantitative techniques for stroke rehabilitation:

[0003] The stratification is crude: it only classifies diseases according to their course (acute / subacute / chronic) without considering differences in lesion localization (such as the different rehabilitation pathways for cortical and brainstem types).

[0004] Rigid data utilization: Using fixed weights to fuse multimodal data, when a certain type of data is abnormal (such as motion sensor failure), it is still calculated according to the original weight, resulting in evaluation distortion.

[0005] Insufficient scalability: Although it supports the access of new data, it directly assigns fixed weights (such as newly added eye-tracking data always accounting for 15%), without verifying its medical relevance. Summary of the Invention

[0006] In view of the shortcomings of the prior art, the purpose of the invention is to provide a method and device for quantifying stroke rehabilitation by stratification, to realize a method for precise medical stratification, dynamic anti-interference weight allocation, and verifiable and expandable quantification of stroke rehabilitation, and to solve the problem of the disconnect between assessment results and clinical practice.

[0007] In a first aspect, the present invention provides a quantitative method for stratified rehabilitation of stroke patients, comprising:

[0008] Historical data of stroke patients are collected and input into the stroke intelligent stratification engine to determine stratification labels.

[0009] The system acquires multimodal data and a basic weight allocation table of stroke patients. The dynamic weight allocator adjusts the weights of the multimodal data of the patients based on the patient's multimodal data and the weight adjustment logic.

[0010] The stratification correction coefficient is determined based on the stratification label, the baseline severity coefficient is determined based on the NIHSS baseline score, and the standardized score of the multimodal data is determined.

[0011] The rehabilitation progress index was determined based on the adjusted weights of the patient's multimodal data, the standardized scores of the multimodal data, the stratification correction coefficient, and the baseline severity coefficient.

[0012] The dimensions of the content in the layer include: disease stage, lesion location, and functional impairment level. The subcategories of disease stage include: acute phase, subacute phase, and chronic phase. The subcategories of lesion location include: cortical type, brainstem type, and cerebellar type. The subcategories of functional impairment level include: mild, moderate, and severe.

[0013] Multimodal data includes: motion data, physiological characteristics, brain-computer interfaces, and demographic / medical history.

[0014] Furthermore, in the aforementioned quantitative method for stratified rehabilitation of stroke, the weight adjustment logic includes:

[0015] Data credibility compensation, recovery phase transition, and outlier penalty;

[0016] The data reliability includes: if the missing rate of one of the modal data in the multimodal data is greater than the preset missing rate, its weight is proportionally transferred to the preset modal data;

[0017] The stage transition includes a weighting scheme that automatically switches to the next disease stage when the weekly growth rate of the Fugl-Meyer rating scale score is greater than or equal to the preset growth rate.

[0018] The outlier penalty includes: if an anomaly is detected in one of the modal data in the multimodal data, temporarily reducing the weight of the preset modal data to a preset percentage.

[0019] Furthermore, in the aforementioned method for quantifying stroke stratified rehabilitation, the stratification correction coefficient is determined based on the stratification label, and the baseline severity coefficient is determined based on the NIHSS baseline score, including:

[0020] In the hierarchical labeling, the hierarchical correction factor is 1.0 for the cortical type, 0.8 for the brainstem type, and 0.65 for the cerebellar type.

[0021] The baseline severity coefficient β, determined based on the NIHSS baseline score, is expressed by the following formula:

[0022] β = 1 + NIHSS baseline score / NIHSS baseline maximum score

[0023] The NIHSS baseline maximum score is 42.

[0024] Furthermore, in the above-mentioned quantitative method for stratified rehabilitation of stroke, the formula for determining the standardized score of multimodal data is as follows:

[0025] The standardized score of motion data is determined by gait symmetry, using the following formula:

[0026]

[0027] The standardized score of physiological characteristics is determined by blood oxygen stability, using the following formula:

[0028]

[0029] The standardized score of a brain-computer interface is determined by the motor intent recognition rate, using the following formula:

[0030]

[0031] Among them, a blood oxygen fluctuation range of >20% will receive 0 points.

[0032] Furthermore, in the aforementioned method for quantifying stratified rehabilitation for stroke, the rehabilitation progress index is determined using the following formula, based on the adjusted weights of the patient's multimodal data, the standardized scores of the multimodal data, the stratification correction coefficient, and the baseline severity coefficient:

[0033]

[0034] Among them, W i N represents the adjusted weights of the i-th modality data. i denoted as the standardized score of the i-th data category, K represents the stratification correction coefficient, β represents the baseline severity coefficient, and RehabIndex represents the recovery progress index.

[0035] Furthermore, in the aforementioned method for quantifying stratified rehabilitation for stroke, if the missing rate of one modality in the multimodal data is greater than a preset missing rate, its weight is proportionally transferred to the preset modality, including:

[0036] If the missing rate of motion data in multimodal data is greater than the preset missing rate, the weight of motion data is transferred to physiological features;

[0037] If the missing rate of physiological features in the multimodal data is greater than the preset missing rate, the weights of the physiological features are transferred to the motion data;

[0038] If the missing rate of brain-computer interfaces in the multimodal data is greater than the preset missing rate, the weight of brain-computer interfaces is transferred to the motion data;

[0039] If the demographic / medical history missing rate in the multimodal data is greater than the preset missing rate, the demographic / medical history weights are transferred to the motion data.

[0040] Furthermore, the aforementioned quantitative method for stratified rehabilitation of stroke also includes:

[0041] After receiving new modal data, the medical relevance extension gateway performs cross-modal correlation analysis with the existing multimodal data;

[0042] If the cross-modal correlation analysis shows that the Pearson correlation coefficient is greater than the preset coefficient, then the corresponding new modal data will be retained.

[0043] The importance score of new modality data for predicting rehabilitation progress was calculated using a random forest model;

[0044] The new modal data includes: biochemical indicators, medical imaging data, microenvironment and nutrition data, biomechanical data, and patient-reported outcomes.

[0045] A second aspect of the present invention also provides a stratified rehabilitation quantification device for stroke, comprising:

[0046] The data acquisition module is used to collect historical data of stroke patients and input the historical data of stroke patients into the stroke intelligent stratification engine to determine the stratification labels.

[0047] Adjustment module: Used to acquire multimodal data and basic weight allocation table of stroke patients. The dynamic weight allocator adjusts the weights of the patient's multimodal data according to the patient's multimodal data and weight adjustment logic.

[0048] The first determination module is used to determine the stratification correction coefficient based on the stratification label, the baseline severity coefficient based on the NIHSS baseline score, and the standardized score of the multimodal data.

[0049] The second determination module is used to determine the rehabilitation progress index based on the adjusted weights of the patient's multimodal data, the standardized scores of the multimodal data, the stratification correction coefficient, and the baseline severity coefficient.

[0050] The dimensions of the content in the layer include: disease stage, lesion localization, and functional impairment level. The subcategories of disease stage include: acute phase, subacute phase, and chronic phase. The subcategories of lesion localization include: cortical type, brainstem type, and cerebellar type. The subcategories of functional impairment level include: mild, moderate, and severe. The multimodal data includes: motor data, physiological characteristics, brain-computer interface, and demographic / medical history.

[0051] A third aspect of the present invention also provides an electronic device comprising: a processor and a memory;

[0052] The processor executes a stroke stratification rehabilitation quantification method by calling programs or instructions stored in memory, as described above.

[0053] In a fourth aspect, the present invention also provides a computer-readable storage medium that stores a program or instructions that cause a computer to execute a stroke stratification rehabilitation quantification method as described above.

[0054] The beneficial effects of this invention are as follows: This invention collects historical data from stroke patients and inputs this data into a stroke intelligent stratification engine to determine stratification labels; it acquires multimodal data and a basic weight allocation table for stroke patients, and a dynamic weight allocator adjusts the weights of the multimodal data based on the patient's multimodal data and weight adjustment logic; it determines stratification correction coefficients based on stratification labels, baseline severity coefficients based on NIHSS baseline scores, and standardized scores of multimodal data; and it determines a rehabilitation progress index based on the adjusted weights of the patient's multimodal data, standardized scores, stratification correction coefficients, and baseline severity coefficients. This invention provides a precise medical stratification method for stroke rehabilitation that integrates lesion localization, disease stage, and functional impairment, utilizes a dynamic weight allocator for dynamic anti-interference weight allocation, and employs a medical relevance extension gateway to verify and extend new modal data, thus solving the problem of the disconnect between assessment results and clinical reality. Attached Figure Description

[0055] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts. It is obvious that the drawings described below are merely some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings.

[0056] Figure 1 A Quantitative Method for Stratified Rehabilitation of Stroke Provided by Embodiments of the Invention Figure 1 ;

[0057] Figure 2 A Quantitative Method for Stratified Rehabilitation of Stroke Provided by Embodiments of the Invention Figure 2 ;

[0058] Figure 3 A diagram of a stratified rehabilitation and quantitative device for stroke provided in an embodiment of the present invention;

[0059] Figure 4 This is a schematic block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0060] To enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. It should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0061] Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concepts disclosed in this invention.

[0062] In the description of this invention, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance. The terms "installed," "connected," and "linked" should be interpreted broadly; for example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0063] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of methods and systems consistent with some aspects of the invention as detailed in the appended claims.

[0064] This invention proposes a method, device, electronic device, and storage medium for the quantitative assessment of stroke stratified rehabilitation. This invention provides a precise medical stratification method for stroke rehabilitation with dynamic anti-interference weight allocation and verifiable and scalable features, solving the problem of the disconnect between assessment results and clinical practice.

[0065] Before introducing the embodiments of the present invention, the technical terms involved in the present invention will be introduced first.

[0066] The NIHSS score (National Institutes of Health Stroke Scale) is an internationally recognized tool for quantitatively assessing neurological deficits in the field of stroke. Its core value lies in objectively, rapidly, and in a standardized manner assessing the severity of stroke.

[0067]

[0068] Grading criteria (adopted in this invention): Mild: ≤5 points (able to take care of oneself) Moderate: 6-14 points (requires assisted rehabilitation) Severe: ≥15 points (bedridden / disordered consciousness).

[0069] The Fugl-Meyer Assessment Scale (FMA) is an authoritative assessment tool for motor function in the field of stroke rehabilitation, and it is particularly adept at quantifying subtle changes in limb motor recovery.

[0070] Evaluation modules and weights

[0071]

[0072] Method Implementation Examples

[0073] Figure 1 A Quantitative Method for Stratified Rehabilitation of Stroke Provided by Embodiments of the Invention Figure 1 .

[0074] In a first aspect, this invention proposes a quantitative method for stratified rehabilitation of stroke patients, combined with... Figure 1 It includes four steps, S1 to S4:

[0075] S1: Collect historical data of stroke patients and input the historical data of stroke patients into the stroke intelligent stratification engine to determine the stratification label.

[0076] Specifically, in this embodiment of the invention, the sources of the patient's historical data include: electronic medical records, admission records, rehabilitation assessment scales, historical sensor data, family follow-up data, ADL questionnaires, MRI / CT images, fNIRS brain region activation maps, gait analysis, swallowing imaging videos, inertial sensor (IMU) trajectory data, cognitive test scores, task completion time, motor segmentation evaluation, number of incorrect movements, EEG awareness index, and pain expression recognition; the dimensions of the content in the layer include: disease stage, lesion localization, and functional impairment level. The subcategories of disease stage include: acute phase, subacute phase, and chronic phase; the subcategories of lesion localization include: cortical type, brainstem type, and cerebellar type; and the subcategories of functional impairment level include: mild, moderate, and severe; for example, the layer label is "cortical-subacute-moderate" group.

[0077] Historical data of stroke patients are input into a stroke intelligent stratification engine for stratification. The stratification rules and medical basis for determining stratification labels are as follows: Based on electronic medical records and admission records, it is determined whether the disease stage is acute. The clinical significance of the acute phase is: active period of neuroedema, focusing on vital sign monitoring. Based on rehabilitation assessment scales and historical sensor data, it is determined whether the disease stage is subacute. The clinical significance of the subacute phase is: golden period of neuroplasticity, strengthening motor function training. Based on family follow-up data and ADL questionnaires, it is determined whether the disease stage is chronic. The clinical significance of the chronic phase is: focusing on optimizing compensatory strategies and rebuilding daily living abilities. Based on MRI / CT images and fNIRS brain region activation maps, it is determined whether the lesion localization is cortical. The clinical significance of cortical lesions is: Significant upper limb fine motor impairment; gait analysis and swallowing angiography video were used to determine whether the lesion was brainstem type, the clinical significance of which is: balance dysfunction and dysphagia; inertial measurement unit (IMU) trajectory data were used to determine whether the lesion was cerebellar type, the clinical significance of which is: ataxia and abnormal postural control; cognitive test scores and task completion time were used to determine whether the functional impairment level was mild, the clinical significance of which is: complex task training is possible; motor segmentation evaluation and the number of erroneous movements were used to determine whether the functional impairment level was moderate, the clinical significance of which is: decomposed movement training is required; EEG awareness index and pain expression recognition were used to determine whether the functional impairment level was severe, the clinical significance of which is: passive movement and arousal are the main treatments.

[0078] S2: Obtain multimodal data and basic weight allocation table of stroke patients. The dynamic weight allocator adjusts the weights of the patient's multimodal data according to the patient's multimodal data and weight adjustment logic.

[0079] Specifically, in this embodiment of the invention, multimodal data includes: motion data, physiological characteristics, brain-computer interface, and demographic / medical history.

[0080] The basic weight allocation table is shown below:

[0081]

[0082] The method by which the dynamic weight allocator adjusts the weights of a patient's multimodal data based on the patient's multimodal data and weight adjustment logic is described in detail below.

[0083] S3: Determine the stratification correction coefficient based on the stratification label, determine the baseline severity coefficient based on the NIHSS baseline score, and determine the standardized score of the multimodal data.

[0084] Specifically, in this embodiment of the invention, the method for determining the stratification correction coefficient based on the stratification label, the baseline severity coefficient based on the NIHSS baseline score, and the standardized score of the multimodal data is described in detail below.

[0085] S4: Determine the rehabilitation progress index based on the adjusted weights of the patient's multimodal data, the standardized scores of the multimodal data, the stratification correction coefficient, and the baseline severity coefficient.

[0086] Specifically, in this embodiment of the invention, the method for determining the rehabilitation progress index based on the adjusted weights of the patient's multimodal data, the standardized score of the multimodal data, the stratification correction coefficient, and the baseline severity coefficient is described in detail below.

[0087] Furthermore, in the aforementioned quantitative method for stratified rehabilitation of stroke, the weight adjustment logic includes:

[0088] Data credibility compensation, recovery phase transition, and outlier penalty;

[0089] Data reliability includes: if the missing rate of one modality in the multimodal data is greater than the preset missing rate, its weight is proportionally transferred to the preset modality.

[0090] Specifically, in this embodiment of the invention, if the missing rate of a certain modality data is greater than 20%, its weight is proportionally transferred to the preset modality data that has the highest correlation with the modality data. For example, if the missing rate of motion data is greater than 20%, the weight of the brain-computer interface is increased by 0.1.

[0091] The stage transition includes a weighting scheme that automatically switches to the next disease stage when the weekly growth rate of the Fugl-Meyer rating scale score is greater than or equal to the preset growth rate.

[0092] Specifically, in this embodiment of the invention, when the weekly growth rate of the Fugl-Meyer score is ≥10%, the weighting scheme for the next stage of the disease is automatically switched, such as switching from the weighting scheme for the subacute stage to the weighting scheme for the chronic stage.

[0093] Outlier penalties include: if an anomaly is detected in one of the modal data in the multimodal data, the weight of the preset modal data is temporarily reduced to a preset percentage.

[0094] Specifically, in this embodiment of the invention, if abnormal physiological data is detected, such as blood oxygen <90% for 1 minute, the weight of the exercise data is temporarily reduced by 50%.

[0095] Furthermore, in the aforementioned method for quantifying stroke stratified rehabilitation, the stratification correction coefficient is determined based on the stratification label, and the baseline severity coefficient is determined based on the NIHSS baseline score, including:

[0096] In the hierarchical labeling, the hierarchical correction factor is 1.0 for the cortical type, 0.8 for the brainstem type, and 0.65 for the cerebellar type.

[0097] Specifically, in this embodiment of the invention, the principle of the baseline severity coefficient is as follows:

[0098] Cortical type – hemiplegia, muscle weakness, sensory disturbance – recovery curve is relatively standard and serves as the benchmark for calculation – 1.0 – benchmark value, without discount.

[0099] Brainstem type – balance disorder (vestibular system), cranial nerve symptoms, bilateral symptoms – due to poor balance foundation, all training progress involving standing and walking needs to be slowed down – 0.8 – progress is discounted.

[0100] Cerebellar type – ataxia, balance disorder, poor coordination, hypotonia – fundamental damage to motor control mechanisms, low task completion efficiency, and the pace needs to be significantly slowed down – 0.65 – the pace is discounted.

[0101] The baseline severity coefficient β, determined based on the NIHSS baseline score, is expressed by the following formula:

[0102] β = 1 + NIHSS baseline score / NIHSS baseline maximum score

[0103] The NIHSS baseline maximum score is 42.

[0104] For example, for patients with a baseline NIHSS score of 20, β = 1 + 20 / 42 ≈ 1.48, which can prevent the progression of critically ill patients from being overestimated.

[0105] Furthermore, in the above-mentioned quantitative method for stratified rehabilitation of stroke, the formula for determining the standardized score of multimodal data is as follows:

[0106] The standardized score of motion data is determined by gait symmetry, using the following formula:

[0107]

[0108] Specifically, in this embodiment of the invention, the smaller the gait symmetry value, the better; the reciprocal value should be taken.

[0109] The standardized score of physiological characteristics is determined by blood oxygen stability, using the following formula:

[0110]

[0111] Among them, a blood oxygen fluctuation range >20% will receive 0 points;

[0112] The standardized score of a brain-computer interface is determined by the motor intent recognition rate, using the following formula:

[0113]

[0114] Furthermore, in the aforementioned method for quantifying stratified rehabilitation for stroke, the rehabilitation progress index is determined using the following formula, based on the adjusted weights of the patient's multimodal data, the standardized scores of the multimodal data, the stratification correction coefficient, and the baseline severity coefficient:

[0115]

[0116] Among them, W i N represents the adjusted weights of the i-th modality data. i denoted as the standardized score of the i-th data category, K represents the stratification correction coefficient, β represents the baseline severity coefficient, and RehabIndex represents the recovery progress index.

[0117] Example: Patient A (cortical-subacute-moderate group):

[0118] The score for the motion data is N1 = 0.7 (weight W1 = 0.6).

[0119] Brain-computer interface score N² = 0.8 (weight W² = 0.2)

[0120] Physiological data score N3 = 0.9 (weight W3 = 0.2)

[0121] With a stratification coefficient K = 1.0 and an NIHSS baseline score of 12, the baseline severity coefficient was 1.29.

[0122] Rehabilitation progress index = [(0.6×0.7+0.2×0.8+0.2×0.9) / (1.0×1.29)]×100% = 58.1%.

[0123] Furthermore, in the aforementioned method for quantifying stratified rehabilitation for stroke, if the missing rate of one modality in the multimodal data is greater than a preset missing rate, its weight is proportionally transferred to the preset modality, including:

[0124] If the missing rate of motion data in multimodal data is greater than the preset missing rate, the weight of motion data is transferred to physiological features;

[0125] If the missing rate of physiological features in the multimodal data is greater than the preset missing rate, the weights of the physiological features are transferred to the motion data;

[0126] If the missing rate of brain-computer interfaces in the multimodal data is greater than the preset missing rate, the weight of brain-computer interfaces is transferred to the motion data;

[0127] If the demographic / medical history missing rate in the multimodal data is greater than the preset missing rate, the demographic / medical history weights are transferred to the motion data.

[0128] For example, if the missing rate of physiological features in multimodal data is greater than 20%, 0.3 of the physiological feature weight is transferred to the motion data, the physiological feature weight is reduced by 0.3 to become 0, and the weight of the motion data is increased by 0.3 on the original basis.

[0129] Figure 2 A Quantitative Method for Stratified Rehabilitation of Stroke Provided by Embodiments of the Invention Figure 2 .

[0130] Furthermore, the above-mentioned quantitative method for stratified rehabilitation of stroke patients, combined with... Figure 2 It also includes three steps, S21 to S23:

[0131] S21: After receiving new modal data, the medical relevance extension gateway performs cross-modal correlation analysis with the existing multimodal data.

[0132] S22: If the cross-modal correlation analysis shows that the Pearson correlation coefficient is greater than the preset coefficient, then the corresponding new modal data shall be retained.

[0133] S23: Calculate the importance score of new modal data for predicting rehabilitation progress using a random forest model;

[0134] The new modal data includes: biochemical indicators, medical imaging data, microenvironment and nutrition data, biomechanical data, and patient-reported outcomes.

[0135] Specifically, in this embodiment of the invention, biochemical indicators include: such as the level of BDNF (brain-derived neurotrophic factor) in the blood, which directly reflects the strength of neural plasticity and is highly correlated with rehabilitation potential. Other indicators include inflammatory markers (such as CRP), blood glucose / glycated hemoglobin, etc. Medical imaging data includes: such as the integrity of the corticospinal tract shown by diffusion tensor imaging (DTI), or brain activation patterns shown by functional magnetic resonance imaging (fMRI). These data can visually demonstrate the location of brain injury and brain remodeling. Microenvironment and nutritional data include: such as sleep quality data monitored through wearable devices, or nutritional status data obtained through questionnaires and tests. These factors significantly affect the energy and physical strength required for rehabilitation. Biomechanical data includes: data such as joint torques and muscle synergistic activation patterns obtained through a high-precision motion capture system, representing deep motor control information. Patient-reported outcomes (PROs) include: patient-submitted scale data on pain, fatigue, and quality of life, which are important supplements to functional assessment.

[0136] In some embodiments, after receiving new modality data, the medical relevance extension gateway performs the following three-step judgment to achieve relevance analysis, and allocates weights through a dynamic weight allocator:

[0137] First, verify the reliability of the new modal data and the original modal data to determine whether the data is within the physiologically possible range.

[0138] Second: Clinical relevance filtering is performed on new and existing modal data to determine whether the data is important in the current clinical context. For example, in the acute phase, a slight increase in blood pressure may be more important than the number of steps taken in a day. The gateway will assign different relevance weights to the data based on patient type (brainstem type, cortical type, etc.) and recovery stage.

[0139] Third: Security and priority arbitration are performed on new modal data and existing modal data. The medical relevance extension gateway continuously monitors physiological characteristic data. Once an abnormality that may endanger patient safety is detected (such as a sharp drop in blood oxygen saturation or severe arrhythmia), it will immediately issue the highest priority instruction, requiring the dynamic weight allocator to significantly reduce or even suspend the weight of exercise training and focus all resources on handling safety risks.

[0140] Here, the medical relevance gateway is the decision-maker: based on clinical rules and real-time security conditions, it decides which data should be given more importance and which data should be ignored. The dynamic weight allocator is the executor: it receives instructions from the medical relevance gateway and executes them technically, dynamically calculating and allocating the final fusion weights of each modality of data through mathematical algorithms (such as weighted average, meta-learning-based networks, gating mechanisms, etc.).

[0141] Device Examples

[0142] Figure 3 This is a diagram of a stratified rehabilitation and quantitative device for stroke provided in an embodiment of the present invention.

[0143] In a second aspect, the present invention also proposes a stratified rehabilitation quantification device for stroke, combined with Figure 3 ,include:

[0144] Data Acquisition Module 31: Used to collect historical data of stroke patients and input the historical data of stroke patients into the stroke intelligent stratification engine for stratification and determination of stratification labels;

[0145] Adjustment module 32: used to acquire multimodal data and basic weight allocation table of stroke patients. The dynamic weight allocator adjusts the weights of the multimodal data of patients according to the situation of multimodal data and weight adjustment logic.

[0146] First determination module 33: used to determine the stratification correction coefficient based on the stratification label, determine the baseline severity coefficient based on the NIHSS baseline score, and determine the standardized score of the multimodal data;

[0147] The second determining module 34 is used to determine the rehabilitation progress index based on the adjusted weights of the patient's multimodal data, the standardized scores of the multimodal data, the stratification correction coefficient, and the baseline severity coefficient.

[0148] The dimensions of the content in the layer include: disease stage, lesion localization, and functional impairment level. The subcategories of disease stage include: acute phase, subacute phase, and chronic phase. The subcategories of lesion localization include: cortical type, brainstem type, and cerebellar type. The subcategories of functional impairment level include: mild, moderate, and severe. The multimodal data includes: motor data, physiological characteristics, brain-computer interface, and demographic / medical history.

[0149] A third aspect of the present invention also provides an electronic device comprising: a processor and a memory;

[0150] The processor executes a stroke stratification rehabilitation quantification method by calling programs or instructions stored in memory, as described above.

[0151] In a fourth aspect, the present invention also provides a computer-readable storage medium that stores a program or instructions that cause a computer to execute a stroke stratification rehabilitation quantification method as described above.

[0152] Figure 4 This is a schematic block diagram of an electronic device provided in an embodiment of the present invention.

[0153] like Figure 4 As shown, the electronic device includes at least one processor 401, at least one memory 402, and at least one communication interface 403. The various components of the electronic device are coupled together via a bus system 404. The communication interface 403 is used for information transmission with external devices. It is understood that the bus system 404 is used to implement communication between these components. In addition to a data bus, the bus system 404 also includes a power bus, a control bus, and a status signal bus. However, for clarity, ... Figure 4 The general designated all buses as Bus System 404.

[0154] It is understood that the memory 402 in this embodiment can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory.

[0155] In some implementations, memory 402 stores elements such as executable units or data structures, or subsets thereof, or extended sets thereof: operating systems and applications.

[0156] The operating system includes various system programs, such as the framework layer, core library layer, and driver layer, used to implement various basic business functions and handle hardware-based tasks. The application programs include various applications, such as media players and browsers, used to implement various application functions. A program implementing any method in the stroke stratified rehabilitation quantification method provided in this embodiment of the invention can be included in the application programs.

[0157] In this embodiment of the invention, the processor 401 executes the steps of various embodiments of the stroke stratified rehabilitation quantification method provided by the present invention by calling the program or instructions stored in the memory 402, specifically, the program or instructions stored in the application program.

[0158] Historical data of stroke patients are collected and input into the stroke intelligent stratification engine to determine stratification labels.

[0159] The system acquires multimodal data and a basic weight allocation table of stroke patients. The dynamic weight allocator adjusts the weights of the multimodal data of the patients based on the patient's multimodal data and the weight adjustment logic.

[0160] The stratification correction coefficient is determined based on the stratification label, the baseline severity coefficient is determined based on the NIHSS baseline score, and the standardized score of the multimodal data is determined.

[0161] The rehabilitation progress index was determined based on the adjusted weights of the patient's multimodal data, the standardized scores of the multimodal data, the stratification correction coefficient, and the baseline severity coefficient.

[0162] The dimensions of the content in the layer include: disease stage, lesion location, and functional impairment level. The subcategories of disease stage include: acute phase, subacute phase, and chronic phase. The subcategories of lesion location include: cortical type, brainstem type, and cerebellar type. The subcategories of functional impairment level include: mild, moderate, and severe.

[0163] Multimodal data includes: motion data, physiological characteristics, brain-computer interfaces, and demographic / medical history.

[0164] Any method in the stroke stratified rehabilitation quantification method provided in this embodiment of the invention can be applied to, or implemented by, the processor 401. The processor 401 can be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 401 or by instructions in software form. The processor 401 can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor.

[0165] The steps of any method in the stroke stratified rehabilitation quantification method provided in this embodiment of the invention can be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software units in the decoding processor. The software units can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory 402, and processor 401 reads the information in memory 402 and combines it with hardware to complete the steps of the method.

[0166] Those skilled in the art will understand that although some embodiments described herein include certain features included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of the invention and form different embodiments.

[0167] Those skilled in the art will understand that the descriptions of the various embodiments have different focuses, and for parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0168] Although embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention. All such modifications and variations fall within the scope defined by the appended claims. The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0169] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A quantitative method for stratified rehabilitation of stroke patients, characterized in that, include: Historical data of stroke patients are collected and input into the stroke intelligent stratification engine to determine stratification labels. The system acquires multimodal data and a basic weight allocation table of stroke patients. The dynamic weight allocator adjusts the weights of the multimodal data of the patients based on the patient's multimodal data and the weight adjustment logic. The stratification correction coefficient is determined based on the stratification label, the baseline severity coefficient is determined based on the NIHSS baseline score, and the standardized score of the multimodal data is determined. The rehabilitation progress index was determined based on the adjusted weights of the patient's multimodal data, the standardized scores of the multimodal data, the stratification correction coefficient, and the baseline severity coefficient. The dimensions of the content in the layer include: disease stage, lesion location, and functional impairment level. The subcategories of disease stage include: acute phase, subacute phase, and chronic phase. The subcategories of lesion location include: cortical type, brainstem type, and cerebellar type. The subcategories of functional impairment level include: mild, moderate, and severe. Multimodal data includes: motion data, physiological characteristics, brain-computer interfaces, and demographic / medical history.

2. The method for quantifying stratified rehabilitation of stroke patients according to claim 1, characterized in that, The weight adjustment logic includes: Data credibility compensation, recovery phase transition, and outlier penalty; The data reliability includes: if the missing rate of one of the modal data in the multimodal data is greater than the preset missing rate, its weight is proportionally transferred to the preset modal data; The stage transition includes a weighting scheme that automatically switches to the next disease stage when the weekly growth rate of the Fugl-Meyer rating scale score is greater than or equal to the preset growth rate. The outlier penalty includes: if an anomaly is detected in one of the modal data in the multimodal data, temporarily reducing the weight of the preset modal data to a preset percentage.

3. The method for quantifying stratified rehabilitation of stroke patients according to claim 1, characterized in that, The stratification correction factor is determined based on the stratification label, and the baseline severity factor is determined based on the NIHSS baseline score, including: In the hierarchical labeling, the hierarchical correction factor is 1.0 for the cortical type, 0.8 for the brainstem type, and 0.65 for the cerebellar type. The baseline severity coefficient β, determined based on the NIHSS baseline score, is expressed by the following formula: β = 1 + NIHSS baseline score / NIHSS baseline maximum score The NIHSS baseline maximum score is 42.

4. The method for quantifying stratified rehabilitation of stroke patients according to claim 1, characterized in that, The formula for determining the standardized score of multimodal data is as follows: The standardized score of motion data is determined by gait symmetry, using the following formula: The standardized score of physiological characteristics is determined by blood oxygen stability, using the following formula: The standardized score of a brain-computer interface is determined by the motor intent recognition rate, using the following formula: Among them, a blood oxygen fluctuation range of >20% will receive 0 points.

5. The method for quantifying stratified rehabilitation of stroke patients according to claim 1, characterized in that, The rehabilitation progress index is determined based on the adjusted weights of the patient's multimodal data, the standardized scores of the multimodal data, the stratification correction factor, and the baseline severity factor, using the following formula: Among them, W i N represents the adjusted weights of the i-th modality data. i denoted as the standardized score of the i-th data category, K represents the stratification correction coefficient, β represents the baseline severity coefficient, and RehabIndex represents the recovery progress index.

6. The method for quantifying stratified rehabilitation of stroke patients according to claim 1, characterized in that, If the missing rate of one modality in the multimodal data is greater than the preset missing rate, its weight is proportionally transferred to the preset modality data, including: If the missing rate of motion data in multimodal data is greater than the preset missing rate, the weight of motion data is transferred to physiological features; If the missing rate of physiological features in the multimodal data is greater than the preset missing rate, the weights of the physiological features are transferred to the motion data; If the missing rate of brain-computer interfaces in the multimodal data is greater than the preset missing rate, the weight of brain-computer interfaces is transferred to the motion data; If the demographic / medical history missing rate in the multimodal data is greater than the preset missing rate, the demographic / medical history weights are transferred to the motion data.

7. The method for quantifying stratified rehabilitation of stroke patients according to claim 1, characterized in that, The method further includes: After receiving new modal data, the medical relevance extension gateway performs cross-modal correlation analysis with the existing multimodal data; If the cross-modal correlation analysis shows that the Pearson correlation coefficient is greater than the preset coefficient, then the corresponding new modal data will be retained. The importance score of new modality data for predicting rehabilitation progress was calculated using a random forest model; The new modal data includes: biochemical indicators, medical imaging data, microenvironment and nutrition data, biomechanical data, and patient-reported outcomes.

8. A stratified rehabilitation and quantitative device for stroke, characterized in that, include: The data acquisition module is used to collect historical data of stroke patients and input the historical data of stroke patients into the stroke intelligent stratification engine to determine the stratification labels. Adjustment module: Used to acquire multimodal data and basic weight allocation table of stroke patients. The dynamic weight allocator adjusts the weights of the patient's multimodal data according to the patient's multimodal data and weight adjustment logic. The first determination module is used to determine the stratification correction coefficient based on the stratification label, the baseline severity coefficient based on the NIHSS baseline score, and the standardized score of the multimodal data. The second determination module is used to determine the rehabilitation progress index based on the adjusted weights of the patient's multimodal data, the standardized scores of the multimodal data, the stratification correction coefficient, and the baseline severity coefficient. The dimensions of the content in the layer include: disease stage, lesion localization, and functional impairment level. The subcategories of disease stage include: acute phase, subacute phase, and chronic phase. The subcategories of lesion localization include: cortical type, brainstem type, and cerebellar type. The subcategories of functional impairment level include: mild, moderate, and severe. The multimodal data include: motor data, physiological characteristics, brain-computer interface, and demographic / medical history.

9. An electronic device, characterized in that, include: Processor and memory; The processor executes a stroke stratified rehabilitation quantification method as described in any one of claims 1 to 7 by calling the program or instructions stored in the memory.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program or instructions that cause a computer to perform a stratified rehabilitation quantification method for stroke as described in any one of claims 1 to 7.

Images