Method and device for designing lower limb rehabilitation program using multi-source information and case-based reasoning

Abstract

The present invention provides a method and device for designing a lower limb rehabilitation program using multi-source information and case reasoning, which collects multi-source data of patients to be rehabilitated, generates a target feature attribute vector and an attribute weight vector based on the multi-source data, wherein the multi-source data includes limb movement information, contact force with a rehabilitation robot, and bioelectric information during the patient's movement; based on the target feature attribute vector and the attribute weight vector, determines a target case matching the patient to be rehabilitated in a rehabilitation patient case library, wherein the rehabilitation patient case library includes feature attribute vectors and rehabilitation program attribute vectors of multiple rehabilitation patients; based on the feature attribute vector of the matching rehabilitation patient, the target feature attribute vector of the patient to be rehabilitated, and the attribute weight vector, the rehabilitation program attribute vector of the matching rehabilitation patient is corrected to obtain a target rehabilitation program for the patient to be rehabilitated. The purpose of reducing the workload of doctors is achieved, and the adaptation degree of rehabilitation prescriptions is higher.

Description

Technical Field

[0001] The present invention relates to the technical field of rehabilitation program design, and in particular to a method and device for designing a lower limb rehabilitation program using multi-source information and case reasoning. Background Art

[0002] Due to the high incidence of accidents and the accelerated aging process, people's demand for rehabilitation treatment is increasing. With the development of bio-electromechanical technology, rehabilitation robots have become an important technical means for rehabilitation training. Rehabilitation program design is an important part of rehabilitation treatment. The quality of the program plays a decisive role in the effect of rehabilitation treatment. How to design scientific, reasonable and objective rehabilitation programs for different patients is an urgent problem to be solved.

[0003] Existing similar published patents:

[0004] A method and system for rehabilitation training management (CN 114242203 A), the steps of this patent are: Step 1: Determine the attribute information and initial physiological characteristic information of the target user; Step 2: Input the attribute information and initial physiological characteristic information into a pre-trained rehabilitation training model to obtain a rehabilitation training plan for the target user within the training cycle, the rehabilitation training plan includes at least one training monitoring item, the training cycle includes a preceding training cycle and a subsequent training cycle, and the preceding training cycle is located before the subsequent training cycle. Step 3: Determine the updated physiological characteristic information of the target user under at least one training monitoring item in the preceding training cycle. Step 4: Input the attribute information and updated physiological characteristic information of the target user into the rehabilitation training model to obtain a rehabilitation training plan for the target user in the subsequent training cycle.

[0005] A method for evaluating the motor function of lower limb rehabilitation robot rehabilitation training (CN110570946A) includes the following steps: Step 1: determining the evaluation index of the factor set; Step 2: determining the weight of the evaluation index; Step 3: determining the comments of the evaluation index at each level.

[0006] Existing technologies focus on how to obtain reasonable rehabilitation evaluations, but rarely study how to design rehabilitation prescriptions suitable for current patients after obtaining rehabilitation evaluations. There are few medical semi-autonomous systems that systematically design rehabilitation prescriptions and allow doctors to decide whether to use the prescriptions. The large amount of motion data during patient training is not fully utilized.

[0007] Therefore, how to get rid of the dilemma that the design of patient rehabilitation training programs depends on the doctor's experience, and how to design rehabilitation programs based on the differences in the characteristics of different patients to be rehabilitated are still urgent problems to be solved by technical personnel in this field. Summary of the invention

[0008] The present invention provides a method and device for designing a lower limb rehabilitation program using multi-source information and case reasoning, so as to solve the problem that the existing patient rehabilitation training program design relies on the experience of doctors and the rehabilitation program is designed according to the differences in the characteristics of different patients to be rehabilitated.

[0009] The present invention provides a method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning, comprising:

[0010] Collecting multi-source data of the patient to be rehabilitated, and generating a target feature attribute vector and an attribute weight vector based on the multi-source data, wherein the multi-source data includes limb movement information, contact force with the rehabilitation robot, and bioelectric information of the patient during movement;

[0011] Based on the target characteristic attribute vector and the target characteristic attribute weight vector, determining a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated, wherein the rehabilitation patient case library includes characteristic attribute vectors and rehabilitation program attribute vectors of multiple rehabilitation patients;

[0012] The rehabilitation program attribute vector of the matched rehabilitation patient is modified based on the characteristic attribute vector of the matched rehabilitation patient, the target characteristic attribute vector and the target characteristic attribute weight vector to obtain the target rehabilitation program for the patient to be rehabilitated.

[0013] According to a method for designing a lower limb rehabilitation program using multi-source information and case reasoning, the limb movement information includes the movement position, speed and acceleration of the lower limb hip, the movement position, speed and acceleration of the knee, and the movement position, speed and acceleration of the ankle joint; the contact force includes the contact force between the end pedal of the rehabilitation robot and the contact surface of the sole of the foot; and the bioelectric information includes electrocardiogram signals and surface electromyography signals of the lower limbs.

[0014] According to a method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning provided by the present invention, the target feature attribute weight vector is generated by manually evaluating each feature attribute in the target feature attribute vector.

[0015] According to a method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning provided by the present invention, based on the target feature attribute vector and the target feature attribute weight vector, a matching rehabilitation patient in a rehabilitation patient case library that matches the patient to be rehabilitated is determined, specifically comprising:

[0016] The similarity between the characteristic attribute vector of each rehabilitation patient in the rehabilitation patient case library and the target characteristic attribute vector of the patient to be rehabilitated is calculated, and the rehabilitation patient with the greatest similarity is determined as the matching rehabilitation patient, and the matching rehabilitation patient case is determined based on the matching rehabilitation patient.

[0017] According to a method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning provided by the present invention, the calculation of the similarity is implemented by the following formula:

[0018]

[0019] Among them, the rehabilitation patient case library C = {c1,c2,…,c p ,…,c m}, the pth recovered patient c in C [ The characteristic attribute vector of n is the number of characteristic attributes, i = 1, 2, ..., n, and the characteristic attribute vector of the patient to be recovered is [x1, x2, ..., x n ] T , is the weight value of the i-th feature attribute, k p is the similarity between the characteristic attribute vector of the pth recovered patient in C and the target characteristic attribute vector of the patient to be recovered.

[0020] According to a method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning provided by the present invention, the rehabilitation program attribute vector of the matched rehabilitation patient is corrected based on the characteristic attribute vector of the matched rehabilitation patient, the target characteristic attribute vector and the target characteristic attribute weight vector to obtain the target rehabilitation program for the patient to be rehabilitated, specifically comprising:

[0021] Determining a feature attribute difference matrix based on the feature attribute vector of the matched rehabilitation patient and the target feature attribute vector;

[0022] Determining a correction matrix based on the feature attribute difference matrix and the target feature attribute weight vector;

[0023] Based on the correction matrix and the rehabilitation program attribute vector of the matching rehabilitation patient, a target rehabilitation program attribute vector of the patient to be rehabilitated is determined.

[0024] According to a method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning provided by the present invention, the characteristic attribute difference matrix is ​​determined based on matching the characteristic attribute vector of the rehabilitation patient and the target characteristic attribute vector, specifically including:

[0025] The difference matrix DIFF is determined by the following formula:

[0026]

[0027] Among them, the characteristic attribute vector of matching rehabilitation patients is n is the number of feature attributes, s = 1, 2, ..., n, and the target feature attribute vector X = [x1, x2, ..., x s ,…,xn ] T , To match the qth program content in the rehabilitation program attribute vector of the rehabilitation patient, q = 1, 2, ..., k, k is the total number of rehabilitation program contents;

[0028] Correspondingly, the determining of the correction matrix based on the feature attribute difference matrix and the target feature attribute weight vector specifically includes:

[0029] The correction matrix R is determined by the following formula:

[0030]

[0031] Among them, the target feature attribute weight vector

[0032] The step of determining a target rehabilitation program attribute vector for the patient to be rehabilitated based on the correction matrix and the rehabilitation program attribute vector of the matched rehabilitation patient specifically includes:

[0033] The target rehabilitation program attribute vector of the patient to be rehabilitated is determined by the following formula

[0034]

[0035] Among them, the rehabilitation plan attribute vector matching the rehabilitation patient

[0036] According to the present invention, a lower limb rehabilitation program design device using multi-source information and case reasoning includes:

[0037] A feature unit, used to collect multi-source data of the patient to be rehabilitated, and generate a target feature attribute vector and an attribute weight vector based on the multi-source data, wherein the multi-source data includes limb movement information, contact force with the rehabilitation robot, and bioelectric information of the patient during movement;

[0038] A matching unit, configured to determine, based on the target feature attribute vector and the target feature attribute weight vector, a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated, wherein the rehabilitation patient case library includes feature attribute vectors and rehabilitation program attribute vectors of a plurality of rehabilitation patients;

[0039] The correction unit is used to correct the rehabilitation program attribute vector of the matching rehabilitation patient based on the characteristic attribute vector of the matching rehabilitation patient, the target characteristic attribute vector and the target characteristic attribute weight vector to obtain the target rehabilitation program for the patient to be rehabilitated.

[0040] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the method for designing a lower limb rehabilitation program using multi-source information and case reasoning as described in any one of the above are implemented.

[0041] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for designing a lower limb rehabilitation program using multi-source information and case reasoning as described in any one of the above.

[0042] The method and device for designing a lower limb rehabilitation program using multi-source information and case reasoning provided by the present invention collects multi-source data of patients to be rehabilitated, generates a target feature attribute vector and an attribute weight vector based on the multi-source data, wherein the multi-source data includes limb movement information, contact force with a rehabilitation robot, and bioelectric information during the patient's movement; based on the target feature attribute vector and the target feature attribute weight vector, determines a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated, wherein the rehabilitation patient case library includes feature attribute vectors and rehabilitation program attribute vectors of multiple rehabilitation patients; based on the feature attribute vector of the matching rehabilitation patient, the target feature attribute vector and the target feature attribute weight vector, the rehabilitation program attribute vector of the matching rehabilitation patient is corrected to obtain a target rehabilitation program for the patient to be rehabilitated. The method and device provided by the present invention can assist doctors in formulating rehabilitation training prescriptions by the system, which reduces the workload of doctors, and the prescriptions adopt a large number of historical successful rehabilitation training cases, making the prescription design more standardized and more adaptable. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 A schematic diagram of a flow chart of a method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning provided by the present invention;

[0045] Figure 2 A schematic flow chart of an overall design method for a rehabilitation program design system using multi-source information and case-based reasoning provided by the present invention;

[0046] Figure 3 A case reasoning flow chart provided by the present invention;

[0047] Figure 4A flowchart of case correction provided by the present invention;

[0048] Figure 5 A schematic diagram of the structure of a lower limb rehabilitation program design device using multi-source information and case-based reasoning provided by the present invention;

[0049] Figure 6 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] The existing rehabilitation program design technology generally has the problem that the rehabilitation program design relies on the doctor's experience and does not design rehabilitation programs for the differences in the characteristics of different patients to be rehabilitated. Figure 1 The present invention describes a method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning. Figure 1 A flow chart of a method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning provided by the present invention is shown in FIG. Figure 1 As shown, the method includes:

[0052] Step 110, collect multi-source data of the patient to be rehabilitated, and generate a target feature attribute vector and an attribute weight vector based on the multi-source data, wherein the multi-source data includes limb movement information, contact force with the rehabilitation robot, and bioelectric information of the patient during movement.

[0053] Specifically, it is necessary to first collect multi-source data of the patient to be rehabilitated. It should be noted here that the multi-source data includes limb movement information, contact force with the rehabilitation robot and bioelectric information of the patient during movement. Limb movement information includes the position, speed and acceleration of important nodes of the lower limbs, which is collected by collecting images of the lower limbs of the person and / or inertial sensors installed at the corresponding nodes of the lower limbs; the contact force with the rehabilitation robot is the contact force between the end pedal of the rehabilitation robot and the contact surface of the sole of the foot, which is collected by the pressure sensor placed on the pedal; bioelectric information includes electrocardiogram and lower limb surface electromyography signals, which are collected by sensors worn by the patient to be rehabilitated. The collected multi-source data are sorted into a multi-source data set, and then the multi-source data set is converted into a feature attribute vector in a specific way. Here, the feature attribute vector of the patient to be rehabilitated is the target feature attribute vector, and each feature attribute is given a feature attribute weight value according to its influence in rehabilitation training to obtain a target feature attribute weight vector.

[0054] Step 120, based on the target feature attribute vector and the target feature attribute weight vector, determine a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated, wherein the rehabilitation patient case library includes feature attribute vectors and rehabilitation program attribute vectors of multiple rehabilitation patients.

[0055] Specifically, the rehabilitation patient case library is first explained. The rehabilitation patient case library stores characteristic attribute vectors of multiple patients who have been rehabilitated and corresponding rehabilitation program attribute vectors. The rehabilitation program attribute vector usually includes elements such as the time required for rehabilitation training, the rehabilitation training cycle, etc. A set of parameters related to rehabilitation training is used as a rehabilitation training vector. The cases of successful rehabilitation of these patients are combined into two vectors to form a rehabilitation patient case library. The characteristic attribute vectors of all rehabilitation patients in the rehabilitation patient case library are searched for the characteristic attribute vectors of matching rehabilitation patient cases that match the target characteristic attribute vectors of the patients to be rehabilitated. The matching here can be the highest similarity or the largest distance between the two. In the calculation process, the weight values ​​of each element in the vector are used to calculate the similarity or distance between the two vectors. There are many common ways, which are not specifically limited here.

[0056] It should be noted here that the rehabilitation patient case library stores case information of multiple recovered patients. The case information includes characteristic attribute vectors and rehabilitation plan attribute vectors. The recovered patients in the matching rehabilitation patient case are the matching rehabilitation patients. The characteristic attribute vector of the matching rehabilitation patient is the characteristic attribute vector of the matching rehabilitation patient case, and the rehabilitation plan attribute vector of the matching rehabilitation patient is the rehabilitation plan attribute vector of the matching rehabilitation patient case.

[0057] Step 130, based on the characteristic attribute vector of the matched rehabilitation patient, the target characteristic attribute vector and the target characteristic attribute weight vector, the rehabilitation program attribute vector of the matched rehabilitation patient is modified to obtain the target rehabilitation program for the patient to be rehabilitated.

[0058] Specifically, after matching the rehabilitation patient database, we obtain matching rehabilitation patient cases that are most similar to the conditions of the patients to be rehabilitated. At this time, the rehabilitation plan of the matching rehabilitation patient cases cannot be directly used as the rehabilitation plan for the patients to be rehabilitated. We must use the difference between the target characteristic attribute vector of the patients to be rehabilitated and the characteristic attribute vector of the matching rehabilitation patients to correct the rehabilitation plan attribute vector of the matching rehabilitation patients. Only in this way can the target rehabilitation plan attribute vector obtained be more suitable for the patients to be rehabilitated.

[0059] The method provided by the present invention collects multi-source data of patients to be rehabilitated, generates a target feature attribute vector and an attribute weight vector based on the multi-source data, wherein the multi-source data includes limb movement information, contact force with a rehabilitation robot, and bioelectric information during patient movement; based on the target feature attribute vector and the target feature attribute weight vector, determines a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated, wherein the rehabilitation patient case library includes feature attribute vectors and rehabilitation program attribute vectors of multiple rehabilitation patients; based on the feature attribute vector of the matching rehabilitation patient, the target feature attribute vector and the target feature attribute weight vector, the rehabilitation program attribute vector of the matching rehabilitation patient is corrected to obtain a target rehabilitation program for the patient to be rehabilitated. The method and device provided by the present invention can assist doctors in formulating rehabilitation training prescriptions by the system, which reduces the workload of doctors, and the prescriptions adopt a large number of historical successful rehabilitation training cases, making the prescription design more standardized and more adaptable.

[0060] Based on the above embodiment, in this method, the limb movement information includes the movement position, speed and acceleration of the lower limb hip, the movement position, speed and acceleration of the knee, and the movement position, speed and acceleration of the ankle joint; the contact force includes the contact force between the end pedal of the rehabilitation robot and the contact surface of the sole of the foot; the bioelectric information includes electrocardiogram signals and surface electromyography signals of the lower limbs.

[0061] Specifically, in the robot-assisted lower limb rehabilitation training scenario, the information that needs to be obtained mainly includes: limb movement information, contact force with the rehabilitation robot, and bioelectric information during movement. Among them, limb movement information includes the movement position, speed, and acceleration of the hip, knee, and ankle joints of the lower limbs, which are collected by images and 7 inertial sensors located at the waist, left and right thighs, left and right calves, and left and right insteps; the contact force with the rehabilitation robot is the contact force between the end pedal of the rehabilitation robot and the contact surface of the sole, which is collected by the pressure sensor placed on the pedal; bioelectric information includes electrocardiogram and surface electromyography signals of the lower limbs. The electrocardiogram signals are collected by the wrist electrocardiometer worn by the individual, and the electromyography signals of the lower limbs are collected by 8 surface-mounted electromyography sensors distributed in the tibialis anterior, gastrocnemius, rectus femoris, and biceps femoris of the bilateral lower limbs. For example, the multi-source data set B obtained as above may include joint position B1, movement speed B2, movement acceleration B3, contact force B4, electromyography B5, and electrocardiogram B6.

[0062] A feature attribute vector set A is selected, and the elements in A include muscle strength A1, muscle tension A2, joint mobility A3, muscle coordination A4, movement smoothness A5, and fatigue A6. Table 1 is a relationship table between multi-source information and feature attributes, and the acquisition method is shown in Table 1.

[0063] Table 1: Relationship between multi-source information and feature attributes

[0064]

[0065] Each characteristic attribute in Table 1 is obtained from multi-source information. For example, fatigue level A6 can be assessed by electromyography B5 and electrocardiogram B6.

[0066] Select all elements in the feature attribute set A and the patient's age, gender, and rehabilitation evaluation level to generate the patient's feature attribute vector X = [x1, x2, ..., x9] T , obtain the feature attribute weight vector corresponding to the patient feature attribute vector

[0067] Based on the above embodiment, in this method, the target feature attribute weight vector is generated by manually evaluating each feature attribute in the target feature attribute vector.

[0068] Specifically, each characteristic attribute in the target characteristic attribute vector is evaluated by manual evaluation. For example, Table 2 is a patient rehabilitation information attribute table. The rehabilitation information characteristic attributes and their weight values ​​are shown in Table 2 below.

[0069] Table 2: Patient rehabilitation information attribute table

[0070]

[0071] The embodiment of the present invention adopts an expert survey method to obtain the characteristic attribute weight value of the patient, and the steps are as follows:

[0072] (1) The nine characteristic attributes in Table 2 are used as evaluation indicators.

[0073] (2) Select h experts, send the nine indicators and related information to the selected experts, and ask them to independently give the weight value ω of each indicator. ij Among them, ω ij is the weight value given by the j-th expert to the i-th indicator, j = 1, 2, …, h, i = 1, 2, …, 9.

[0074] (3) Calculate the average weight of each indicator in is the average weight value of the i-th indicator.

[0075] Based on the above embodiment, in the method, based on the target feature attribute vector and the target feature attribute weight vector, determining a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated specifically includes:

[0076] The similarity between the characteristic attribute vector of each rehabilitation patient in the rehabilitation patient case library and the target characteristic attribute vector of the patient to be rehabilitated is calculated, and the rehabilitation patient with the greatest similarity is determined as the matching rehabilitation patient, and the matching rehabilitation patient case is determined based on the matching rehabilitation patient.

[0077] Specifically, the similarity between the characteristic attribute vectors of all rehabilitation patients in the rehabilitation patient case library and the target characteristic attribute vector is traversed. Here, Euclidean distance, Mahalanobis distance, Chebyshev distance, etc. can be used, which are not specifically limited here. It should be noted here that the calculation of similarity also uses the target characteristic attribute weight vector. Finally, the rehabilitation patient corresponding to the characteristic attribute vector with the greatest similarity is used as the matching rehabilitation patient.

[0078] Based on the above embodiment, in this method, the calculation of the similarity is implemented by the following formula:

[0079]

[0080] Among them, the rehabilitation patient case library C = {c1,c2,…,c p ,…,c m}, the pth recovered patient c in C p The characteristic attribute vector of n is the number of characteristic attributes, i = 1, 2, ..., n, and the characteristic attribute vector of the patient to be recovered is [x1, x2, ..., x n ] T , is the weight value of the i-th feature attribute, k p is the similarity between the characteristic attribute vector of the pth recovered patient in C and the target characteristic attribute vector of the patient to be recovered.

[0081] Specifically, the similarity between the characteristic attribute vector of each rehabilitation patient and the target characteristic attribute vector of the patient to be rehabilitated is calculated, that is, the similarity between each rehabilitation patient and the patient to be rehabilitated is calculated, and the calculation formula of the similarity is:

[0082]

[0083] Among them, the rehabilitation patient case library C = {c1,c2,…,c p ,…,c m}, the pth recovered patient c in C p The characteristic attribute vector of n is the number of characteristic attributes, i = 1, 2, ..., n, and the characteristic attribute vector of the patient to be recovered is [x1, x2, ..., x n ] T , is the weight value of the i-th feature attribute, k pis the similarity between the characteristic attribute vector of the pth recovered patient in C and the target characteristic attribute vector of the patient to be recovered.

[0084] For example, suppose the rehabilitation patient case library C = {c1,c2,…,c p ,…,c m}The pth recovered patient c p The attribute vector is The target feature attribute vector of the patient c to be recovered is [x1,x2,…,x9] T , then the patient to be recovered c and the recovered patient c p The similarity is:

[0085]

[0086] Where: k p ——Patients to be recovered c and recovered patients c p similarity;

[0087] ——The i-th attribute of the p-th recovered patient;

[0088] x i ——The i-th attribute of the patient to be recovered.

[0089] Based on the above embodiment, in the method, the rehabilitation program attribute vector of the matched rehabilitation patient is corrected based on the characteristic attribute vector of the matched rehabilitation patient, the target characteristic attribute vector and the target characteristic attribute weight vector to obtain the target rehabilitation program for the patient to be rehabilitated, specifically including:

[0090] Determining a feature attribute difference matrix based on the feature attribute vector of the matched rehabilitation patient and the target feature attribute vector;

[0091] Determining a correction matrix based on the feature attribute difference matrix and the target feature attribute weight vector;

[0092] Based on the correction matrix and the rehabilitation program attribute vector of the matching rehabilitation patient, a target rehabilitation program attribute vector of the patient to be rehabilitated is determined.

[0093] Specifically, it is necessary to use the first difference between the characteristic attribute vector of the matched rehabilitation patient and the target characteristic attribute vector of the patient to be rehabilitated to derive the second difference between the rehabilitation plan attribute vector of the matched rehabilitation patient and the target rehabilitation plan for the patient to be rehabilitated. The target rehabilitation plan for the patient to be rehabilitated can be obtained by correcting the rehabilitation plan attribute vector of the matched rehabilitation patient with the second difference.

[0094] Based on the above embodiment, in the method, determining the feature attribute difference matrix based on matching the feature attribute vector of the rehabilitation patient and the target feature attribute vector specifically includes:

[0095] The difference matrix DIFF is determined by the following formula:

[0096]

[0097]

[0098] Among them, the characteristic attribute vector of matching rehabilitation patients is n is the number of feature attributes, s = 1, 2, ..., n, and the target feature attribute vector X = [x1, x2, ..., x s ,…,x n ] T , To match the qth program content in the rehabilitation program attribute vector of the rehabilitation patient, q = 1, 2, ..., k, k is the total number of rehabilitation program contents;

[0099] Correspondingly, the determining of the correction matrix based on the feature attribute difference matrix and the target feature attribute weight vector specifically includes:

[0100] The correction matrix R is determined by the following formula:

[0101]

[0102] Among them, the target feature attribute weight vector

[0103] The step of determining a target rehabilitation program attribute vector for the patient to be rehabilitated based on the correction matrix and the rehabilitation program attribute vector of the matched rehabilitation patient specifically includes:

[0104] The target rehabilitation program attribute vector of the patient to be rehabilitated is determined by the following formula

[0105]

[0106] Among them, the rehabilitation plan attribute vector matching the rehabilitation patient

[0107] Specifically, assuming the specific content of the rehabilitation training program: training cycle Number of training sessions Training time Training Method Training speed The characteristic attribute vector X of the patient to be recovered is [x1, x2, ..., x9] T , the characteristic attribute set matching the rehabilitation patients is The attribute vector of the rehabilitation program matching the rehabilitation patient is

[0108]

[0109] Calculate the difference matrix: For the qth rehabilitation program content of the matched rehabilitation patient For example, the deviation value of the sth characteristic attribute of the patient to be recovered is s qs Among them, q = 1, 2, ..., 5, is the number of elements of the rehabilitation program content that matches the rehabilitation patients, and s = 1, 2, ..., 9, is the number of characteristic attributes that match the rehabilitation patients.

[0110] The specific calculation process is shown in the following formula and Table 3.

[0111]

[0112] Table 3: Description of patient training program and characteristic attributes

[0113]

[0114] The difference attribute matrix DIFF between the characteristic attributes of the patients to be recovered and the matched patients to be recovered is:

[0115]

[0116] Get the correction matrix: for the qth rehabilitation program content of the matched rehabilitation patient For example, combined with the weighted average matrix in the previous step The correction matrix R can be obtained as:

[0117]

[0118] Among them, r b is the correction operator of the bth solution element of the case solution, b = 1, 2, …, 5.

[0119] Modified rehabilitation training program: The rehabilitation training program for patients to be rehabilitated can be obtained using the following formula:

[0120]

[0121] Based on the above embodiments, the present invention provides a method for designing a rehabilitation program system using multi-source information and case-based reasoning. Figure 2 The flowchart of the overall design method of the rehabilitation program design system using multi-source information and case-based reasoning provided by the present invention is as follows: Figure 2 As shown, the method includes:

[0122] Step 1: Multi-source data collection, feature attribute determination, and feature attribute weight value acquisition.

[0123] In the robot-assisted lower limb rehabilitation training scenario, the information that needs to be obtained mainly includes: limb movement information, contact force with the rehabilitation robot, and bioelectric information during movement. Among them, limb movement information includes the movement position, speed, and acceleration of the hip, knee, and ankle joints of the lower limbs, which are collected by images and 7 inertial sensors located at the waist, left and right thighs, left and right calves, and left and right insteps; the contact force with the rehabilitation robot is the contact force between the end pedal of the rehabilitation robot and the contact surface of the sole, which is collected by the pressure sensor placed on the pedal; bioelectric information includes electrocardiogram and surface electromyography signals of the lower limbs. The electrocardiogram signals are collected by the wrist electrocardiometer worn by the individual, and the electromyography signals of the lower limbs are collected by 8 surface-mounted electromyography sensors distributed in the tibialis anterior, gastrocnemius, rectus femoris, and biceps femoris of the bilateral lower limbs. The multi-source data set B obtained as above includes joint position B1, movement speed B2, movement acceleration B3, contact force B4, electromyography B5, and electrocardiogram B6.

[0124] The feature attribute set A is selected, including muscle strength A1, muscle tension A2, joint mobility A3, muscle coordination A4, movement smoothness A5, and fatigue A6, and the acquisition method is shown in Table 1.

[0125] The characteristic attributes in Table 1 are obtained from multi-source information. For example, fatigue level A6 can be assessed by electromyography B5 and electrocardiogram B6.

[0126] Select all elements in the feature attribute set A and the patient's age, gender, and rehabilitation evaluation level to generate the patient's feature attribute vector X = [x1, x2, ..., x9] T , obtain the feature attribute weight vector corresponding to the patient feature attribute vector The characteristic attributes of rehabilitation information and their weight values ​​are shown in Table 2.

[0127] The present invention adopts the expert survey method to obtain the characteristic attribute weight value of the patient, and the steps are as follows:

[0128] (1) The nine characteristic attributes in Table 2 are used as evaluation indicators.

[0129] (2) Select n experts, send the nine indicators and related information to the selected experts, and ask them to independently give the weight value ω of each indicator. ij Among them, ω ij is the weight value given by the j-th expert to the i-th indicator, j = 1, 2, …, n, i = 1, 2, …, 9.

[0130] (3) Calculate the average weight of each indicator in is the average weight value of the i-th indicator.

[0131] Step 2: Rehabilitation plan matching based on case reasoning.

[0132] The feature attributes and feature attribute weight values ​​in step 1 are used to generate a temporary case that does not contain a rehabilitation plan, and the temporary case is indexed and compared with all cases in the case library to obtain a matching case. Figure 3 The case reasoning flowchart provided by the present invention is specifically matched with the flowchart as follows Figure 3 shown.

[0133] The case with the highest similarity to the temporary case is the matching case. The similarity is calculated using the K-nearest neighbor method. The specific calculation process is as follows:

[0134] Assume that the case library C = {c1,c2,…,c p ,…,c m} in case p p The attribute vector is The attribute vector of temporary case c (i.e., the patient's characteristic attribute vector X) is [x1, x2, ..., x9] T , then the pth case c p The similarity with the temporary case c is:

[0135]

[0136] Where: k p ——the similarity between the temporary case and the pth case;

[0137] ——the i-th attribute of the p-th case;

[0138] x i ——The i-th attribute of the temporary case.

[0139] Step 3: Case modification based on the difference attribute matrix.

[0140] The matching case includes the specific content of the rehabilitation training program: training cycle Number of training sessions Training time Training Method Training speed Therefore, the characteristic attribute set of the matching case is The rehabilitation program attribute set is Because of the differences between different patients, the matching cases were corrected. Figure 4 The case correction flow chart provided by the present invention is as follows: Figure 4 shown.

[0141] The case correction process based on the difference attribute matrix is ​​as follows:

[0142] (1) Extract the characteristic attribute set of matching cases and the attribute set of rehabilitation programs.

[0143] (2) Calculate the difference matrix. For the qth solution of the matching case For example, the deviation value of the sth conditional attribute of the temporary case is d qs . Among them, q = 1, 2, ..., 5, is the number of elements of the case solution matching the case, and s = 1, 2, ..., 9, is the number of characteristic attributes of the matching case. The specific calculation process is shown in the following formula and Table 3.

[0144]

[0145] The difference attribute matrix DIFF of the characteristic attributes of the temporary case and the matching case is:

[0146]

[0147] (3) Obtain the correction matrix. For the qth solution of the matching case For example, combined with the weighted average matrix in step 1 The correction matrix R can be obtained as:

[0148]

[0149] Among them, r b is the correction operator of the bth solution element of the case solution, b = 1, 2, …, 5.

[0150] (4) Modify the rehabilitation training program. Using Formula 4, the rehabilitation training program for the temporary case is:

[0151]

[0152] Step 4: Experts decide whether to update the case library.

[0153] The doctor decides whether the patient should follow the revised rehabilitation training program. If the patient follows the generated program, the patient's movement data during the rehabilitation training process needs to be collected. After a course of rehabilitation, the doctor uses the collected data to determine whether to generate the program as a new case and add it to the case library.

[0154] The advantages of the present invention are: the rehabilitation prescription is formulated by the system with the assistance of the doctor, which reduces the workload of the doctor; the prescription is generated before the patient training, and the rehabilitation prescription is designed by the system in combination with a large amount of case data, and the doctor decides whether to use the prescription, making the prescription design more standardized and objective; the generation process of the rehabilitation prescription of the present invention is more explainable, and the process of obtaining the rehabilitation prescription in a deep learning manner is carried out inside the computer, and the doctor cannot explain the reason for the generation of the rehabilitation prescription.

[0155] The following is a description of the lower limb rehabilitation program design device using multi-source information and case-based reasoning provided by the present invention. The lower limb rehabilitation program design device using multi-source information and case-based reasoning described below and the lower limb rehabilitation program design method using multi-source information and case-based reasoning described above can be referenced to each other.

[0156] Figure 5 The schematic diagram of the structure of the lower limb rehabilitation program design device using multi-source information and case reasoning provided by the present invention is as follows: Figure 5 As shown, the device includes a feature unit 510, a matching unit 520 and a correction unit 530, wherein:

[0157] The feature unit 510 is used to collect multi-source data of the patient to be rehabilitated, and generate a target feature attribute vector and an attribute weight vector based on the multi-source data, wherein the multi-source data includes limb movement information, contact force with the rehabilitation robot, and bioelectric information of the patient during movement;

[0158] The matching unit 520 is used to determine a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated based on the target feature attribute vector and the target feature attribute weight vector, wherein the rehabilitation patient case library includes feature attribute vectors and rehabilitation program attribute vectors of multiple rehabilitation patients;

[0159] The correction unit 530 is used to correct the rehabilitation plan attribute vector of the matched rehabilitation patient based on the characteristic attribute vector of the matched rehabilitation patient, the target characteristic attribute vector and the target characteristic attribute weight vector to obtain the target rehabilitation plan for the patient to be rehabilitated.

[0160] The lower limb rehabilitation program design device provided by the present invention utilizes multi-source information and case reasoning, collects multi-source data of patients to be rehabilitated, generates target feature attribute vectors and attribute weight vectors based on the multi-source data, wherein the multi-source data includes limb movement information, contact force with a rehabilitation robot, and bioelectric information during patient movement; based on the target feature attribute vector and the target feature attribute weight vector, determines a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated, wherein the rehabilitation patient case library includes feature attribute vectors and rehabilitation program attribute vectors of multiple rehabilitation patients; based on the feature attribute vector of the matching rehabilitation patient, the target feature attribute vector and the target feature attribute weight vector, the rehabilitation program attribute vector of the matching rehabilitation patient is corrected to obtain a target rehabilitation program for the patient to be rehabilitated. The method and device provided by the present invention can assist doctors in formulating rehabilitation training prescriptions by the system, which reduces the workload of doctors, and the prescriptions adopt a large number of historical successful rehabilitation training cases, making the prescription design more standardized and more adaptable.

[0161] Based on the above embodiments, in the device, the limb movement information includes the movement position, speed and acceleration of the lower limb hip, the movement position, speed and acceleration of the knee, and the movement position, speed and acceleration of the ankle joint; the contact force includes the contact force between the end pedal of the rehabilitation robot and the contact surface of the sole of the foot; and the bioelectric information includes electrocardiogram signals and surface electromyography signals of the lower limbs.

[0162] On the basis of the above embodiment, in the device, the target feature attribute weight vector is generated by manually evaluating each feature attribute in the target feature attribute vector.

[0163] Based on the above embodiments, in the device, the matching unit is specifically used to: calculate the similarity between the characteristic attribute vector of each rehabilitation patient in the rehabilitation patient case library and the target characteristic attribute vector of the patient to be rehabilitated, determine the rehabilitation patient with the greatest similarity as the matching rehabilitation patient, and determine the matching rehabilitation patient case based on the matching rehabilitation patient.

[0164] Based on the above embodiment, in the device, the calculation of the similarity is implemented by the following formula:

[0165]

[0166] Among them, the rehabilitation patient case library C = {c1,c2,…,c p ,…,c m}, the pth recovered patient c in C p The characteristic attribute vector of n is the number of characteristic attributes, i = 1, 2, ..., n, and the characteristic attribute vector of the patient to be recovered is [x1, x2, ..., x n ] T , is the weight value of the i-th feature attribute, k p is the similarity between the characteristic attribute vector of the pth recovered patient in C and the target characteristic attribute vector of the patient to be recovered.

[0167] Based on the above embodiment, in the device, the correction unit is specifically used for:

[0168] Determining a feature attribute difference matrix based on the feature attribute vector of the matched rehabilitation patient and the target feature attribute vector;

[0169] Determining a correction matrix based on the feature attribute difference matrix and the target feature attribute weight vector;

[0170] Based on the correction matrix and the rehabilitation program attribute vector of the matching rehabilitation patient, a target rehabilitation program attribute vector of the patient to be rehabilitated is determined.

[0171] On the basis of the above embodiment, in the device, determining the characteristic attribute difference matrix based on matching the characteristic attribute vector of the rehabilitation patient and the target characteristic attribute vector specifically includes:

[0172] The difference matrix DIFF is determined by the following formula:

[0173]

[0174] Among them, the characteristic attribute vector of matching rehabilitation patients is n is the number of feature attributes, s = 1, 2, ..., n, and the target feature attribute vector X = [x1, x2, ..., x s ,…,x n ] T , To match the qth program content in the rehabilitation program attribute vector of the rehabilitation patient, q = 1, 2, ..., k, k is the total number of rehabilitation program contents;

[0175] Correspondingly, the determining of the correction matrix based on the feature attribute difference matrix and the target feature attribute weight vector specifically includes:

[0176] The correction matrix R is determined by the following formula:

[0177]

[0178] Among them, the target feature attribute weight vector

[0179] The step of determining a target rehabilitation program attribute vector for the patient to be rehabilitated based on the correction matrix and the rehabilitation program attribute vector of the matched rehabilitation patient specifically includes:

[0180] The target rehabilitation program attribute vector of the patient to be rehabilitated is determined by the following formula

[0181]

[0182] Among them, the rehabilitation plan attribute vector matching the rehabilitation patient

[0183] Figure 6 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 6As shown, the electronic device may include: a processor (processor) 610, a communication interface (Communications Interface) 620, a memory (memory) 630 and a communication bus 640, wherein the processor 610, the communication interface 620, and the memory 630 communicate with each other through the communication bus 640. The processor 610 can call the logic instructions in the memory 630 to execute a lower limb rehabilitation program design method using multi-source information and case reasoning, the method comprising: collecting multi-source data of the patient to be rehabilitated, generating a target feature attribute vector and an attribute weight vector based on the multi-source data, wherein the multi-source data includes limb movement information, contact force with the rehabilitation robot and bioelectric information during the patient's movement; based on the target feature attribute vector and the target feature attribute weight vector, determining a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated, wherein the rehabilitation patient case library includes feature attribute vectors and rehabilitation program attribute vectors for multiple rehabilitation patients; based on the feature attribute vector of the matching rehabilitation patient, the target feature attribute vector and the target feature attribute weight vector, correcting the rehabilitation program attribute vector of the matching rehabilitation patient to obtain a target rehabilitation program for the patient to be rehabilitated.

[0184] In addition, the logic instructions in the above-mentioned memory 630 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0185] On the other hand, the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the lower limb rehabilitation program design method using multi-source information and case reasoning provided by the above-mentioned methods, the method including: collecting multi-source data of the patient to be rehabilitated, generating a target feature attribute vector and an attribute weight vector based on the multi-source data, wherein the multi-source data includes limb movement information, contact force with a rehabilitation robot, and bioelectric information during the patient's movement; based on the target feature attribute vector and the target feature attribute weight vector, determining a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated, wherein the rehabilitation patient case library includes feature attribute vectors and rehabilitation program attribute vectors of multiple rehabilitation patients; based on the feature attribute vector of the matching rehabilitation patient, the target feature attribute vector and the target feature attribute weight vector, the rehabilitation program attribute vector of the matching rehabilitation patient is corrected to obtain a target rehabilitation program for the patient to be rehabilitated.

[0186] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the above-mentioned lower limb rehabilitation program design method using multi-source information and case reasoning, the method comprising: collecting multi-source data of the patient to be rehabilitated, generating a target feature attribute vector and an attribute weight vector based on the multi-source data, wherein the multi-source data includes limb movement information, contact force with a rehabilitation robot, and bioelectric information during the patient's movement; based on the target feature attribute vector and the target feature attribute weight vector, determining a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated, wherein the rehabilitation patient case library includes feature attribute vectors and rehabilitation program attribute vectors for multiple rehabilitation patients; based on the feature attribute vector of the matching rehabilitation patient, the target feature attribute vector and the target feature attribute weight vector, correcting the rehabilitation program attribute vector of the matching rehabilitation patient to obtain a target rehabilitation program for the patient to be rehabilitated.

[0187] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0188] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0189] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning, characterized in that: include: Collecting multi-source data of the patient to be rehabilitated, and generating a target feature attribute vector and an attribute weight vector based on the multi-source data, wherein the multi-source data includes limb movement information, contact force with the rehabilitation robot, and bioelectric information of the patient during movement; Based on the target characteristic attribute vector and the target characteristic attribute weight vector, determining a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated, wherein the rehabilitation patient case library includes characteristic attribute vectors and rehabilitation program attribute vectors of multiple rehabilitation patients; The rehabilitation program attribute vector of the matched rehabilitation patient is modified based on the characteristic attribute vector of the matched rehabilitation patient, the target characteristic attribute vector and the target characteristic attribute weight vector to obtain the target rehabilitation program of the patient to be rehabilitated; Based on the target feature attribute vector and the target feature attribute weight vector, determining a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated specifically includes: Calculate the similarity between the characteristic attribute vector of each rehabilitation patient in the rehabilitation patient case library and the target characteristic attribute vector of the patient to be rehabilitated, determine the rehabilitation patient with the greatest similarity as the matching rehabilitation patient, and determine the matching rehabilitation patient case based on the matching rehabilitation patient; The similarity is calculated by the following formula: Among them, the rehabilitation patient case database , Middle Recovered patients The characteristic attribute vector of , is the number of characteristic attributes, i =1,2,…, n , the characteristic attribute vector of the patient to be recovered is , For the i The weight value of the feature attribute, for Middle The method comprises: determining the similarity between the characteristic attribute vector of the matched rehabilitation patient and the target characteristic attribute vector of the patient to be rehabilitated; and modifying the rehabilitation program attribute vector of the matched rehabilitation patient based on the characteristic attribute vector of the matched rehabilitation patient, the target characteristic attribute vector and the target characteristic attribute weight vector to obtain the target rehabilitation program of the patient to be rehabilitated, specifically comprising: Determining a feature attribute difference matrix based on the feature attribute vector of the matched rehabilitation patient and the target feature attribute vector; Determining a correction matrix based on the feature attribute difference matrix and the target feature attribute weight vector; Determining a target rehabilitation program attribute vector for the patient to be rehabilitated based on the correction matrix and the rehabilitation program attribute vector of the matched rehabilitation patient; The determining of the characteristic attribute difference matrix based on the characteristic attribute vector of the matched rehabilitation patient and the target characteristic attribute vector specifically includes: The difference matrix DIFF is determined by the following formula: Among them, the characteristic attribute vector of matching rehabilitation patients is , is the number of characteristic attributes, s =1,2,…, n , target feature attribute vector = To match the rehabilitation plan attribute vector of the rehabilitation patient q The content of the plan, q =1,2,…, k , k The total number of rehabilitation program contents; Correspondingly, the determining of the correction matrix based on the feature attribute difference matrix and the target feature attribute weight vector specifically includes: The correction matrix R is determined by the following formula: Among them, the target feature attribute weight vector ; The step of determining a target rehabilitation program attribute vector for the patient to be rehabilitated based on the correction matrix and the rehabilitation program attribute vector of the matched rehabilitation patient specifically includes: The target rehabilitation program attribute vector of the patient to be rehabilitated is determined by the following formula : Among them, the rehabilitation program attribute vector matching the rehabilitation patient .

2. The method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning according to claim 1, characterized in that: The limb movement information includes the movement position, speed and acceleration of the lower limb hip, the movement position, speed and acceleration of the knee, and the movement position, speed and acceleration of the ankle joint; the contact force includes the contact force between the end pedal of the rehabilitation robot and the contact surface of the sole of the foot; the bioelectric information includes electrocardiogram signals and surface electromyography signals of the lower limbs.

3. The method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning according to claim 1, characterized in that: The target feature attribute weight vector is generated by manually evaluating each feature attribute in the target feature attribute vector.

4. A lower limb rehabilitation program design device using multi-source information and case-based reasoning, used to implement a lower limb rehabilitation program design method using multi-source information and case-based reasoning as described in any one of claims 1 to 3, characterized in that: include: A feature unit, used to collect multi-source data of the patient to be rehabilitated, and generate a target feature attribute vector and an attribute weight vector based on the multi-source data, wherein the multi-source data includes limb movement information, contact force with the rehabilitation robot, and bioelectric information of the patient during movement; A matching unit, configured to determine, based on the target feature attribute vector and the target feature attribute weight vector, a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated, wherein the rehabilitation patient case library includes feature attribute vectors and rehabilitation program attribute vectors of a plurality of rehabilitation patients; Based on the target feature attribute vector and the target feature attribute weight vector, determining a matching rehabilitation patient case in a rehabilitation patient case library that matches the patient to be rehabilitated specifically includes: Calculate the similarity between the characteristic attribute vector of each rehabilitation patient in the rehabilitation patient case library and the target characteristic attribute vector of the patient to be rehabilitated, determine the rehabilitation patient with the greatest similarity as the matching rehabilitation patient, and determine the matching rehabilitation patient case based on the matching rehabilitation patient; The similarity is calculated by the following formula: Among them, the rehabilitation patient case database , Middle Recovered patients The characteristic attribute vector of , is the number of characteristic attributes, i =1,2,…, n , the characteristic attribute vector of the patient to be recovered is , For the i The weight value of the feature attribute, Middle The similarity between the characteristic attribute vector of the rehabilitation patient and the target characteristic attribute vector of the patient to be rehabilitated; A correction unit, configured to correct the rehabilitation program attribute vector of the matched rehabilitation patient based on the characteristic attribute vector of the matched rehabilitation patient, the target characteristic attribute vector and the target characteristic attribute weight vector, so as to obtain a target rehabilitation program for the patient to be rehabilitated; The step of correcting the rehabilitation program attribute vector of the matched rehabilitation patient based on the characteristic attribute vector of the matched rehabilitation patient, the target characteristic attribute vector and the target characteristic attribute weight vector to obtain the target rehabilitation program for the patient to be rehabilitated specifically includes: Determining a feature attribute difference matrix based on the feature attribute vector of the matched rehabilitation patient and the target feature attribute vector; Determining a correction matrix based on the feature attribute difference matrix and the target feature attribute weight vector; Determining a target rehabilitation program attribute vector for the patient to be rehabilitated based on the correction matrix and the rehabilitation program attribute vector of the matched rehabilitation patient; The determining of the characteristic attribute difference matrix based on the characteristic attribute vector of the matched rehabilitation patient and the target characteristic attribute vector specifically includes: The difference matrix DIFF is determined by the following formula: Among them, the characteristic attribute vector of matching rehabilitation patients is , is the number of characteristic attributes, s =1,2,…, n , target feature attribute vector = To match the rehabilitation plan attribute vector of the rehabilitation patient q The content of the plan, q =1,2,…, k , k The total number of rehabilitation program contents; Correspondingly, the determining of the correction matrix based on the feature attribute difference matrix and the target feature attribute weight vector specifically includes: The correction matrix R is determined by the following formula: Among them, the target feature attribute weight vector ; The step of determining a target rehabilitation program attribute vector for the patient to be rehabilitated based on the correction matrix and the rehabilitation program attribute vector of the matched rehabilitation patient specifically includes: The target rehabilitation program attribute vector of the patient to be rehabilitated is determined by the following formula : Among them, the rehabilitation program attribute vector matching the rehabilitation patient .

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning as described in any one of claims 1 to 3 are implemented.

6. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for designing a lower limb rehabilitation program using multi-source information and case-based reasoning as described in any one of claims 1 to 3 are implemented.

Images