Double-threshold decision-making spasm evaluation method and device

Abstract

The invention belongs to the technical field of rehabilitation medical evaluation, and particularly relates to a spasm evaluation method and device based on double-threshold decision. A rehabilitation doctor drives a patient to bend and stretch an upper limb elbow joint in a cycle at a proper speed, and electromyographic signals x (t) of biceps brachii or triceps brachii of the patient are collected through an electromyographic signal collection module; the joint angle theta (t) of the elbow joint of the patient is collected through an angle measurement sensor; an impedance force measuring module (1) installed on the forearm of the patient is used for collecting impedance force F (t) when the elbow joint stretches or bends; according to the method, multi-dimensional information in the passive flexion and extension process of a subject is utilized, the characteristic that nerve reflex is earlier than biomechanical reflex is utilized, and the time difference of spasm reactions of the nerve reflex and the biomechanical reflex is controlled to be between-30ms and 200ms, so that false positive judgment is avoided, the spasm occurrence time and the joint angle value during spasm occurrence can be accurately and timely judged, and the joint angle value can be accurately and accurately judged. A doctor only needs to drive a patient to do stretching / buckling movement once, and then spasm evaluation can be completed.

Description

Technical Field

[0001] The present invention belongs to the technical field of rehabilitation medical assessment, and in particular relates to a double-threshold decision-making method and device for spasticity assessment. Background Art

[0002] Spasticity is a movement disorder characterized by increased muscle tone and tendon hyperreflexia due to increased stretch reflex excitability. Spasticity treatment is a systematic process, and objective and reliable quantitative assessment is an important basis for physicians to formulate rehabilitation plans.

[0003] Existing methods for assessing spastic muscle tension mostly rely on expert experience, such as rapid PROM assessment and modified Ashworth scale. The test results are highly subjective and it is difficult to accurately quantify the degree of muscle tension.

[0004] Objective assessment of spasticity is influenced by numerous factors, including the subject's etiology, physical condition, subjective participation, and the stretching technique used by the examiner. Existing objective methods require rehabilitation physicians to repeatedly guide patients through stretching exercises, which can easily cause muscle fatigue and boredom, leading to inaccurate assessment results.

[0005] To solve this problem, the present invention utilizes the characteristic that physiological signals are generated before physical signals, designs a multi-dimensional data acquisition device for detecting upper limb elbow joint spasm, and proposes a new method for spasm assessment based on stretch reflex threshold. Summary of the Invention

[0006] The purpose of the present invention is to overcome the defects in the prior art that the spasm grade assessment by physician examiners is highly subjective, easily judged as false positive, and difficult to accurately quantify the spasm grade, and to provide a spasm assessment method and device with dual threshold decision-making.

[0007] The technical solution adopted by the present invention to solve its technical problem is:

[0008] A double-threshold decision-making method for spasticity assessment, characterized by comprising the following steps:

[0009] S1. The rehabilitation physician guides the patient to perform a cycle of flexion and extension of the upper limb elbow joint at an appropriate speed, and uses the electromyographic signal acquisition module to collect the electromyographic signal x(t) of the patient's biceps or triceps; uses the angle measurement sensor to collect the joint angle θ(t) of the patient's elbow joint; and uses the impedance force measurement module (1) installed on the patient's forearm to collect the impedance force F(t) when the elbow joint is extended or flexed;

[0010] S2. Determine the duration of spasms using physiological signals; process the EMG signal x(t) using the SSA-VMD-Garrote combined threshold algorithm to reduce baseline activity and improve the signal-to-noise ratio; and extract the sEMG envelope based on the Hilbert transform.

[0011] Set the extracted sEMG envelope segment of the electromyographic data The reflex is a spasm reflex, where x is the mean value of the electromyographic signal of the flexion / extension cycle, s is the standard deviation of the electromyographic signal of the flexion / extension cycle, and the starting point of the spasm reflex is defined as the spasm occurrence point, that is, the time when the first trough point of the electromyographic signal of this segment occurs is the first threshold t1;

[0012] S3. Determine the spasm time by physical signals; define the time when the impedance step reaction occurs as the second threshold t2;

[0013] If S4.t1-30ms<t2<t1+200ms, it is determined that spasm has occurred, that is, it is determined to be positive; otherwise it is determined to be a false positive.

[0014] Furthermore, in step S2, the method for processing the electromyographic signal x(t) by the SSA-VMD-Garrote combined threshold algorithm is:

[0015] Initialization: Set the parameters of the SSA algorithm. The population size is 20, the maximum number of iterations is 10, the penalty factor search range is [100, 3000], the modal number search range is [3, 10], and each sparrow individual represents a set of α and k values;

[0016] Fitness function: Define the fitness function and use envelope entropy as the fitness function. The goal is to minimize the envelope entropy. The envelope entropy can reflect the changes and fluctuations in muscle activity intensity.

[0017] Parameter optimization: Use SSA to optimize the VMD parameters α and k to find the parameter combination that optimizes the fitness function value;

[0018] VMD decomposition: Use the optimal parameters α and k obtained by optimization to perform VMD decomposition on the signal;

[0019] Assume that the original signal x(t) is decomposed into k components, ensuring that the decomposition sequence is a modal component with a finite bandwidth and a center frequency, and that the sum of the estimated bandwidths of each mode is minimized. The constraint is that the sum of all modes is equal to the original signal. The idea of VMD is to minimize the sum of the bandwidths of each mode near its center frequency while ensuring that the sum of the modes approximates x(t). This is considered an optimization problem and solved using the calculus of variations.

[0020] The objective function of the optimization problem is

[0021]

[0022] And satisfy

[0023]

[0024] Among them, {u k (t)} is the set of all modal functions, {ω k} is the set of all modal center frequencies,

[0025] In the VMD algorithm, optimization is achieved by iteratively solving the Lagrange multiplier method;

[0026]

[0027] Where α is the penalty factor and λ is the Lagrange multiplier

[0028] In specific implementation, all quantities are often transformed by Fourier transform first. Indicates u k (t) is represented in the frequency domain. represents the representation of the original signal x(t) in the frequency domain. The Lagrange multiplier Then it corresponds to the constraint;

[0029] In each iteration, for the kth mode, we first assume that the other modes are known and fixed, and then update

[0030]

[0031] Calculate the new center frequency:

[0032]

[0033] Finally update the Lagrange multiplier:

[0034]

[0035] Where γ is the step size, which is used to control the update speed;

[0036] Garrote function threshold denoising:

[0037]

[0038] Among them, x i is a coefficient of some IMF, is the coefficient after denoising, λ is the threshold (2 times the standard deviation);

[0039] Signal reconstruction: The filtered IMF is reconstructed to obtain the final output signal. The reconstruction process is to superimpose each IMF to restore the original electromyographic signal x(t). Through reconstruction, we can obtain the signal after the Garrote function threshold denoising, thereby achieving the purpose of extracting effective signal information.

[0040] Hilbert transform to extract the EMG signal envelope;

[0041] The Hilbert transform of the electromyographic signal x(t) is defined as:

[0042]

[0043] Where H represents the Hilbert change operation, P represents the principal value integral, x(t) is the original signal, is the Hilbert transform of x(t)

[0044] The analytical signal z(t) is defined as

[0045]

[0046] The modulus A(t) of the analytical signal is the envelope of the signal

[0047]

[0048] Furthermore, the joint angle θ corresponding to the first threshold t1 of the EMG signal spasm p is the stretch reflex threshold; the assessment process affects the patient's stretch reflex threshold θ p , maximum angle θ rom Recording, using θ p / θ rom Numerical experiments on the objective evaluation of spasticity level; During the evaluation of flexor spasticity, θ p Indicates the angle from the point where the spasm occurs to the limit of extension; θ rom Indicates the range of motion of the elbow joint of the upper limb during the test, which is defined as the range of motion from the maximum flexion angle to the maximum extension angle; during the assessment of extensor spasm, θ p The angle from the point where the spasm occurs to the flexion limit is calculated in the same way as θ. p / θ rom .

[0049] The present invention also discloses a dual-threshold decision-making spasm assessment device, which is used for signal acquisition of the above-mentioned dual-threshold decision-making spasm assessment method, and is characterized in that it includes an upper arm fixing rod, a lower arm length adjustment rod assembly rotatably connected to the upper arm fixing rod through an elbow joint rotating shaft, an angle measurement sensor installed on the elbow joint rotating shaft, and an impedance force measurement module installed on the lower arm length adjustment rod assembly; the impedance force measurement module includes an arc-shaped mounting seat installed on the lower arm length adjustment rod assembly and two force sensors arranged opposite to each other respectively installed at both ends of the arc-shaped mounting seat; the force sensor is installed on the arc-shaped mounting seat through a ball pair.

[0050] Furthermore, a fixing frame is fixedly installed on the upper arm fixing rod, and the upper arm of the human body is fixed by a strap installed on the fixing frame.

[0051] Furthermore, it also includes a sEMG sensor for collecting electromyographic signals and a sEMG acquisition card connected to the electrical signals of the sEMG sensor; the sEMG sensor has a wet electrode patch, and a sEMG sensor is respectively set on the biceps and triceps.

[0052] Furthermore, the sEMG sensor is fixedly attached to the upper arm by a strap.

[0053] Furthermore, it also includes a host computer, the sEMG acquisition card is wirelessly connected to the host computer via Bluetooth; the force sensor and the angle measurement sensor are connected to the host computer via RS484 communication.

[0054] The beneficial effects of the dual-threshold decision-making method for spasticity assessment of the present invention are:

[0055] This patented method utilizes multidimensional information from the subject's passive flexion and extension process, taking advantage of the fact that neural reflexes precede biomechanical reflexes. By controlling the time difference between the two spasm responses to between -30ms and 200ms, false positives can be avoided, allowing accurate and timely determination of the spasm timing and joint angle values during the spasm. The system also records multidimensional data during the extension / flexion process in real time, allowing the physician to complete the assessment of spasm simply by leading the patient through a single extension / flexion exercise. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0057] Figure 1 1 is a schematic diagram of the overall flow of the spasticity assessment method using dual-threshold decision-making according to an embodiment of the present invention;

[0058] Figure 2 is a flow chart of an evaluation method according to an embodiment of the present invention;

[0059] Figure 3 It is a flowchart of the SSA-VMD-Garrote joint threshold algorithm;

[0060] Figure 4 1 is an overall structural diagram of a spasm assessment device with dual threshold decision-making according to an embodiment of the present invention;

[0061] Figure 5 is a partial structural diagram of a spasticity assessment device with dual threshold decision-making according to an embodiment of the present invention;

[0062] Figure 6 This is a picture of a field-collected signal according to an embodiment of the present invention;

[0063] Figure 7 This is a schematic diagram of the SSA-VMD-Garrote combined threshold algorithm for processing electromyographic signals;

[0064] Figure 8 is the change in elbow joint parameters during one extension cycle of a patient with extensor muscle disorder according to an embodiment of the present invention;

[0065] Figure 9 is the change in elbow joint parameters during one extension cycle of a patient with flexor muscle disorder according to an embodiment of the present invention;

[0066] Figure 10 There are two examiners θ p / θ rom Scatterplot with MAS ratings.

[0067] In the figure: 1. Impedance force measurement module, 11. Force sensor, 12. Arc mounting base, 2. Forearm length adjustment rod assembly, 3. Angle measurement sensor, 4. Fixing bracket, 5. sEMG acquisition card, 51. Wet electrode patch, 6. sEMG sensor, 7. Reference electrode, 8. Upper arm fixing rod. DETAILED DESCRIPTION

[0068] The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic diagrams that illustrate the basic structure of the present invention in a schematic manner.

[0069] like Figure 4-Figure 5 The dual-threshold decision-making spasm assessment device of the present invention shown includes an upper arm fixing rod 8, a lower arm length adjustment rod assembly 2 rotatably connected to the upper arm fixing rod 8 via an elbow joint rotating shaft, an angle measurement sensor 3 installed on the elbow joint rotating shaft, and an impedance force measurement module 1 installed on the lower arm length adjustment rod assembly 2; the impedance force measurement module 1 includes an arc-shaped mounting seat 12 installed on the lower arm length adjustment rod assembly 2 and two force sensors 11 respectively installed at both ends of the arc-shaped mounting seat 12 and arranged opposite to each other; the force sensor 11 is installed on the arc-shaped mounting seat 12 via a ball pair.

[0070] The impedance force measurement module 1 is used to collect the impedance force during the extension or flexion movement of the elbow joint. Since the axis of the human joint is not completely matched with the axis of the device, and the center of rotation of the elbow joint will slide slightly during movement, the force sensor 11 is placed on the arc-shaped mounting seat 12 and designed as a ball pair to ensure that the angle between the forearm and the force sensor 11 is always vertical, thereby reducing the interference of the elbow joint displacement.

[0071] The forearm length adjustment rod assembly 2 is used to adjust the length of the forearm of the device so that the force sensor 11 is located at the 2 / 5 end of the ulna to reduce individual differences; the angle measurement sensor 3 is used to collect angle information and angular velocity information.

[0072] A fixing frame 4 is fixedly mounted on the upper arm fixing rod 8, and the upper arm of the human body is fixed by a strap installed on the fixing frame 4. According to ergonomic design, the radian design ensures that it conforms to the contour of the upper limb.

[0073] The spasticity assessment device also includes sEMG sensors 6 for collecting electromyographic signals and an sEMG acquisition card 5 electrically connected to the sEMG sensors 6. The sEMG sensors 6 are equipped with wet electrode patches 51, one for each of the biceps and triceps. The sEMG acquisition card 5 utilizes the OpenSignals wireless acquisition system from PLUX, a US company. The sEMG sensors 6 are used to collect sEMG signals from antagonist muscles. Each sEMG sensor 6 utilizes a wet electrode patch 51 containing electrolyte gel. These patches are attached to the center of each muscle along its fiber direction. Before attaching the electrodes, the corresponding muscle area is wiped with 75% alcohol to remove surface oil and improve conductivity, thereby reducing noise and enhancing the signal-to-noise ratio. The device is made of PLA plastic and lightweight aluminum alloy to minimize the impact of weight on the experiment.

[0074] If the patient has an extensor muscle disorder, a spasm will occur when the patient flexes the elbow, and the sEMG sensor 6 located on the triceps brachii collects the myoelectric signals. If the patient has a flexor muscle disorder, a spasm will occur when the patient extends the elbow, and the sEMG sensor 6 located on the biceps brachii collects the signals. A reference electrode 7 is also connected to the sEMG acquisition card 5. To ensure that the sEMG sensor 6 fits tightly against the upper arm muscles, a strap is used to press the sEMG sensor 6 against the upper arm.

[0075] The sEMG acquisition card 5 is wirelessly connected to the host computer via Bluetooth; the force sensor 11 and the angle measurement sensor 3 are connected to the host computer via RS484 communication.

[0076] In order to better illustrate the principle underlying the spasticity assessment method of this patent application, the upper limb elbow joint spasticity process is first analyzed.

[0077] Based on nerve reflex angle

[0078] When a joint reaches its stretch reflex threshold, receptors in the muscle spindles sense the stretch and activate motor neurons through a reflex arc, causing muscle contraction. This reflex is completed through the spinal cord and has a relatively fast effect. When the joint angle reaches the threshold, the sEMG signal increases immediately. This potential change occurs before the movement and is typically manifested as:

[0079] x(t)∝Muscle·Contraction·Force∝θ(t)·K m

[0080] Where x(t) is the myoelectric signal that changes with time; θ(t) is the joint angle; K m is muscle stiffness.

[0081] Near the stretch reflex threshold, the frequency and amplitude of the sEMG signal usually increase, appearing as a short "surge." In order to respond to rapid stretch, the nervous system increases the frequency of nerve firing to quickly trigger muscle contraction:

[0082]

[0083] where K n The gain of the nervous system indicates the strength of the neural reflex in regulating muscle contraction. When the joint angle exceeds the stretch reflex threshold, the muscle begins to reflexively increase contraction, resulting in a significant increase in sEMG activity. The time at which the spasm reflex, detected by the biosignal, occurs is used as the first threshold.

[0084] Based on the biomechanical perspective

[0085] During passive joint movement, resistance is determined by the stiffness and damping of the joint. When the stretch reflex is activated, muscle stiffness increases, causing a sharp change in joint resistance.

[0086] Model the motion of the elbow joint as a spring-damper system

[0087]

[0088] Where J is the moment of inertia of the system; c is the coefficient of viscous friction; and k is the spring constant.

[0089] In normal passive motion, the stiffness and damping coefficient of the joint are relatively small. When the joint angle exceeds the stretch reflex threshold, the reflex contraction of the muscle will cause a sudden increase in resistance, which is mainly manifested in the increase of stiffness k and damping c:

[0090]

[0091] When a joint reaches the stretch reflex threshold, friction and viscosity increase, and the system's total impedance F exhibits a sudden surge. Because the physical signal occurs after the movement, later than the sEMG, the time corresponding to the starting point of the impedance signal's step response is used as the second threshold to verify the authenticity of the first threshold, thereby improving the accuracy of stretch reflex threshold detection.

[0092] The double-threshold decision-making method for spasticity assessment specifically includes the following steps:

[0093] S1. The rehabilitation physician guides the patient to perform a cycle of flexion and extension of the upper limb elbow joint at an appropriate speed, and uses the electromyographic signal acquisition module to acquire the electromyographic signal x(t) of the patient's biceps or triceps; uses the angle measurement sensor (3) to acquire the joint angle θ(t) of the patient's elbow joint; and uses the impedance force measurement module (1) installed on the patient's forearm to acquire the impedance force F(t) when the elbow joint is extended or flexed. When the patient has a muscle disorder, the electromyographic signal of the triceps is acquired; when the patient has a flexion disorder, the electromyographic signal of the biceps is acquired.

[0094] S2. Determine the duration of spasms using physiological signals: Process the EMG signal x(t) using the SSA-VMD-Garrote combined threshold algorithm to reduce baseline activity and improve the signal-to-noise ratio; extract the sEMG envelope based on the Hilbert transform;

[0095] High-quality signals can often more accurately detect the onset of reflection activity; VMD decomposes the input signal into several physically meaningful intrinsic mode functions (IMFs) by iteratively searching for the best estimate of a series of finite-bandwidth modes. These mode functions have different frequencies and amplitudes and can represent different components in the original signal. However, the decomposition effect of VMD is heavily dependent on the selection of the penalty factor α and the number of modes k. The parameter α controls the modal bandwidth, and the parameter k determines the number of decomposed modes. However, this method has the problems of slow convergence and easy to fall into local optimality. In order to improve the performance of the algorithm, this patent introduces the SSA optimization algorithm.

[0096] Initialization: Set the parameters of the SSA algorithm. The population size is 20, the maximum number of iterations is 10, the penalty factor search range is [100, 3000], the modal number search range is [3, 10], and each sparrow individual represents a set of α and k values;

[0097] Fitness function: Define the fitness function and use envelope entropy as the fitness function. The goal is to minimize the envelope entropy. The envelope entropy can reflect the changes and fluctuations in muscle activity intensity.

[0098] Parameter optimization: Use SSA to optimize the VMD parameters α and k to find the parameter combination that optimizes the fitness function value;

[0099] VMD decomposition: Use the optimal parameters α and k obtained by optimization to perform VMD decomposition on the signal;

[0100] Assume that the original signal x(t) is decomposed into k IMF components, ensuring that the decomposition sequence is a modal component with a finite bandwidth and a center frequency, and that the sum of the estimated bandwidths of each mode is minimized. The constraint is that the sum of all modes is equal to the original signal. The idea of VMD is to minimize the sum of the bandwidths of each mode near its center frequency while ensuring that the sum of the modes approximates x(t). This is considered an optimization problem and solved using the calculus of variations.

[0101] The objective function of the optimization problem is

[0102]

[0103] And satisfy

[0104]

[0105] Among them, {u k (t)} is the set of all modal functions, {ω k} is the set of all modal center frequencies,

[0106] In the VMD algorithm, optimization is achieved by iteratively solving the Lagrange multiplier method;

[0107]

[0108] Where α is the penalty factor and λ is the Lagrange multiplier

[0109] When implementing it, first perform Fourier transform on all quantities. Indicates u k (t) is represented in the frequency domain. represents the representation of the original signal x(t) in the frequency domain. The Lagrange multiplier Then it corresponds to the constraint;

[0110] In each iteration, for the kth mode, we first assume that the other modes are known and fixed, and then update

[0111]

[0112] Calculate the new center frequency:

[0113]

[0114] Last updated Lagrange multiplier

[0115]

[0116] Where γ is the step size, which is used to control the update speed;

[0117] Garrote function threshold denoising:

[0118]

[0119] Among them, x i is a coefficient of some IMF, is the coefficient after denoising, λ is the threshold (2 times the standard deviation);

[0120] Signal reconstruction: The filtered IMF is reconstructed to obtain the final output signal. The reconstruction process is to superimpose each IMF to restore the original electromyographic signal x(t). Through reconstruction, we can obtain the signal after the Garrote function threshold denoising, thereby achieving the purpose of extracting effective signal information.

[0121] Hilbert transform to extract the EMG signal envelope;

[0122] The Hilbert transform of the signal x(t) is defined as

[0123]

[0124] Where H represents the Hilbert change operation, P represents the principal value integral, x(t) is the original signal, is the Hilbert transform of x(t)

[0125] The analytical signal z(t) is defined as

[0126]

[0127] The modulus A(t) of the analytical signal is the envelope of the signal

[0128]

[0129] The joint angle θ corresponding to the first threshold t1 of the EMG signal spasm p is the stretch reflex threshold.

[0130] Figure 7 The graph shows the changes in electromyographic signals processed by the SSA-VMD-Garrote combined threshold algorithm.

[0131] Set the extracted sEMG envelope to include the electromyographic data outside this range. The reflex is a spasmodic reflex, in which is the mean value of the electromyogram signal for the flexion / extension cycle, s is the standard deviation of the electromyogram signal for the flexion / extension cycle. The starting point of the spasm reflex is defined as the spasm occurrence point, that is, the time when the first trough point of this segment of the electromyogram signal occurs is the first threshold t1;

[0132] S3. Determine the spasm time through physical signals: Define the time when the impedance force has a step response as the second threshold t2;

[0133] S4. If t1 - 30ms < t2 < t1 + 200ms, it is determined that spasm occurs, that is, it is determined as positive, otherwise it is determined as false positive.

[0134] Figure 8 and Figure 9 respectively show the changes in elbow joint parameters within one movement cycle of patients with flexor disorder and patients with extensor disorder.

[0135] The moment t1 is the starting point of the electromyogram reflex, and the moment t2 is the starting point of the impedance force step response (t1 - 30ms < t2 < t1 + 200ms). It is determined that spasm occurs. The joint angle corresponding to the electromyogram reflex threshold t1 is the stretch reflex threshold θ p .

[0136] For the spasm assessment method of this patent application, 12 patients with dystonia who were undergoing rehabilitation treatment in the Rehabilitation Department of the Affiliated Hospital of Hebei University of Engineering were selected as subjects (average age 57 years, 9 males, 3 females, 6 left affected sides, 6 right affected sides, 9 flexor spasm patients, 3 extensor spasm patients). Each patient signed an informed consent form for the experimental study before the experiment. The personal relevant information of the subjects is shown in Table 1.

[0137] Table 1 Basic information table of patients

[0138] serial number gender age Ill-affected side Onset time Spastic muscles S1 male 67 right 45 flexor muscles S2 female 58 right 32 flexor muscles S3 male 33 Left 100+ flexor muscles S4 male 68 right 42 extensor muscles S5 male 70 right 45 flexor muscles S6 male 24 Left 200+ extensor muscles S7 female 60 right 200+ flexor muscles S8 male 66 Left 87 extensor muscles S9 male 68 Left 75 flexor muscles S10 female 66 Left 100+ flexor muscles S11 male 37 Left 53 flexor muscles S12 male 73 right 82 flexor muscles

[0139] Inclusion criteria: ① Patients with upper limb spasm caused by stroke and brain injury; ② No skin breakage and unobstructed passive movement; ③ Normal mental state and cognition, and willing to cooperate with the operation process of this experiment.

[0140] Exclusion criteria: ① Participated in other clinical experiments one month ago; ② Suffering from other peripheral nerve or central nervous system diseases; ③ Accompanied by other skeletal diseases that can affect the movement of the upper limb elbow joint; ④ Unsuitable for this wearable device; ⑤ Poor consciousness and unstable condition.

[0141] To reduce the influence of environmental factors on the spasm assessment results, each data collection is carried out at the same time period. Each subject takes a lying position and is ensured to be in a relaxed state before being evaluated. Such as Figure 4As shown, the rehabilitation physician places the assessment device on the subject's affected upper limb, ensuring that the axis of the angle measurement sensor 3 aligns with the center of the elbow joint. The device's mobile end is positioned at the distal 2 / 5 of the ulna, and the contact surface with the body is wrapped with an elastic bandage to prevent indentation during flexion and extension and to prevent fat from affecting the test results. The fixed end is aligned with the long axis of the upper arm, and a fixed strap ensures that the device does not loosen during exercise. This ensures that the device's rotation axis is coaxial with the elbow joint's rotation axis. Spasticity assessment can only be performed after the subject is completely relaxed.

[0142] Rehabilitation physicians performed elbow flexion and extension on each participant at an appropriate speed based on their experience and simultaneously assessed the MAS score. The MAS categorizes spasticity on a scale of 0 to 4. For statistical analysis purposes, a score of 1+ on the scale was recorded as 1.5 in this article.

[0143] The data of 12 cases collected by the above method were analyzed. In order to ensure the accuracy of the data, two physicians conducted spasticity assessment on each test subject. p The angle from the point where the spasm occurs to the limit of extension, θ rom The range of motion of the elbow joint of the upper limb during the test is defined as the range of motion from the maximum flexion angle to the maximum extension angle. p , maximum angle θ rom Recording, using θ p / θ rom The objective evaluation of the spasm level in the numerical experiment. During the evaluation of extensor spasm, θ p The angle from the point where the spasm occurs to the flexion limit is calculated in the same way as θ. p / θ rom The results of the measurement using this device are shown in Table 2.

[0144] Table 2 Test results and MAS evaluation of this device

[0145]

[0146] S1-S12: Testers 1-12; E1, E2: Examiner 1, Examiner 2; “-” means no spasm was detected

[0147] Figure 10 There are two examiners θ p / θ rom Scatterplot with MAS ratings.

[0148] Pearson correlation coefficient was used to analyze the relationship between MAS and θ. p / θ rom The correlation between the device and the spasticity level was tested to verify the validity of the device and the effectiveness of the quantitative evaluation of the spasticity level. Figure 10 As shown, N represents the number of subjects. The results show that θ p / θ rom Significantly correlated with MAS score, θ p / θ rom The correlation with MAS score satisfies (r E1 =0.929, r E2 =0.949, *p<0.05). E1 The detection result θ of the first inspector p / θ rom Pearson correlation coefficient between the MAS evaluation, r E2 The detection result θ of the second examiner p / θ rom Pearson correlation coefficient between the evaluation and MAS, *p<0.05 indicates statistical significance.

[0149] Finally, based on the above experimental data, two patients with upper limb flexor spasticity in the Affiliated Hospital of Hebei Engineering University were randomly selected to use the θ p / θ rom The same physician used MAS and this device to evaluate upper limb spasticity. The corresponding spasticity level when it is the minimum value is the subject's spasticity level, where Δθ represents the random sampling of patients and the above subjects θ p / θ rom The difference between the two subjects and the test results are shown in Table 3. The results show that the device-predicted MAS of the two subjects was consistent with the subjectively assessed MAS, indicating that the stretch reflex threshold can determine the subject's MAS score to a certain extent.

[0150] Table 3 Device verification results

[0151]

[0152] Compared with the traditional 0-4 level MAS assessment, this method can more finely divide the spasticity level. p / θ rom The maximum difference in the test results was 0.09, but the MAS grade was all 1, indicating that the degree of spasticity of subject S11 was higher than that of subject S12 on the basis of MAS score 1.

[0153] The objective assessment of spasticity is influenced by many factors, including the subject's etiology, physical condition, subjective participation, and the examiner's stretching technique, all of which can affect the test results to some extent. This paper avoids the TSRT method of multiple stretches to detect spasticity, thereby ignoring the influence of the examiner's stretching speed.

[0154] This patented invention is based on neural reflex mechanisms and biomechanical models, utilizing multidimensional information from the subject's passive flexion and extension process. It can objectively quantify spasticity within at least one test cycle, improving detection efficiency and providing a theoretical basis for developing adaptive control strategies in the field of rehabilitation robotics. It should be understood that the specific embodiments described above are intended only to explain the present invention and are not intended to limit it. Obvious changes or modifications derived from the spirit of the present invention remain within the scope of protection of the present invention.

Claims

1. A dual-threshold decision-making method for spasticity assessment, characterized by: The following steps are involved: S1. The physician examiner guides the patient to perform a cycle of flexion and extension of the upper limb elbow joint at an appropriate speed, and uses the electromyographic signal acquisition module to collect the electromyographic signal x(t) of the patient's biceps or triceps; uses the angle measurement sensor (3) to collect the joint angle θ(t) of the patient's elbow joint; and uses the impedance force measurement module (1) installed on the patient's forearm to collect the impedance force F(t) when the elbow joint is extended or flexed; S2. Determine the onset of spasms using physiological signals; process the EMG signal x(t) using the SSA-VMD-Garrote combined threshold algorithm to reduce baseline activity and improve the signal-to-noise ratio; and extract the sEMG envelope using the Hilbert transform. Set the value beyond the extracted sEMG envelope data The reflex is a spasmodic reflex, in which is the mean value of the electromyographic signal of the flexion / extension cycle, s is the standard deviation of the electromyographic signal of the flexion / extension cycle, and the starting point of the spasm reflex is defined as the spasm occurrence point, that is, the time when the first trough point of the electromyographic signal segment where the spasm reflex occurs is the first threshold t1; S3. Determine the time of spasm occurrence by physical signals; define the time when the impedance step reaction occurs as the second threshold t2; S4. If t1-30ms<t2<t1+200ms, it is determined that spasm has occurred, that is, it is determined to be positive; otherwise, it is determined to be a false positive.

2. The dual-threshold decision-making method for spasticity assessment according to claim 1, characterized in that: The method for processing the electromyographic signal x(t) by the SSA-VMD-Garrote combined threshold algorithm in step S2 is: Initialization: Set the parameters of the Sparrow Optimization Algorithm (SSA); the population size is 20, the maximum number of iterations is 10, the penalty factor search range is [100, 3000], the modal number search range is [3, 10], and each sparrow individual represents a set of penalty factor α and modal number k values; Fitness function: Define the fitness function and use envelope entropy as the fitness function. The goal is to minimize the envelope entropy. The envelope entropy can reflect the changes and fluctuations in muscle activity intensity. Parameter optimization: Use SSA to optimize the VMD parameter α and the modal number k to find the parameter combination that optimizes the fitness function value; VMD decomposition: Use the optimal parameters α and k obtained by optimization to perform VMD decomposition on the signal; Assume that the original signal x(t) is decomposed into k IMF components, ensuring that the decomposition sequence is a modal component with a finite bandwidth and a center frequency. At the same time, the sum of the estimated bandwidths of each mode is minimized. The constraint is that the sum of all modes is equal to the original signal. The VMD idea is: while ensuring that the sum of each mode approaches x(t), the sum of the bandwidths of each mode near its center frequency is minimized. This is regarded as an optimization problem and solved by variational methods. The objective function of the optimization problem is And satisfy Among them, {u k (t)} is the set of all modal functions, {ω k } is the set of all modal center frequencies, In the VMD algorithm, optimization is achieved by iteratively solving the Lagrange multiplier method; Where α is the penalty factor and λ is the Lagrange multiplier When implementing it, first perform Fourier transform on all quantities; Indicates u k (t) is represented in the frequency domain. Represents the representation of the original signal x(t) in the frequency domain; and the Lagrange multiplier Then it corresponds to the constraint; In each iteration, for the kth mode, we first assume that the other modes are known and fixed, and then update Calculate the new center frequency: Finally update the Lagrange multiplier: Where γ is the step size, which is used to control the update speed; Garrote function threshold denoising: Among them, x i is a coefficient of some IMF, is the coefficient after denoising, λ is the threshold (2 times the standard deviation); Signal reconstruction: The filtered IMF is reconstructed to obtain the final output signal. The reconstruction process is to superimpose each IMF to restore the original electromyographic signal x(t). Through reconstruction, we can obtain the signal after the Garrote function threshold denoising, thereby achieving the purpose of extracting effective signal information. Hilbert transform to extract the EMG signal envelope; The Hilbert transform of the electromyographic signal x(t) is defined as: Where H represents the Hilbert change operation, P represents the principal value integral, x(t) is the original signal, is the signal after Hilbert transform of x(t); The analytical signal z(t) is defined as The modulus A(t) of the analytical signal is the envelope of the signal:

3. The dual-threshold decision-making method for spasticity assessment according to claim 1, characterized in that: The joint angle θ corresponding to the first threshold t1 of the EMG signal spasm p is the stretch reflex threshold; the assessment process affects the patient's stretch reflex threshold θ p , maximum angle θ rom Recording, using θ p / θ rom Numerical experiments on the objective evaluation of spasticity level; During the evaluation of flexor spasticity, θ p Indicates the angle from the point where the spasm occurs to the limit of extension; θ rom Indicates the range of motion of the elbow joint of the upper limb during the test, which is defined as the range of motion from the maximum flexion angle to the maximum extension angle; during the assessment of extensor spasm, θ p The angle from the point where the spasm occurs to the flexion limit is calculated in the same way as θ. p / θ rom .

4. A dual-threshold decision-making spasticity assessment device, used for signal acquisition in the dual-threshold decision-making spasticity assessment method according to claim 1, characterized in that: The invention comprises a large arm fixing rod (8), a small arm length adjustment rod assembly (2) rotatably connected to the large arm fixing rod (8) via an elbow joint rotating shaft, an angle measurement sensor (3) mounted on the elbow joint rotating shaft, and an impedance force measurement module (1) mounted on the small arm length adjustment rod assembly (2); the impedance force measurement module (1) comprises an arc-shaped mounting seat (12) mounted on the small arm length adjustment rod assembly (2) and two force sensors (11) respectively mounted at two ends of the arc-shaped mounting seat (12) and arranged opposite to each other; the force sensors (11) are mounted on the arc-shaped mounting seat (12) via a ball joint.

5. The dual-threshold decision-making spasticity assessment device according to claim 4, characterized in that: A fixing frame (4) is fixedly mounted on the upper arm fixing rod (8), and the upper arm of a human body is fixed by a strap mounted on the fixing frame (4).

6. The dual-threshold decision-making spasticity assessment device according to claim 4, characterized in that: It also includes a sEMG sensor (6) for collecting electromyographic signals and a sEMG acquisition card (5) connected to the electrical signals of the sEMG sensor (6); the sEMG sensor (6) has a wet electrode patch (51), and one sEMG sensor (6) is respectively provided on the biceps brachii and the triceps brachii.

7. The dual-threshold decision-making spasticity assessment device according to claim 6, characterized in that: The sEMG sensor (6) is fixedly attached to the upper arm via a strap.

8. The dual-threshold decision-making spasticity assessment device according to claim 6, characterized in that: It also includes a host computer, the sEMG acquisition card (5) is wirelessly connected to the host computer via Bluetooth; the force sensor (11) and the angle measurement sensor (3) are connected to the host computer via RS484 communication.

Images