CN115337003A - Multi-dimensional lower limb rehabilitation evaluation device and method for stroke patient - Google Patents

Abstract

A multi-dimensional lower limb rehabilitation evaluation device for stroke patients is composed of an upper computer module, a main controller module, an inertial sensor module, a pressure sensor module and a communication module; in the evaluation process, the evaluation method is divided into three layers of joint activity, gait track and gait phase, and the weight is determined by an analytic hierarchy process. The joint mobility is evaluated through a maximum and minimum joint mobility algorithm; the evaluation of the gait track is realized by a cosine similarity algorithm; realizing the identification of gait phase by a proportional fuzzy logic split-phase method so as to evaluate; the device has the advantages of simple structure, low price, convenience in carrying, strong universality of the evaluation method, high precision and easiness in implementation.

Description

A multi-dimensional lower limb rehabilitation evaluation device and evaluation method for stroke patients

technical field

The invention belongs to the field of rehabilitation robot technology and biological feature recognition technology, and in particular relates to a multi-dimensional lower limb rehabilitation evaluation device and evaluation method for stroke patients.

Background technique

The effect of rehabilitation training is often reflected through the rehabilitation evaluation system. Rehabilitation evaluation is the process of accurately judging the disorder and forming a diagnostic conclusion by collecting and analyzing various data of patients. Compared with clinical diagnosis, rehabilitation assessment focuses on function, which is a process of making qualitative and quantitative judgments on dysfunction. The scientific and comprehensive rehabilitation assessment method can guide rehabilitation training methods, accelerate the recovery of limb functions, and greatly reduce the recovery time. Traditional rehabilitation evaluation methods mostly use artificial joint range of motion measurement, muscle strength detection, quantitative table scoring and other methods, which may lead to problems such as large measurement errors and too subjective evaluation methods. Secondly, the evaluation method is too single, and the equipment is expensive. The portability is poor, and there are limitations in the applicable population. In order to solve the above problems, the present invention proposes a multi-dimensional lower limb rehabilitation evaluation device and evaluation method. The evaluation method is composed of three levels of joint mobility, gait trajectory and gait phase. The weights of the three levels are quantified by the analytic hierarchy process. Data is collected by inertial sensors and thin-film pressure sensors. The above method can effectively improve the accuracy and scientificity of rehabilitation assessment.

Invention content:

The purpose of the present invention is to provide a multi-dimensional lower limb rehabilitation evaluation device and evaluation method for stroke patients, which can solve the problems of traditional lower limb rehabilitation evaluation, and is a portable device with small equipment, low price and easy carrying, and The method is capable of collecting and processing data on joint angles of lower extremities, gait trajectory and plantar pressure, and determining and scoring weights through analytic hierarchy process. This method is simple and easy to implement.

The technical solution of the present invention: a multi-dimensional lower limb rehabilitation evaluation device for stroke patients, characterized in that it includes a host computer module, a main controller module, an inertial sensor module, a pressure sensor module and a communication module; wherein, the The input end of the main controller module is connected with the inertial sensor module and the pressure sensor module for the collection and transmission of joint and plantar data, and the output end of the main controller module is connected with the input end of the communication module; the communication module is used for Set up the communication between the main controller module and the upper computer module, the input end of the communication module is connected with the output end of the main controller module, and its output end is connected with the input end of the upper computer module; the inertial sensor module is used for collecting joint angles and gait trajectory information; the pressure sensor module is used to collect plantar pressure information.

The pressure sensor module is a module composed of a thin film pressure sensor and is installed in the insole.

The quantity of described thin-film pressure sensor is 8 pressure sensors; Described 8 thin-film pressure sensors are connected with main controller module through 8 analog-to-digital converters respectively; Described 8 thin-film pressure sensors are respectively installed in two insoles, There are 4 in one insole, three in the forefoot and one in the heel.

The positions of the four thin-film pressure sensors in the insole correspond to the heel, the fourth metatarsal, the first metatarsal and the thumb respectively, and the pressure signals at the positions of the four thin-film pressure sensors are obtained respectively.

The inertial sensor module is composed of 4 inertial sensors; the 4 inertial sensors are connected to the main controller module; the 4 inertial sensors are respectively placed at the thighs and shanks of the two legs.

The communication module is one of WiFi, 5G, 4G or serial communication.

A multi-dimensional lower limb rehabilitation assessment method for stroke patients, characterized in that it comprises the following steps:

(1) Use the inertial sensor to obtain the joint angle signal, use the signal to obtain and record the patient's joint flexion and extension angle, obtain the patient's joint flexion and extension angle, and compare them with the medical standard joint angle to obtain the proportion; The proportion is used as the basis for the evaluation of joint range of motion;

The step (1) is to use the maximum and minimum joint degree algorithm to process the joint angle signal obtained by the inertial sensor, that is, to obtain the patient's hip joint flexion angle A ₁ , the patient's hip joint extension angle B ₁ , and the patient's knee joint flexion angle C ₁ , the patient's knee joint extension angle D ₁ , and calculate the ratios to the standard hip joint flexion angle A ₁₁ , standard hip joint extension angle B ₁₁ , standard knee joint flexion angle C ₁₁ , and standard knee joint extension angle D ₁₁ , as As shown in Table 1; mid-term, ω _A , ω _B , ω _C , and ω _D are the weights of the total 4 index scores of the knee and hip joints of the patient's legs during flexion and extension, respectively;

Table 1

Then, according to the formula (1), the score R _S of the range of motion of the joint can be obtained;

(2) Use the inertial sensor to obtain the gait trajectory signal, record the trajectory of the patient's knee and hip joints, and use the similarity between the trajectory of the patient's knee and hip joints and the standard knee and hip joints as the gait The evaluation basis of the state trajectory;

The step (2) adopts the cosine similarity algorithm to process the gait track signal acquired by the inertial sensor, obtains and records the motion track of the patient's knee joint and hip joint, that is: select the patient's hip joint motion track and the standard hip joint motion track The same number of points form two vectors a1 and vector b1, select the same number of points of the patient’s knee joint motion trajectory and the standard knee joint motion trajectory to form two vectors a2 and vector b2, and calculate the cosine angles of the four vectors respectively, Such as formula (2), formula (3), and this cosine angle is used as the evaluation basis of gait track;

Then, the score of gait trajectory is formula (4)

T _S ＝(cosθ ₁ ×w ₁ +cosθ ₂ ×w ₂ )×100 (4)

In the formula (4), w ₁ =w ₂ =0.5 can be taken, w ₁ is the weight of the hip joint movement trajectory score to the gait trajectory score, and w ₂ is the weight of the knee joint movement trajectory score to the gait trajectory score.

(3) Use the pressure sensor to obtain plantar pressure signals to identify the gait phase of the patient, record the sequence of bilateral gait phases, obtain the number of abnormal gait cycles and the total number of gait cycles, and calculate the abnormal gait cycle The ratio of the number to the total number of gait cycles is used as the basis for evaluating the gait phase;

Described step (3) adopts proportional fuzzy logic phase separation method to carry out the recognition of gait phase, record bilateral gait phase order, according to gait phase recognition result, to standing early phase, standing mid-term phase, standing late phase and swing phase The time interval of the phase is obtained to obtain the number of abnormal gait cycles and the total number of gait cycles, so as to obtain the ratio of the number of abnormal gait cycles to the total number of gait cycles, as shown in Figure 4, which specifically consists of the following steps:

(3-1) Carry out proportional fusion processing of the plantar pressure signal, that is, perform summation and proportional operation on the plantar pressure signal at the same time to solve the recognition error caused by individual differences, such as formulas (5)-(8 ) as shown:

Among them, K ₁ , K ₂ , K ₃ , and K ₄ are the _{ratios of} the signals of the _four pressure sensors to the sum of the plantar pressure signals _; under pressure;

(3-2) Perform fuzzy processing on the proportionally fused data obtained in step (3-1), and determine an appropriate membership function. Since the four pressure sensor signals after step (3-1) are proportionally fused The proportional data of the sum of plantar pressure signals, that is, K ₁ , K ₂ , K ₃ , and K ₄ , are all in the range of 0-1. In order to eliminate the sharp boundary of the traditional gait stage division, an exponential function is selected as Membership function, as shown in formula (9):

Among them, f(K _i ) is the output of the function, and K _i is the ratio of the four pressure sensor signals obtained in step (3-1) to the sum of plantar pressure signals K ₁ , K ₂ , K ₃ , K ₄ , K _0i is the ratio threshold, s is the sensitivity coefficient;

(3-3) Since the membership function shown in formula (9) in step (3-2) is a symmetric function, therefore, in order to describe the degree to which the proportional value _Ki is close to 1 or close to 0, step (3 -2) is the inverse function of formula (9), and f(K _i ) is divided into f ^H (K _i ) and f ^L (K _i ) two parts, f ^H (K _i ) represents the proportional value K _i close to 1, f ^L (K _i ) represents the degree to which the proportional value K _i is close to 0, as shown in formula (10) and formula (11) respectively;

(3-4) Based on the function output of the formula (10) and the formula (11) in the step (3-3), the fuzzy logic rule table is set, as shown in Table 2, the medical standard divides the gait into six phases, Initial contact, load response, mid-stance, late-stance, anticipatory swing, and post-swing, respectively, due to initial contact and load response, with low mechanical discrimination between late-stance and post-swing. Therefore, they are divided into one category, called the early stance and the late stance. When the proportional fuzzy logic phase separation method is used to identify the gait phase, the gait is re-divided into four phases, namely: early stance, mid-stance, and late stance , swing period;

Table 2

Heel K₁ Fourth metatarsal K₂ First metatarsal K₃ Thumb K₄ Pre-stand AS h * L L Standing Mid BS * h h L Standing Late CS L * * h swing period DS L L L L

From Table 2, it can be obtained that the phasing formulas of the early stage of stance, the middle stage of stance, the late stage of stance and the swing stage as shown in formulas (12)-(15) are the phasing results of the patient's gait phase;

AS＝f ^H (K ₁ )×f ^L (K ₃ )×f ^L (K ₄ ) (12)

BS＝f ^H (K ₂ )×f ^H (K ₃ )×f ^L (K ₄ ) (13)

CS＝f ^L (K ₁ )×f ^H (K ₄ ) (14)

DS＝f ^L (K ₁ )×f ^L (K ₂ )×f ^L (K ₃ )×f ^L (K ₄ ) (15)

Among them, when K _i >K _0i , the fuzzy set is in a "big" state, marked as "H", when K _i <K _0i , the fuzzy set is in a "small" state, marked as "L", if When there is no state, it is recorded as "*";

(3-5) Based on the output of formulas (12)-(15) in step (3-4), calculate the duration of the gait phase in a gait cycle: a gait cycle includes the early stage of standing, the middle stage of standing, and the period of standing late stage and swing phase, and the order of gait phase is early stance -> mid stance -> late stance -> swing phase; the gait phase duration is compared with the medical standard gait phase time, and in order to consider individual differences, the The medical standard gait phase duration is divided into pre-stance AS which accounts for 8%-12% of a gait cycle, mid-stance BS which accounts for 18%-22% of a gait cycle, and late-stance CS which accounts for 28%-28% of a gait cycle. 32%, the swing phase DS accounts for 38-42% of a gait cycle; finally judge whether the gait phase duration exceeds the range specified by the medical standard gait phase duration, if it exceeds, it is an abnormal gait cycle, if it does not exceed, then is the normal gait cycle; the ratio of the number of abnormal gait cycles to the total number of gait cycles is used as the evaluation basis for the gait phase, the total number of gait cycles is the sum of the number of abnormal gait cycles and the number of normal gait cycles, abnormal The number of gait cycles is the sum of abnormal gait cycles, and the normal gait cycle is the sum of normal gait cycles; at this time, the gait phase score is shown in formula (16):

Among them, PS is the gait phase score, _{A n} is the number of abnormal gait cycles, and T _n is the total number of gait cycles.

(4) Since the range of motion, gait trajectory and gait phase play different roles in the multi-dimensional lower limb rehabilitation evaluation, in order to obtain scientific and accurate rehabilitation evaluation results, it is necessary to comprehensively evaluate the three, and use the analytic hierarchy process to determine The weights of the joint activity evaluation basis obtained in step (1), the gait trajectory evaluation basis obtained in step (2), and the gait phase evaluation basis obtained in step (3):

(4-1) According to the analysis method of the AHP, the AHP is divided into three levels, namely: the target layer, the criterion layer and the program layer; wherein, the target layer is the degree of recovery of lower limb joint function; the The criterion layer is the basis for evaluating the range of motion obtained in step (1), the basis for evaluating the gait trajectory obtained in step (2), and the basis for evaluating the gait phase obtained in step (3); the scheme layer is obtained in step (1). The evaluation basis of joint range of motion, the gait trajectory evaluation basis obtained in step (2) and the weight coefficient occupied by the gait phase evaluation basis obtained in step (3);

(4-2) and use the "1-9 scale method" to quantify the proportion of joint range of motion, gait trajectory, and plantar pressure, and use the evaluation basis and steps of joint range of motion obtained in step (1) in the criterion layer (2) The gait trajectory evaluation basis obtained and the gait phase evaluation basis obtained in step (3) are compared in pairs to form a paired comparison table. The title items of the rows and columns of the table are evaluation indicators, and the order consistent;

table 3

range of motion Gait track gait phase range of motion a₁₁ a₁₂ a₁₃ Gait track a₂₁ a₂₂ a₂₃ gait phase a₃₁ a₃₂ a₃₃

(4-4) The table of three rows and three columns obtained by step (4-2) can obtain the pairwise comparison matrix shown in formula (17), that is: judgment matrix A:

Each element in the judgment matrix A corresponds to the data in each cell in the table with three rows and three columns;

(4-5) Normalize the column vectors of the judgment matrix A, and at the same time sum and normalize the rows, then formula (18) can be obtained:

Among them, w ₁ , w ₂ , w ₃ joint mobility weights, gait trajectory weights and gait phase weights;

(4-6) Verify the consistency of formula (18), and get formula (19):

Because Aω＝λω, then:

where n is the matrix order;

(4-7) Random consistency index RI, as shown in Table 4:

Table 4

no 1 2 3 4 5 6 7 8 9 10 11 RI 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 1.51

Further, the inconsistency index CI and the consistency ratio CR can be obtained, as shown in formula (21) and formula (22):

If the consistency ratio CR<0.1, it is considered that the degree of inconsistency in Table 3 is within the allowable range, there is satisfactory consistency, and the consistency test is passed, and the normalized feature vector ω of formula (18) can be used as the weight vector; if If the consistency ratio does not meet the requirements, it is necessary to construct a pairwise comparison matrix according to steps (4-2) and (4-3) to construct a judgment matrix, and then use steps (4-5) to (4-7) to check the consistency Verify, and finally obtain the weight of each level.

The numerical value of Table 3 in the described step (4) is the official numerical value that " 1-9 scale method " provides, and what numerical value uses is that the 1-9 scale method of Santy provides, is public knowledge. In the embodiment, the numerical content of Table 3 is determined by the rehabilitation physician with reference to Santy's 1-9 scale method based on experience, and the specific numerical values shown in Table 3-1 are obtained;

Table 3-1

range of motion Gait trajectory gait phase range of motion 1 1/3 1/4 Gait trajectory 3 1 1/2 gait phase 4 2 1

In the table, "Number 1" means that the two evaluation bases of row and column are equally important, "Number 3" means that row is slightly more important than column, "Number 5" means that row is significantly more important than column, and "Number 2" means that row is more important than column. The degree of importance is between "Number 1" and "Number 3", "Number 4" means that the importance of the row ratio is between "Number 3" and "Number 5", and the score part is the importance of the column ratio than the row , such as "Number 1/3" indicates that the column is slightly more important than the row;

then further get:

Normalize the column vector of the judgment matrix A, and sum the rows and normalize at the same time, you can get

That is: the weight of joint mobility is 0.123, the weight of gait trajectory is 0.32, and the weight of gait phase is 0.557;

Perform consistency verification to get:

Because Aω＝λω, then:

where n is the matrix order;

Further, the inconsistency index CI and the consistency ratio CR can be obtained:

Therefore, the degree of inconsistency in Table 3-1 is within the allowable range, and there is satisfactory consistency. After passing the consistency test, the normalized feature vector ω obtained at this time, that is, the weight of joint mobility is 0.123, and the weight of gait trajectory is 0.32, and the weight of gait phase is 0.557, which can be used as the weight of the evaluation index.

(5) According to the weight of the evaluation index determined in step (4), combined with the score R _S of the joint range of motion in step (1), the score T _S of the gait track in step (2), and the gait phase of step (3) Score P _S , the comprehensive rehabilitation evaluation score can be obtained, as shown in formula (23)

Z _S ＝w ₁ ×R _S +w ₂ ×T _S +w ₃ ×P _s (23)

Where Z _s is the comprehensive rehabilitation evaluation score;

(6) According to the comprehensive rehabilitation evaluation score obtained by formula (23), evaluate the rehabilitation effect of the patient: it is stipulated that the rehabilitation effect above 90 points is excellent, the rehabilitation effect between 80-90 points is good, and the rehabilitation effect between 70-80 points It is average, and the rehabilitation effect below 60 points is poor; when the comprehensive rehabilitation evaluation score of the patient is below 60 points, the conclusion of "need to strengthen the intensity of rehabilitation training" is given, and the training method at this stage should tend to be passive rehabilitation training; When the patient's comprehensive rehabilitation evaluation score is between 70-80 points, the conclusion that "rehabilitation training intensity can be appropriately reduced" is given, and the active training method is appropriately increased on the basis of passive training methods; when the patient's comprehensive rehabilitation evaluation score is between 80-90 When the score is between, give the conclusion that "the active training method can be used as the main training method", and appropriately increase the confrontational training method; when the patient's comprehensive rehabilitation evaluation score is above 90 points, give the conclusion that "the confrontational training method can be used Recommendations as a primary training modality".

The superiority of the present invention: the multi-dimensional lower limb rehabilitation evaluation method and device, through the analysis of the evaluation method composed of three levels, respectively use the maximum and minimum joint degree algorithm, cosine similarity algorithm, proportional fuzzy logic phasing method to evaluate the patient's lower limbs Joint angle, gait trajectory and plantar pressure are processed, and the weight of the three-level evaluation method of joint mobility, gait trajectory and gait phase is determined through the analytic hierarchy process; it solves the problems caused by individual differences Gait phase error, expensive equipment, poor portability and other issues have high evaluation accuracy and accurate evaluation results, which have high research significance.

Description of drawings

FIG. 1 is a schematic diagram of the overall structure of a multi-dimensional lower limb rehabilitation assessment device for stroke patients according to the present invention.

Fig. 2 is a schematic diagram of the installation position distribution of the thin-film pressure sensor and the inertial sensor in a multi-dimensional lower limb rehabilitation evaluation device for stroke patients according to the present invention.

Fig. 3 is a flowchart of a multi-dimensional lower limb rehabilitation assessment method for stroke patients according to the present invention.

FIG. 4 is a schematic flow chart of a proportional fuzzy logic phase separation method in a multi-dimensional lower limb rehabilitation assessment method for stroke patients according to the present invention.

Detailed ways

Embodiment: The embodiment of the present invention aims at the problem of lower limb rehabilitation evaluation of stroke patients, and designs a multi-dimensional lower limb rehabilitation evaluation method and device.

Figure 1 is a schematic diagram of the overall structure of a multi-dimensional lower limb rehabilitation evaluation device for stroke patients according to the present invention. First, the acquisition system consists of a main controller module, a pressure sensor module, and the pressure sensor module includes 8 membrane pressure sensors. Sensor, inertial sensor module, the inertial sensor module consists of 4 inertial sensors, 1 communication module, and 1 pair of insoles. The communication module is not limited to Bluetooth, WiFi, 5G, 4G, serial port, etc. The thin-film pressure sensor needs to be converted by the ADC and then transmitted to the main controller module. The main controller module converts the digital signal converted by the ADC into a pressure value, and transmits it to the host computer module for processing through the communication module. The inertial sensor transmits the data after sending the zero-degree calibration command through the main controller module, and finally transmits the data to the upper computer module through the communication module through the main controller module for processing. Inertial sensor data includes joint angle and joint trajectory data. Joint angle, joint trajectory data and pressure data processing are all carried out in the host computer module, which includes the maximum and minimum joint degree algorithm, cosine similarity algorithm, proportional fuzzy logic phase separation method and comprehensive rehabilitation evaluation.

Fig. 2 is a schematic diagram of the installation position distribution of a thin film pressure sensor and an inertial sensor in a multi-dimensional lower limb rehabilitation assessment device for stroke patients according to the present invention. The inertial sensor 1 is located at the thigh and is used to collect the range of motion of the hip joint and the Joint track, the inertial sensor 2 is located at the calf, and is used to collect the range of motion of the knee joint and the track of the hip joint, and the inertial sensor fixes the thigh and calf of the patient through an adhesive tape. 8 thin-film pressure sensors are installed in two insoles, 4 in one insole, three of them are at the forefoot and the other is at the heel, pressure sensor 4, pressure sensor 3, pressure sensor 2, pressure sensor 1, Corresponding to the heel, the fourth metatarsal, the first metatarsal and the thumb respectively, the pressure sensor is fixed on the insole by superglue, and the insole is placed in the patient's shoe.

Fig. 3 is a flowchart of a multi-dimensional lower limb rehabilitation assessment method for stroke patients according to the present invention. First, the main controller module is controlled by the host computer module to realize the collection of joint angles through inertial sensors, and the collected data is transmitted to the communication module, and then transmitted to the host computer module by the communication module, and processed by the maximum and minimum joint degree algorithm in the host computer module , to obtain the evaluation result of the range of motion of the joint, the content of the evaluation of the range of motion of the joint, the specific steps are as follows, in order to ensure the output accuracy of the inertial sensor, when initializing the inertial sensor, it is necessary to send a zero-degree calibration command with the current position as the zero angle to the inertial sensor:

(1) Use the inertial sensor to obtain the joint angle signal, use the signal to obtain and record the patient's joint flexion and extension angle, obtain the patient's joint flexion and extension angle, and compare them with the medical standard joint angle to obtain the proportion; The proportion is used as the basis for the evaluation of joint range of motion;

The step (1) is to use the maximum and minimum joint degree algorithm to process the joint angle signal obtained by the inertial sensor, that is, to obtain the patient's hip joint flexion angle A ₁ , the patient's hip joint extension angle B ₁ , and the patient's knee joint flexion angle C ₁ , the patient's knee joint extension angle D ₁ , and calculate the ratios to the standard hip joint flexion angle A ₁₁ , standard hip joint extension angle B ₁₁ , standard knee joint flexion angle C ₁₁ , and standard knee joint extension angle D ₁₁ , as As shown in Table 1; mid-term, ω _A , ω _B , ω _C , and ω _D are the weights of the total 4 index scores of the knee and hip joints of the patient's legs during flexion and extension, respectively;

Table 1

Then, according to the formula (1), the score R _S of the range of motion of the joint can be obtained;

Secondly, the main controller module is controlled by the upper computer module to realize the collection of joint trajectory through the inertial sensor, and the collected data is transmitted to the communication module, and then transmitted to the upper computer module by the communication module, and processed by the cosine similarity algorithm in the upper computer module , to obtain the evaluation result of gait trajectory, the specific evaluation content of gait trajectory is as follows:

(2) Use the inertial sensor to obtain the gait trajectory signal, record the trajectory of the patient's knee and hip joints, and use the similarity between the trajectory of the patient's knee and hip joints and the standard knee and hip joints as the gait The evaluation basis of the state trajectory;

The step (2) adopts the cosine similarity algorithm to process the gait track signal acquired by the inertial sensor, obtains and records the motion track of the patient's knee joint and hip joint, that is: select the patient's hip joint motion track and the standard hip joint motion track The same number of points form two vectors a1 and vector b1, select the same number of points of the patient’s knee joint motion trajectory and the standard knee joint motion trajectory to form two vectors a2 and vector b2, and calculate the cosine angles of the four vectors respectively, Such as formula (2), formula (3), and this cosine angle is used as the evaluation basis of gait trajectory;

Then, the score of gait trajectory is formula (4)

T _S ＝(cosθ ₁ ×w ₁ +cosθ ₂ ×w ₂ )×100 (4)

In the formula (4), w ₁ =w ₂ =0.5 can be taken, w ₁ is the weight of the hip joint movement trajectory score to the gait trajectory score, and w ₂ is the weight of the knee joint movement trajectory score to the gait trajectory score.

Secondly, the main controller module is controlled by the host computer module to realize the collection of plantar pressure through the thin film pressure sensor, and the collected data is transmitted to the communication module, and then transmitted to the host computer module by the communication module. In the host computer module and through proportional fuzzy logic The gait phase evaluation results are obtained through the phase separation method. The specific gait phase evaluation content is as follows:

(3) Use the pressure sensor to obtain plantar pressure signals to identify the gait phase of the patient, record the sequence of bilateral gait phases, obtain the number of abnormal gait cycles and the total number of gait cycles, and calculate the abnormal gait cycle The ratio of the number to the total number of gait cycles is used as the basis for evaluating the gait phase;

Described step (3) adopts proportional fuzzy logic phase separation method to carry out the recognition of gait phase, record bilateral gait phase order, according to gait phase recognition result, to standing early phase, standing mid-term phase, standing late phase and swing phase The time interval of the phase is obtained to obtain the number of abnormal gait cycles and the total number of gait cycles, so as to obtain the ratio of the number of abnormal gait cycles to the total number of gait cycles, as shown in Figure 4, which specifically consists of the following steps:

(3-1) Carry out proportional fusion processing of the plantar pressure signal, that is, perform summation and proportional operation on the plantar pressure signal at the same time to solve the recognition error caused by individual differences, such as formulas (5)-(8 ) as shown:

Among them, K ₁ , K ₂ , K ₃ , and K ₄ are the _{ratios of} the signals of the _four pressure sensors to the sum of the plantar pressure signals _; under pressure;

(3-2) Perform fuzzy processing on the proportionally fused data obtained in step (3-1), and determine an appropriate membership function. Since the four pressure sensor signals after step (3-1) are proportionally fused The proportional data of the sum of plantar pressure signals, that is, K ₁ , K ₂ , K ₃ , and K ₄ , are all in the range of 0-1. In order to eliminate the sharp boundary of the traditional gait stage division, an exponential function is selected as Membership function, as shown in formula (9):

Among them, f(K _i ) is the output of the function, and K _i is the ratio of the four pressure sensor signals obtained in step (3-1) to the sum of plantar pressure signals K ₁ , K ₂ , K ₃ , K ₄ , K _0i is the ratio threshold, s is the sensitivity coefficient;

(3-3) Since the membership function shown in formula (9) in step (3-2) is a symmetric function, therefore, in order to describe the degree to which the proportional value _Ki is close to 1 or close to 0, step (3 -2) is the inverse function of formula (9), and f(K _i ) is divided into f ^H (K _i ) and f ^L (K _i ) two parts, f ^H (K _i ) represents the proportional value K _i close to 1, f ^L (K _i ) represents the degree to which the proportional value K _i is close to 0, as shown in formula (10) and formula (11) respectively;

(3-4) Based on the function output of the formula (10) and the formula (11) in the step (3-3), the fuzzy logic rule table is set, as shown in Table 2, the medical standard divides the gait into six phases, Initial contact, load response, mid-stance, late-stance, anticipatory swing, and post-swing, respectively, due to initial contact and load response, with low mechanical discrimination between late-stance and post-swing. Therefore, they are divided into one category, called the early stance and the late stance. When the proportional fuzzy logic phase separation method is used to identify the gait phase, the gait is re-divided into four phases, namely: early stance, mid-stance, and late stance , swing period;

Table 2

Heel K₁ Fourth metatarsal K₂ First metatarsal K₃ Thumb K₄ Pre-stand AS h * L L Standing Mid BS * h h L Standing Late CS L * * h swing period DS L L L L

From Table 2, it can be obtained that the phasing formulas of the early stage of stance, the middle stage of stance, the late stage of stance and the swing stage as shown in formulas (12)-(15) are the phasing results of the patient's gait phase;

AS＝f ^H (K ₁ )×f ^L (K ₃ )×f ^L (K ₄ ) (12)

BS＝f ^H (K ₂ )×f ^H (K ₃ )×f ^L (K ₄ ) (13)

CS＝f ^L (K ₁ )×f ^H (K ₄ ) (14)

DS＝f ^L (K ₁ )×f ^L (K ₂ )×f ^L (K ₃ )×f ^L (K ₄ ) (15)

Among them, when K _i >K _0i , the fuzzy set is in a "big" state, marked as "H", when K _i <K _0i , the fuzzy set is in a "small" state, marked as "L", if When there is no state, it is recorded as "*";

(3-5) Based on the output of formulas (12)-(15) in step (3-4), calculate the duration of the gait phase in a gait cycle: a gait cycle includes the early stage of standing, the middle stage of standing, and the period of standing late stage and swing phase, and the order of gait phase is early stance -> mid stance -> late stance -> swing phase; the gait phase duration is compared with the medical standard gait phase time, and in order to consider individual differences, the The gait phase duration of medical standards is divided into 8%-12% (10% in the embodiment) of a gait cycle in the early stage of standing, and 18%-22% (in the embodiment) of a gait cycle in the mid-stand BS. Get 20%), stand late stage CS accounts for a gait cycle 28%-32% (get 30% in the embodiment), swing phase DS accounts for a gait cycle 38-42% (get 40% in the embodiment); Finally pass Determine whether the gait phase duration exceeds the range specified by the medical standard for gait phase duration. If it exceeds, it is an abnormal gait cycle, and if it does not exceed it, it is a normal gait cycle; The proportion of the gait phase is used as the evaluation basis for the gait phase. The total number of gait cycles is the sum of the number of abnormal gait cycles and the number of normal gait cycles, the number of abnormal gait cycles is the sum of abnormal gait cycles, and the normal gait cycle is normal The sum of gait cycle numbers; at this time, the gait phase score is shown in formula (16):

Among them, PS is the gait phase score, _{A n} is the number of abnormal gait cycles, and T _n is the total number of gait cycles.

Finally, the analytic hierarchy process is implemented to determine the weight of each assessment level:

(4) Since the range of motion, gait trajectory and gait phase play different roles in the multi-dimensional lower limb rehabilitation evaluation, in order to obtain scientific and accurate rehabilitation evaluation results, it is necessary to comprehensively evaluate the three, and use the analytic hierarchy process to determine The weights of the joint activity evaluation basis obtained in step (1), the gait trajectory evaluation basis obtained in step (2), and the gait phase evaluation basis obtained in step (3):

(4-1) According to the analysis method of the AHP, the AHP is divided into three levels, namely: the target layer, the criterion layer and the program layer; wherein, the target layer is the degree of recovery of lower limb joint function; the The criterion layer is the basis for evaluating the range of motion obtained in step (1), the basis for evaluating the gait trajectory obtained in step (2), and the basis for evaluating the gait phase obtained in step (3); the scheme layer is obtained in step (1). The evaluation basis of joint range of motion, the gait trajectory evaluation basis obtained in step (2) and the weight coefficient occupied by the gait phase evaluation basis obtained in step (3);

(4-2) and use the "1-9 scale method" to quantify the proportion of joint range of motion, gait trajectory, and plantar pressure, and use the evaluation basis and steps of joint range of motion obtained in step (1) in the criterion layer (2) The gait trajectory evaluation basis obtained and the gait phase evaluation basis obtained in step (3) are compared in pairs to form a paired comparison table, as shown in Table 3-1. The title items of the rows and columns of the table All are evaluation indicators, and the sequence is the same;

Table 3-1

range of motion Gait track gait phase range of motion 1 1/3 1/4 Gait track 3 1 1/2 gait phase 4 2 1

In the table, "Number 1" means that the two evaluation bases of row and column are equally important, "Number 3" means that row is slightly more important than column, "Number 5" means that row is significantly more important than column, and "Number 2" means that row is more important than column. The degree of importance is between "Number 1" and "Number 3", "Number 4" means that the importance of the row ratio is between "Number 3" and "Number 5", and the score part is the importance of the column ratio than the row , such as "Number 1/3" indicates that the column is slightly more important than the row;

Under the guidance of rehabilitation physicians, the comparison matrix shown in formula (24) can be obtained, namely: judgment matrix A:

Normalize the column vectors of the judgment matrix A, and at the same time sum and normalize the rows, the formula (25) can be obtained:

Verify the consistency of formula (18), and get formula (26):

Because Aω＝λω, then:

where n is the matrix order;

Random consistency index RI, as shown in Table 4:

Table 4

no 1 2 3 4 5 6 7 8 9 10 11 RI 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 1.51

Through the random consistency index RI shown in Table 4, the inconsistency index CI and the consistency ratio CR can be further obtained, as shown in formula (28) and formula (29):

Since the consistency ratio CR<0.1, it is considered that the degree of inconsistency of the judgment matrix A is within the allowable range, there is satisfactory consistency, and the consistency test is passed, and the normalized eigenvector ω of formula (25) can be used as the weight vector; finally The obtained joint activity weight is 0.123, the gait trajectory weight is 0.32, and the gait phase weight is 0.557. Finally, the comprehensive rehabilitation evaluation score can be deduced as formula (30):

Z _S ＝0.123×R _S +0.32×T _S +0.557×P _s (30)

After the patient obtains the comprehensive rehabilitation evaluation score, he can refer to the rehabilitation training method corresponding to the score range of step (5) in the technical solution to adjust the rehabilitation training content. to speed up recovery.

The specific steps of patient operation are:

(1) According to Figure 2, the patient wears the evaluation device.

(2) Through the control of the host computer module to the main controller module to evaluate the range of motion of the joints, the patient performs the extension and flexion of the hip and knee joints, and the extension and flexion angles must reach their own limits during exercise to improve Accuracy of assessment.

(3) The gait trajectory is evaluated through the control of the host computer module to the main controller module, and the patient's gait walks normally during rehabilitation training. In order to ensure the accuracy of the evaluation, try to avoid turning, avoiding obstacles and other actions during walking.

(4) The gait phase is evaluated through the control of the host computer module to the main controller module, and the patient still performs the same action as step (3).

(5) After the evaluation is completed, the score of the comprehensive rehabilitation evaluation can be obtained through the prompt of the host computer module. According to the score, the patient can refer to the rehabilitation training method corresponding to the score range of step (5) in the technical plan to carry out the rehabilitation training content. Adjustment.

Although the embodiments and drawings of the present invention are disclosed for the purpose of illustration, those skilled in the art can understand that various replacements, changes and modifications are possible without departing from the scope of the present invention and the appended claims. Therefore, the scope of the present invention is not limited to what is disclosed in the embodiments and drawings.

Claims (10)

1. A multi-dimensional lower limb rehabilitation evaluation device for stroke patients is characterized by comprising an upper computer module, a main controller module, an inertial sensor module, a pressure sensor module and a communication module; the input end of the main controller module is connected with the inertial sensor module and the pressure sensor module and used for collecting and transmitting data of joints and soles, and the output end of the main controller module is connected with the input end of the communication module; the communication module is used for establishing communication between the main controller module and the upper computer module, the input end of the communication module is connected with the output end of the main controller module, and the output end of the communication module is connected with the input end of the upper computer module; the inertial sensor module is used for acquiring joint angle and gait track information; the pressure sensor module is used for collecting plantar pressure information.

2. The multi-dimensional lower limb rehabilitation assessment device for stroke patients according to claim 1, wherein said pressure sensor module is a module consisting of a membrane pressure sensor, and is installed in an insole.

3. The multi-dimensional lower limb rehabilitation assessment device for stroke patients according to claim 2, characterized in that the number of said membrane pressure sensors is 8 pressure sensors; the 8 film pressure sensors are respectively connected with the main controller module through 8 analog-to-digital converters; the 8 film pressure sensors are respectively arranged in two insoles, wherein 4 film pressure sensors are arranged in one insole, three film pressure sensors are arranged at the front sole, and the other film pressure sensor is arranged at the heel.

4. The device of claim 3, wherein the positions of the 4 thin film pressure sensors in the insole correspond to the heel, the fourth metatarsal, the first metatarsal and the thumb, respectively, and the pressure signals of the positions of the 4 thin film pressure sensors are obtained respectively.

5. The multi-dimensional lower limb rehabilitation assessment device for stroke patients according to claim 1, characterized in that said inertial sensor module is composed of 4 inertial sensors; the 4 inertial sensors are connected with the main controller module; the 4 inertial sensors are respectively arranged at the thighs and the shanks of the two legs; the communication module is one of WiFi,5G,4G or serial port communication modes.

6. A multi-dimensional lower limb rehabilitation assessment method for stroke patients is characterized by comprising the following steps:

(1) Acquiring a joint angle signal by using an inertial sensor, solving and recording the flexion and extension angles of the joint of the patient by using the signal to obtain the flexion and extension angles of the joint of the patient, and respectively making ratios with medical standard joint angles to obtain a ratio; taking the ratio as the evaluation basis of the joint activity;

(2) Acquiring gait track signals by using an inertial sensor, recording the movement tracks of the knee joint and the hip joint of a patient, and taking the similarity degree of the movement tracks of the knee joint and the hip joint of the patient and the movement tracks of the standard knee joint and the hip joint as the evaluation basis of the gait track;

(3) Acquiring a plantar pressure signal by using a pressure sensor, identifying the gait phase of a patient by using the plantar pressure signal, recording the gait phase sequence of two sides, obtaining the abnormal gait cycle number and the total gait cycle number, calculating the proportion of the abnormal gait cycle number to the total gait cycle number, and using the proportion as the evaluation basis of the gait phase;

(4) Because the joint activity, the gait track and the gait phase have different functions in the multi-dimensional lower limb rehabilitation evaluation, in order to obtain a scientific and accurate rehabilitation evaluation result, the three needs to be comprehensively evaluated, and the weights of the joint activity evaluation basis obtained in the step (1), the gait track evaluation basis obtained in the step (2) and the gait phase evaluation basis obtained in the step (3) are determined by an analytic hierarchy process, namely: weight w of degree of articulation ₁ Gait trajectory weight w ₂ And gait phase weight w ₃ (ii) a The method specifically comprises the following steps:

(4-1) dividing the analytic hierarchy process into three levels according to the analytic mode of the analytic hierarchy process, namely: a target layer, a criterion layer and a scheme layer; wherein the target layer is the degree of recovery of joint function of the lower limb; the criterion layer is a joint activity degree evaluation basis obtained in the step (1), a gait track evaluation basis obtained in the step (2) and a gait phase evaluation basis obtained in the step (3); the scheme layer is a weight coefficient occupied by the joint activity evaluation basis obtained in the step (1), the gait track evaluation basis obtained in the step (2) and the gait phase evaluation basis obtained in the step (3);

(4-2) quantifying the ratio of joint activity, gait track and plantar pressure by using a '1-9 scale method', comparing every two of the joint activity evaluation basis obtained in the step (1), the gait track evaluation basis obtained in the step (2) and the gait phase evaluation basis obtained in the step (3) in a criterion layer to form a pair comparison table, wherein the heading items of rows and columns of the table are evaluation indexes and the sequence of the table is consistent;

TABLE 3

Degree of motion of joint Gait trajectory Gait phase Degree of motion of joint a ₁₁ a ₁₂ a ₁₃ Gait trajectory a ₂₁ a ₂₂ a ₂₃ Gait phase a ₃₁ a ₃₂ a ₃₃

(4-4) from the table of three rows and three columns obtained in step (4-2), a pair-wise comparison matrix as shown in formula (17) can be obtained, namely: judging the matrix A:

judging that each element in the matrix A corresponds to data in each cell in the table with three rows and three columns;

(4-5) normalizing the column vectors of the judgment matrix a, and simultaneously summing and normalizing the rows, the formula (18) can be obtained:

wherein, w ₁ 、w ₂ 、w ₃ Joint activity degree weight, gait track weight and gait phase weight;

(4-6) carrying out consistency verification on the formula (18) to obtain a formula (19):

since a ω = λ ω, then:

wherein n is the order of the matrix;

(4-7) random consistency index RI, as shown in Table 4:

TABLE 4

n 1 2 3 4 5 6 7 8 9 10 11 RI 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 1.51

Further, the inconsistency index CI and the consistency ratio CR can be obtained as shown in formula (21) and formula (22), respectively:

if the consistency ratio CR is less than 0.1, the inconsistency degree in the table 3 is considered to be within the allowable range, and satisfactory consistency is obtained, and the feature vector omega can be normalized by the formula (18) as a weight vector through consistency test; if the consistency ratio does not meet the requirement, constructing a judgment matrix for the comparison matrix according to the step (4-2) and the step (4-3), and performing consistency verification by using the steps (4-5) to (4-7) to finally obtain the weight occupied by each layer.

(5) Combining the score R of the joint activity degree in the step (1) according to the weight of the evaluation index determined in the step (4) _S The score T of the gait trajectory in the step (2) _S The gait phase score P of step (3) _S Then, a comprehensive rehabilitation evaluation score can be obtained, which is shown in formula (23):

Z _S ＝w ₁ ×R _S +w ₂ ×T _S +w ₃ ×P _s (23)

wherein Z _S Scoring for comprehensive rehabilitation evaluation;

(6) And (3) evaluating the rehabilitation effect of the patient according to the comprehensive rehabilitation evaluation score obtained by the formula (23): the rehabilitation effect is excellent for more than 90 minutes, the rehabilitation effect is good between 80 and 90 minutes, the rehabilitation effect is general between 70 and 80 minutes, and the rehabilitation effect is poor for less than 60 minutes; when the comprehensive rehabilitation evaluation score of the patient is below 60 points, a conclusion that the rehabilitation training strength needs to be strengthened is given, and the training mode at the stage is similar to a passive rehabilitation training mode; when the comprehensive rehabilitation evaluation score of the patient is between 70 and 80 points, a conclusion that the rehabilitation training intensity can be properly reduced is given, and an active training mode is properly added on the basis of a passive training mode; when the comprehensive rehabilitation evaluation score of the patient is between 80 and 90 points, a conclusion that the active training mode can be used as the main training mode is given, and the antagonistic training mode is properly added; when the patient composite rehabilitation evaluation score is above 90 points, a suggestion that the antagonistic training mode can be used as the main training mode is given.

7. The method for assessing the rehabilitation of lower limbs of patients with cerebral apoplexy according to claim 6, wherein the step (1) is to process the joint angle signals obtained by the inertial sensor by using the maximum and minimum joint degree algorithm, i.e. to obtain the hip joint flexion angle A of the patient respectively ₁ Hip joint extension angle B of patient ₁ Knee joint flexion angle C of patient ₁ Knee joint extension angle D of patient ₁ And respectively calculating the flexion angle A with the standard hip joint ₁₁ Standard hip joint extension angle B ₁₁ Standard knee joint flexion angle C ₁₁ Standard knee joint extension angle D ₁₁ As shown in table 1; middle period, ω _A 、ω _B 、ω _C 、ω _D Respectively weighing the joint mobility score which is occupied by 4 index scores in total when the knee joints and the hip joints of the two legs of the patient are bent and extended;

TABLE 1

Then, the score R of the joint mobility can be obtained according to the formula (1) _S ；

8. The method for assessing the rehabilitation of the lower limbs of the patient with the stroke as claimed in claim 6, wherein the step (2) uses a cosine similarity algorithm to process the gait track signals obtained by the inertial sensor, so as to obtain and record the movement tracks of the knee joint and the hip joint of the patient, namely: selecting points with the same number as the hip joint motion trail of the patient and the standard hip joint motion trail to form two vectors a1 and b1, selecting points with the same number as the knee joint motion trail of the patient and the standard knee joint motion trail to form two vectors a2 and b2, respectively calculating cosine included angles of the four vectors, such as formula (2) and formula (3), and taking the cosine included angles as evaluation basis of the gait trail;

then, the score of the gait trajectory is the formula (4)

T _S ＝(cosθ ₁ ×w ₁ +cosθ ₂ ×w ₂ )×100 (4)

In the formula (4), w may be taken ₁ ＝w ₂ ＝0.5，w ₁ Weighting the hip joint movement track score to the gait track score, w ₂ And weighting the knee joint movement track score to the gait track score.

9. The method for assessing the rehabilitation of the lower limbs of the patient with the stroke as claimed in claim 6, wherein the step (3) is performed by using a proportional fuzzy logic phase-splitting method to identify the gait phase, record the sequence of the gait phases on both sides, and according to the gait phase identification result, the abnormal gait cycle number and the total gait cycle number are obtained for the time intervals of the phase at the early stage of standing, the phase at the middle stage of standing, the phase at the late stage of standing and the phase at the swing stage, so as to obtain the ratio of the abnormal gait cycle number to the total gait cycle number, and the method specifically comprises the following steps:

(3-1) proportionally fusing the plantar pressure signals, namely: the sole pressure signals are summed at the same time and then subjected to proportional operation, so that the identification error caused by individual difference is solved, as shown in formulas (5) to (8):

wherein, K ₁ 、K ₂ 、K ₃ 、K ₄ The ratios of 4 pressure sensor signals to the total of sole pressure signals are respectively calculated; f ₁ 、F ₂ 、F ₃ 、F ₄ Respectively representing the pressure of the 4 pressure sensors at the same moment;

(3-2) fuzzifying the proportion-fused data obtained in the step (3-1) to determine a proper membership function, wherein the proportion data of 4 pressure sensor signals which are proportionally fused in the step (3-1) to the total sum of sole pressure signals, namely K ₁ 、K ₂ 、K ₃ 、K ₄ And the value ranges of the two groups are all between 0 and 1, and in order to eliminate sharp boundaries of traditional gait stage division, an exponential function is selected as a membership function, as shown in a formula (9):

wherein, f (K) _i ) Is the output of a function, K _i The proportion K of the signals of the 4 pressure sensors obtained in the step (3-1) to the total sum of the pressure signals of the sole ₁ 、K ₂ 、K ₃ 、K ₄ ，K _0i Is a proportional threshold, s is a sensitivity coefficient;

(3-3) since the membership function shown in the formula (9) in the step (3-2) is a symmetric function, in order to describe the proportional value K _i To the extent of approaching 1 or approaching 0, the inverse function of formula (9) in step (3-2) can be directly taken, and f (K) can be calculated _i ) Is divided into f ^H (K _i ) And f ^L (K _i ) Two parts, f ^H (K _i ) Representative of the ratio value K _i Degree close to 1, f ^L (K _i ) Represents the proportional value K _i A degree close to 0, as shown in equation (10) and equation (11), respectively;

(3-4) setting a fuzzy logic rule table based on the function output of the formula (10) and the formula (11) in the step (3-3), as shown in table 2, the medical standard divides the gait into six phases, which are initial contact, load response, mid-stance phase, late-stance phase, expected swing phase and late swing phase, respectively, and the mechanical differentiation degree between the late-stance phase and the late pre-swing phase is low due to the initial contact and the load response. Therefore, the gait phase identification method divides the gait into four time phases again when the proportional fuzzy logic phase separation method identifies the gait phase, namely the early stage of the stance and the late stage of the stance, namely: standing earlier stage, standing middle stage, standing later stage, and swing stage;

TABLE 2

From table 2, the phase splitting formulas of the early stage of standing, the middle stage of standing, the late stage of standing and the swing stage shown in formulas (12) to (15) can be obtained, namely the phase splitting result of the gait phase of the patient;

AS＝f ^H (K ₁ )×f ^L (K ₃ )×f ^L (K ₄ ) (12)

BS＝f ^H (K ₂ )×f ^H (K ₃ )×f ^L (K ₄ ) (13)

CS＝f ^L (K ₁ )×f ^H (K ₄ ) (14)

DS＝f ^L (K ₁ )×f ^L (K ₂ )×f ^L (K ₃ )×f ^L (K ₄ ) (15)

wherein, when K _i ＞K _0i When the fuzzy set is in a large state, the fuzzy set is marked as H, and when K is reached _i ＜K _0i If the fuzzy set is in a small state, the fuzzy set is marked as 'L', and if the fuzzy set is in a non-state, the fuzzy set is marked as 'X';

(3-5) calculating the duration of the gait phase in one gait cycle based on the outputs of equations (12) - (15) in step (3-4): one gait cycle comprises a standing prophase, a standing middle phase, a standing later phase and a swing phase, and the sequence of the gait phases is the standing prophase- > the standing middle phase- > the standing later phase- > the swing phase; comparing the gait phase duration with the gait phase time of the medical standard, and dividing the gait phase duration of the medical standard into 8% -12% of a gait cycle in the early stage of standing, 18% -22% of a gait cycle in the middle stage of standing, 28% -32% of a gait cycle in the late stage of standing and 38% -42% of a gait cycle in the swing stage DS for considering individual difference; finally, judging whether the gait phase duration time exceeds the range specified by the medical standard gait phase duration time, if so, determining the gait phase duration time as an abnormal gait cycle, and if not, determining the gait phase duration time as a normal gait cycle; taking the ratio of the abnormal gait cycle number to the total gait cycle number as the evaluation basis of the gait phase, wherein the total gait cycle number is the sum of the abnormal gait cycle number and the normal gait cycle number, the abnormal gait cycle number is the sum of the abnormal gait cycle number, and the normal gait cycle number is the sum of the normal gait cycle number; at this time, the gait phase score is expressed by equation (16):

wherein, P _S Score gait phase, A _n For abnormal cycles of gait, T _n The total number of gait cycles.

10. The method for assessing the rehabilitation of the lower limbs of the patient with cerebral apoplexy according to claim 6, wherein the numerical values in table 3 in the step (4) are official numerical values provided by a 1-9-scale method, as shown in table 3-1:

TABLE 3-1

Degree of motion of joint Gait trajectory Gait phase Degree of motion of joint 1 1/3 1/4 Gait trajectory 3 1 1/2 Gait phase 4 2 1

In the table, "numeral 1" indicates that two evaluation criteria of row and column are equally important, "numeral 3" indicates that row is slightly more important than column, "numeral 5" indicates that row is significantly more important, "numeral 2" indicates that the degree of importance of row to column is between "numeral 1" and "numeral 3," numeral 4 "indicates that the degree of importance of row to column is between" numeral 3 "and" numeral 5, "and the score part is the degree of importance of column to row, for example," numeral 1/3 "indicates that column is slightly more important than row;

further obtaining:

normalizing the column vector of the judgment matrix A, and simultaneously summing and normalizing the rows to obtain the final product

Namely: the weight of the joint activity degree is 0.123, the weight of the gait track is 0.32, and the weight of the gait phase is 0.557;

and (5) carrying out consistency verification to obtain:

since a ω = λ ω, then:

wherein n is the order of the matrix;

further, an inconsistency index CI and a consistency ratio CR can be obtained:

therefore, the degree of inconsistency in table 3-1 is within the allowable range and has satisfactory consistency, and the normalized feature vector ω obtained at this time is: the joint motion degree weight is 0.123, the gait trajectory weight is 0.32, and the gait phase weight is 0.557, which can be used as the weight of the evaluation index.

Images