CN-122025128-A - ICF-RS-based rehabilitation state parameter generation method

CN122025128ACN 122025128 ACN122025128 ACN 122025128ACN-122025128-A

Abstract

The invention relates to the technical field of biomedical engineering and discloses a rehabilitation state parameter generation method based on ICF-RS. The method comprises the steps of acquiring myoelectricity and torque signals of a target muscle group and corresponding joints at a unified starting time, synchronously acquiring the myoelectricity and torque signals according to a preset frequency, carrying out statistical standardization and integral of the myoelectricity signals, carrying out absolute average on the torque signals for whole seconds, pairing the torque signals with the myoelectricity signals for whole seconds, establishing a first-order linear dynamic model which takes two signal sequences as input and contains a natural fall-back term, carrying out dispersion according to a whole second step length to generate a state sequence, calculating an average state change rate according to the end state and the total time length of the sequence, comparing a reference rate and converting the average state change rate into a numerical value grading parameter, mapping the grading parameter to walking capacity and muscle endurance grade in an ICF rehabilitation scale to generate corresponding function grade parameter, archiving second grade data, the state sequence and the generated series parameter, and calculating the change rate according to days to generate trend characterization parameter.

Inventors

YAN TIEBIN
SUN QIANQIAN
LIU JIEMING
ZHAN YAN

Assignees

中山大学孙逸仙纪念医院

Dates

Publication Date: 20260512
Application Date: 20260128
Priority Date: 20251022

Claims (9)

1. The ICF-RS-based rehabilitation state parameter generation method is characterized by comprising the following steps of: Acquiring myoelectric signals and joint torque signals synchronously acquired at unified starting time according to preset sampling frequency by a surface myoelectricity acquisition unit arranged at the position of a target muscle group and a torque acquisition unit arranged at the position of a corresponding joint, wherein the signals are original data continuously recorded in a preset data acquisition period; carrying out statistical standardization processing on the collected electromyographic signals, and integrating according to a whole second time window to form an electromyographic integration sequence; Absolute average smoothing is carried out on the joint torque signals within a whole second window to form a second-level torque sequence, and the second-level torque sequence corresponds to the myoelectric integration sequence one by one on a second-level time axis; defining a state function reflecting the level of the motion function in a continuous time domain, establishing a first-order linear dynamic model which takes myoelectricity integral and second-level torque as input and takes an attenuation term as constraint, and evolving the model by adopting piecewise constant input in each whole second interval; Performing numerical discrete on the dynamic model by adopting an Euler method with whole second as a step length to generate state sequence parameters covering the whole data acquisition period; Calculating the average state change rate according to the state value at the end of data acquisition and the total duration of data acquisition, comparing the average state change rate with a preset reference rate, and converting the average state change rate into a numerical value scoring parameter; mapping the numerical scoring parameters to two indexes of walking ability and muscle endurance in an ICF-RS system according to the numerical scoring parameters so as to generate corresponding function grade parameters; And archiving second-level data, state sequence parameters, numerical scoring parameters and functional grade parameters to an electronic data file, forming a scoring parameter sequence according to days, and calculating based on the change rate of the sequence to obtain the rehabilitation trend characterization parameters.
2. The ICF-RS-based rehabilitation state parameter generating method according to claim 1, wherein the acquiring the electromyographic signal and the joint torque signal specifically includes: the surface myoelectricity electrode attached to the center of the quadriceps on one side of the subject was obtained and recorded as electrode number Output original myoelectric voltage sequence Wherein, the method comprises the steps of, Is the first collected by the electrode system Myoelectric voltages at the individual sample points; indexing myoelectricity sampling points; the total myoelectricity sampling points are; acquiring a torque sensor arranged at the center of the knee joint on the same side of the subject and recording the torque sensor as a sensor number Output joint torque signal sequence Wherein, the method comprises the steps of, Is the first collected by the torque channel Joint torque at each sample point; Indexing the joint torque sampling points; The total number of torque sampling points is the total number of torque sampling points; Setting myoelectricity sampling frequency as Corresponding to the sampling period ; Setting the torque sampling frequency as Corresponding to the sampling period ; Set the collective duration of single data acquisition as Then the total myoelectricity sample is Total number of torque samples is ; Setting the starting time as The two acquisition systems are synchronously started, and the data sequence of the whole period is recorded 、。
3. The ICF-RS-based rehabilitation state parameter generating method according to claim 2, wherein the statistical normalization processing is performed on the collected myoelectric signals, and integration is performed according to a whole second time window to form a myoelectric integration sequence, and specifically comprising: calculating original electromyographic signal sequence Mean of (2) And standard deviation of : , ; Normalizing the original signal and taking the absolute value to obtain a normalized myoelectric value sequence : Wherein, the method comprises the steps of, To myoelectricity first Normalized absolute values of the individual samples; Setting myoelectric integration window length to And (3) points, constructing a continuous integral function sequence: Wherein, the method comprises the steps of, Is the first Myoelectric integral for seconds; the value range is 1 to 60 for the whole second window.
4. The ICF-RS-based rehabilitation state parameter generating method according to claim 3, wherein the absolute average smoothing of the joint torque signal is performed within a whole second window to form a second-level torque sequence, and the second-level torque sequence corresponds to the myoelectric integration sequence one by one on a second-level time axis, and specifically comprising: setting the torque window length to be And (3) performing absolute value arithmetic average processing on the torque signal: Wherein, the method comprises the steps of, Is the first Absolute average torque in seconds; is the first Absolute values of the individual torque samples; And myoelectric integration results Completing one-to-one correspondence, and obtaining a time alignment data sequence: 。
5. The ICF-RS based rehabilitation state parameter generating method according to claim 4, wherein the defining a state function reflecting a movement function level in a continuous time domain, establishing a first-order linear dynamic model using myoelectric integral and second-level torque as inputs and using attenuation terms as constraints, and evolving the model using piecewise constant inputs in each whole second interval, specifically includes: Setting the state variable function as Define a domain as Wherein, the method comprises the steps of, As to continuous time Is a function of (2); is a continuous time variable; Setting an initial state value ; Constructing a state function dynamic evolution differential equation: ; wherein the constant value is segmented in whole seconds: , , , ; Wherein, the Is the myoelectricity gain coefficient; is a torque gain coefficient; Is the attenuation coefficient; And (3) with As a piecewise constant function, in Taking the constants respectively , ; Setting the discrete step length of the model as ; Is provided with , For any non-empty set The first three indexes of the sequence from small to large in time are recorded as If the elements are less than three, only 1 or 2 elements which are actually existed are taken, wherein, Is that Is a positive index set of (1); Is that Is a positive index set of (1); Setting myoelectricity reference quantity as : Wherein, the method comprises the steps of, Is a collection Is the first of (2) A minimum index; ; Is a collection Number of medium elements; Is a ternary median function and is used for the control of the phase shift, To pair(s) A centering value after sequencing; setting the torque reference quantity as : Wherein, the method comprises the steps of, Is a collection Is the first of (2) A minimum index; Is a collection The number of elements in (a); gain and attenuation coefficients are obtained from this: , , 。
6. The ICF-RS based rehabilitation state parameter generating method according to claim 5, wherein the performing numerical discrete on the dynamic model with the step length of whole second by using an euler method to generate a state sequence parameter covering the whole data acquisition period, and obtaining a state estimation result at each moment according to the state sequence parameter, specifically includes: Using Euler's discrete method in step length Performing dispersion to obtain: , Wherein, the method comprises the steps of, Is the first A state value at the end of second; Repeating the calculation to obtain a complete state sequence 。
7. The ICF-RS-based rehabilitation state parameter generating method according to claim 6, wherein calculating an average state change rate according to the state value at the end of the evaluation and the total evaluation duration, comparing the rate with a preset reference rate, and converting the rate into a numerical scoring parameter, and specifically comprising: take the final state value ; Calculating the average state change rate ; Setting the reference rate to , 。
8. The method for generating rehabilitation status parameters based on ICF-RS according to claim 7, wherein the mapping the numerical scoring parameters to two indicators of walking ability and muscular endurance in ICF-RS system to generate corresponding function grade parameters specifically includes: Selecting Two evaluation indexes are Walking ability and walking ability Muscle endurance function; According to the score Assignment: Wherein, the method comprises the steps of, Is that ; Will be Simultaneous assignment to And (3) with And two, forming the function grade parameter output.
9. The ICF-RS-based rehabilitation state parameter generating method according to claim 8, wherein the archiving the second level data, the state sequence parameter, the numerical scoring parameter and the function level parameter to the electronic data file, forming a scoring parameter sequence by day, and calculating a rehabilitation trend characterization parameter based on the change rate of the sequence, specifically includes: uniformly packaging and archiving information of myoelectric sampling integral data per second into an electronic data file of a subject Torque data State variable Scoring of Mapping A grade; If continuous Day-to-day data acquisition, noted as scoring sequence Wherein, the method comprises the steps of, Is the first Scoring the day; Is a date index; setting the trend change rate parameter as The specific calculation method comprises the following steps: ; If it is The parameter characterization score is in an ascending trend; If it is The parameter characterization score is in a stable trend; If it is The parameter characterization score is in a decreasing trend.

Description

ICF-RS-based rehabilitation state parameter generation method Technical Field The invention relates to the technical field of biomedical engineering, in particular to a rehabilitation state parameter generation method based on ICF-RS. Background In the fields of rehabilitation medicine, sports science and the like, quantitative analysis of human body functional states is a key technical link for evaluating intervention effects and optimizing training schemes. The traditional analysis method depends on subjective scales and observation of professionals, and the method has the problems of strong subjectivity and poor repeatability, and is difficult to accurately capture the fine change of the functional state. With the development of sensing technology, functional quantification by using objective physiological signals such as surface electromyographic signals (EMG) and joint torque becomes a research hotspot. The EMG signals can reflect the electrophysiological activity of the muscles, and the joint torque directly represents the mechanical output capability of the limbs. The simultaneous acquisition of these two signals provides a data basis for a deep understanding of the neuro-muscular-mechanical coupling relationship. At the same time, the international functional, disability and health classification (ICF) and its rehabilitation scale (ICF-RS) provide a standardized framework for functional description. However, the automatic and quantitative conversion of objectively acquired physiological signals into parameters conforming to the ICF-RS framework remains a technical challenge. Most of the existing methods stay at the primary processing and statistics level of signals, lack of a complete technical path which can fuse and convert multi-mode signals into state variables with definite physical meanings and finally generate parameter sets compatible with a standardized scale. Thus, there is a great need in the art for a method that systematically processes raw physiological signals to generate a series of quantitative parameters that can be used to characterize the state of recovery. The invention aims to provide a method for opening a technical link from physiological signal acquisition to dynamic state modeling and then to standardized parameter output. Disclosure of Invention The invention provides a rehabilitation state parameter generation method based on ICF-RS, which facilitates solving the problems mentioned in the background art. The invention provides a rehabilitation state parameter generation method based on ICF-RS, which comprises the following steps: Acquiring myoelectric signals and joint torque signals synchronously acquired at unified starting time according to preset sampling frequency by a surface myoelectricity acquisition unit arranged at the position of a target muscle group and a torque acquisition unit arranged at the position of a corresponding joint, wherein the signals are original data continuously recorded in a preset data acquisition period; carrying out statistical standardization processing on the collected electromyographic signals, and integrating according to a whole second time window to form an electromyographic integration sequence; Absolute average smoothing is carried out on the joint torque signals within a whole second window to form a second-level torque sequence, and the second-level torque sequence corresponds to the myoelectric integration sequence one by one on a second-level time axis; defining a state function reflecting the level of the motion function in a continuous time domain, establishing a first-order linear dynamic model which takes myoelectricity integral and second-level torque as input and takes an attenuation term as constraint, and evolving the model by adopting piecewise constant input in each whole second interval; Performing numerical discrete on the dynamic model by adopting an Euler method with whole second as a step length to generate state sequence parameters covering the whole data acquisition period; Calculating the average state change rate according to the state value at the end of the evaluation and the total evaluation time, comparing the average state change rate with a preset reference rate, and converting the average state change rate into a numerical value scoring parameter; mapping the numerical scoring parameters to two indexes of walking ability and muscle endurance in an ICF-RS system according to the numerical scoring parameters so as to generate corresponding function grade parameters; And archiving second-level data, state sequence parameters, numerical scoring parameters and functional grade parameters to an electronic data file, forming a scoring parameter sequence according to days, and calculating based on the change rate of the sequence to obtain the rehabilitation trend characterization parameters. Optionally, the acquiring the electromyographic signal and the joint torque signal specifically includes: the surface m