Search

CN-122025004-A - Fracture healing period rehabilitation guidance system based on multi-mode data acquisition

CN122025004ACN 122025004 ACN122025004 ACN 122025004ACN-122025004-A

Abstract

The invention relates to the technical field of intelligent medical treatment, in particular to a fracture healing period rehabilitation guidance system based on multi-mode data acquisition. According to the invention, the physiological phase of fracture healing is judged and the corresponding biomechanical tolerance boundary is matched by analyzing the frequency drift gradient of the acceleration response signal and checking the nonlinear characteristic of the broken end stiffness, a cooperative characteristic matrix based on a muscle activation intensity envelope sequence is constructed, the main angle of a geometric projection subspace is calculated, the deviation degree of motion control relative to a standard mode is quantitatively evaluated, the compensation risk is identified, the muscle force vector is decoupled into an axial and shearing component and is compared with the dynamic limit value of the healing phase in real time, the evaluation of biomechanical state is realized, and the rehabilitation training is ensured to be efficiently carried out in a safe mechanical window.

Inventors

  • YANG XIAOPING
  • HOU JINGMING
  • ZHOU XIAOPING

Assignees

  • 中国人民解放军陆军军医大学第一附属医院

Dates

Publication Date
20260512
Application Date
20260128

Claims (10)

  1. 1. Fracture healing period rehabilitation guidance system based on multi-mode data acquisition, characterized in that the system comprises: the data acquisition processing module is used for applying sweep frequency excitation to acquire acceleration response signals of the bone ends of the affected limbs, acquiring surface electromyographic signals of target muscle groups and compensatory muscle groups of the affected limbs, generating a muscle activation intensity envelope sequence, acquiring joint angle data of the limbs, converting the joint angle data into Euler angle data, and establishing a physiological characteristic data set; The healing state evaluation module is used for extracting natural frequency parameters of acceleration response signals in the physiological characteristic data set, calculating a slope increasing along with the amplitude of the surface electromyographic signals, defining the slope as a frequency drift gradient, comparing the frequency drift gradient with a linearity standard, judging a fracture healing stage and obtaining a mechanical tolerance boundary standard; The collaborative mode monitoring module is used for extracting the physiological characteristic data set, decomposing surface electromyographic signals to obtain a muscle collaborative characteristic matrix, orthogonalizing the muscle collaborative characteristic matrix to construct a geometrical projection subspace, calculating a geometrical projection subspace included angle as a collaborative subspace main angle, and obtaining an action compensation evaluation index; The mechanical vector decoupling module is used for positioning a muscle force vector based on the physiological characteristic data set and projecting the muscle force vector to a fracture end local coordinate system, decomposing a projection component to obtain an axial resultant force parameter and a shearing resultant force parameter, and integrating the axial resultant force parameter and the shearing resultant force parameter into a stress analysis data set; and the rehabilitation safety analysis module is used for comparing the main angle of the collaborative subspace in the action compensation evaluation index with a preset angle warning standard, comparing the shearing resultant force parameter in the stress analysis data set with the shearing component limit specified in the mechanical tolerance boundary, and constructing a rehabilitation training evaluation record.
  2. 2. The multi-modal data acquisition-based fracture healing period rehabilitation guidance system according to claim 1, wherein the physiological characteristic data set comprises a time-synchronized acceleration response signal sequence, a muscle activation intensity envelope sequence, euler angle format posture data, the mechanical tolerance boundary comprises a bone fracture end allowable axial force limit and a bone fracture end allowable shear force limit, the motion compensation assessment index comprises a collaborative subspace main angle value and a motion control mode deviation quantification value, the stress analysis data set comprises an axial resultant force parameter along the bone axis direction and a shear resultant force parameter perpendicular to the bone axis plane, and the rehabilitation training assessment record comprises a compensation deviation risk level, a fracture end stress overload risk level and a motion compliance state identification.
  3. 3. The multi-modal data acquisition-based fracture healing period rehabilitation guidance system of claim 2, wherein the data acquisition processing module comprises: The vibration response monitoring submodule controls the vibration exciter, continuously applies sweep frequency excitation load to the bone end of the affected limb according to a preset frequency step length and a scanning period, and acquires an acceleration response signal by combining the mechanical vibration waveform of the bone end under the excitation action acquired by the acceleration sensor array; The myoelectricity characteristic extraction submodule is used for collecting original surface myoelectricity signals of a target muscle group of an affected limb and a compensatory muscle group in a rehabilitation action process by utilizing a multichannel myoelectricity electrode, performing DC component removal and full-wave rectification operation on the original surface myoelectricity signals, extracting signal energy characteristics, constructing a moving average time window with a fixed length, performing smooth filtering treatment on rectified signal amplitude, and generating a muscle activation intensity envelope sequence; And the multimode data fusion sub-module is used for collecting quaternion format joint angle data in the limb movement process, converting the quaternion format joint angle data into Euler angle format data according to the rotation sequence, and combining the acceleration response signals and the muscle activation intensity envelope sequence, performing time stamp alignment and resampling synchronous operation on the data stream, and establishing a physiological characteristic data set.
  4. 4. The multi-modal data acquisition based fracture healing period rehabilitation guidance system of claim 3, wherein the healing state assessment module comprises: The frequency response characteristic analysis sub-module analyzes acceleration response signals in the physiological characteristic data set, extracts corresponding first-order natural frequency parameters according to power spectrum density peaks, synchronously analyzes a muscle activation intensity envelope sequence corresponding to a surface electromyographic signal in the data set, acquires an amplitude change sequence, calculates a change slope generated by the natural frequency parameters along with the increase of the amplitude of the muscle activation intensity envelope sequence, and generates a frequency drift gradient; The rigidity nonlinear verification sub-module invokes the frequency drift gradient, performs numerical interval comparison operation with a preset linearity reference range respectively, verifies nonlinear characteristic expression of the broken end rigidity in the muscle force loading process according to a comparison difference value, and outputs a fracture healing stage judging result; And the stress boundary matching sub-module is used for inquiring a corresponding biomechanical safety standard table in a preset rehabilitation database according to the fracture healing stage judging result, extracting the allowable axial force limit of the fracture end and the allowable shearing force limit of the fracture end which are allowed in the current healing stage, and obtaining the mechanical tolerance boundary containing the axial force limit and the shearing force limit.
  5. 5. The multi-modal data acquisition-based fracture healing period rehabilitation guidance system according to claim 4, wherein the process of performing numerical interval comparison operation with a preset linearity reference range and verifying nonlinear characteristic expression of the broken end stiffness in the muscle force loading process according to a comparison difference value is specifically as follows: Invoking a preset clinical pathological feature library, and reading a fiber connection period gradient distribution boundary value, a cartilage callus formation period gradient distribution boundary value and a callus transformation period gradient distribution boundary value stored in the library; Constructing a fiber connection period checking section based on the fiber connection period gradient distribution boundary value, constructing a cartilage callus formation period checking section based on the cartilage callus formation period gradient distribution boundary value, constructing a callus transformation period checking section based on the callus transformation period gradient distribution boundary value, and setting the fiber connection period checking section, the cartilage callus formation period checking section and the callus transformation period checking section as the preset linearity reference range; Executing logic operation of the frequency drift gradient and the numerical value of the preset linearity reference range; if the operation result shows that the frequency drift gradient falls into the fiber connection period check interval, generating a fracture healing stage judgment result marked as a fiber connection period; If the operation result shows that the frequency drift gradient falls into the cartilage scab formation period checking section, generating a fracture healing stage judging result marked as a cartilage scab formation period; if the operation result shows that the frequency drift gradient falls into the correction interval of the callus transformation period, generating a fracture healing stage judging result marked as the callus transformation period; the process for inquiring the corresponding biomechanical safety standard table in the preset rehabilitation database and extracting the allowable axial force limit of the fracture end and the allowable shearing force limit of the fracture end which are allowed in the current healing stage comprises the following steps: invoking a pre-constructed biomechanical safety standard table, wherein the biomechanical safety standard table is a load limiting data set calculated by multiplying a preset rehabilitation safety coefficient by a yield strength critical value of poroma tissues at each healing stage based on statistics in a clinical biomechanical database; Taking the fracture healing stage judging result as an index key, and retrieving corresponding load limit data from the biomechanical safety standard table; extracting an axial load extremum converted by a safety coefficient from the load limiting data as an allowable axial force limit of the fracture end, and synchronously extracting a shearing load extremum converted by the safety coefficient as an allowable shearing force limit of the fracture end; and data packaging the limit of the allowable axial force of the fracture end and the limit of the allowable shearing force of the fracture end to generate the mechanical tolerance boundary.
  6. 6. The multi-modal data acquisition based fracture healing period rehabilitation guidance system of claim 5, wherein the collaborative mode monitoring module comprises: The collaborative matrix extraction submodule invokes the physiological characteristic dataset, extracts a muscle activation intensity envelope sequence and constructs a multi-channel observation matrix, performs matrix decomposition operation on the observation matrix and iteratively solves a base matrix component containing muscle weight information of each channel to obtain a muscle collaborative characteristic matrix; The projection space construction sub-module extracts the cooperative basis vectors of the current action based on the muscle cooperative characteristic matrix, synchronously acquires the standard basis vector matrix in a preset standard action library, orthogonalizes the two groups of basis vectors respectively, and constructs a corresponding geometric projection sub-space; And the deviation quantization sub-module is used for calculating a cosine value of the minimum main included angle aiming at two characteristic planes representing the current action and the standard action in the geometric projection subspace, performing inverse cosine function operation on the cosine value to obtain a main angle of the collaborative subspace, and quantitatively evaluating the deviation degree of the current action control of the patient relative to the standard action to obtain an action compensation evaluation index.
  7. 7. The multi-modal data acquisition-based fracture healing period rehabilitation guidance system according to claim 6, wherein the process of synchronously acquiring the standard basis vector matrix in the preset standard action library is specifically as follows: Invoking a preset standard action collaborative feature library, wherein the standard action collaborative feature library comprises surface electromyographic signal data sets acquired by a plurality of healthy subjects when the healthy subjects strictly follow rehabilitation action specifications; respectively executing non-negative matrix factorization operation on each group of signals in the surface electromyographic signal data set, and extracting an individual basis vector set reflecting an individual muscle activation mode; Performing vector cluster analysis on all the individual basis vector sets, identifying basis vector clusters with similar features, calculating mass vectors of the clusters, and constructing the mass vector combinations into the standard basis vector matrix; The process of quantitatively evaluating the deviation degree of the current action control of the patient relative to the standard action specifically comprises the following steps of: Performing inverse cosine mathematical operation on the cosine value of the minimum main included angle to obtain a collaborative subspace main angle expressed by an angle value; acquiring a preset full-orthogonal limit angle, wherein the full-orthogonal limit angle is defined as a theoretical maximum included angle value when two vector spaces are in a mutually perpendicular state; Calculating the ratio of the main angle of the collaborative subspace to the full orthogonal limit angle to generate a normalized collaborative deviation coefficient; And mapping the collaborative deviation coefficient to a preset numerical evaluation model to generate the action compensation evaluation index.
  8. 8. The multi-modal data acquisition-based fracture healing period rehabilitation guidance system according to claim 7, wherein the mechanical vector decoupling module comprises: The muscle strength estimation submodule invokes a muscle activation strength envelope sequence in the physiological characteristic data set, combines the physiological cross-sectional area parameters of the target muscle group, performs mechanical mapping operation based on a muscle contraction dynamics model, converts the electrophysiological signal amplitude into a muscle tension value, calculates the real-time contraction strength of each active muscle and antagonistic muscle in the action process, and acquires a muscle contraction strength sequence under a corresponding time step; The vector space mapping submodule is used for reconstructing a space pose matrix of a bone of a patient limb based on Euler angle format pose data in the physiological characteristic data set, calling the muscle contraction force strength sequence, combining an anatomical attachment point database to position action point coordinates and action direction vectors of each muscle force, constructing a three-dimensional coordinate system with the fracture broken end center as an origin, converting and mapping force vectors of each cross-joint muscle into the coordinate system, and generating broken end projection force vector components; And the broken end resultant force decoupling sub-module is used for carrying out multidimensional vector synthesis operation on the broken end projection force vector component, obtaining a total force vector acting on the broken end of the bone, carrying out orthogonal decomposition treatment on the total force vector along a bone anatomical axis and a cross section direction perpendicular to the axis, extracting a pressure component along the axial direction and a shearing component in the cross section, integrating an axial resultant force parameter and a shearing resultant force parameter, and establishing a broken end stress analysis data set.
  9. 9. The multi-modal data acquisition based fracture healing period rehabilitation guidance system of claim 8, wherein the rehabilitation security analysis module comprises: the action mode verification sub-module is used for calling the action compensation evaluation index, extracting a main angle value of the collaborative subspace, comparing the main angle value with a preset angle warning standard, calculating the difference value amplitude of the current action deviating from a standard interval, and judging the compensation deviation risk level according to the difference value amplitude; The stress safety verification sub-module is used for calling the broken end stress analysis data set and extracting shearing resultant force parameters, synchronously acquiring allowable shearing force limit of the broken end of the bone from the mechanical tolerance boundary, comparing the shearing resultant force parameters with the allowable shearing force limit of the broken end of the bone, calculating the amplitude proportion of the actual stress value exceeding the safety limit, judging the stress state and generating the stress overload risk level of the broken end; And the comprehensive compliance evaluation sub-module is used for checking stress safety based on the compensation deviation risk level and the broken end stress overload risk level, judging the compliance state of the current action, recognizing whether the compensation deviation risk and the broken end stress overload risk exist in the current action, and establishing a rehabilitation training compliance evaluation record.
  10. 10. The multi-modal data acquisition-based fracture healing period rehabilitation guidance system according to claim 9, wherein the process of calculating the difference amplitude of the current action from the standard interval by comparing with the preset angle alert standard is specifically as follows: Invoking a preset clinical rehabilitation reference database, reading a plurality of groups of historical collaborative subspace main angle samples of healthy subjects stored in the database under a standard rehabilitation action mode, and executing normal distribution fitting operation on the historical collaborative subspace main angle samples to acquire sample distribution mean values and sample standard deviations; acquiring a preset abnormal judgment sensitivity coefficient, calculating an accumulated sum of products of the sample distribution mean value and the sample standard deviation and the abnormal judgment sensitivity coefficient, and establishing the accumulated sum as the preset angle warning standard; Reading the main angle value of the collaborative subspace in the action compensation evaluation index, and executing the difference operation of subtracting the preset angle warning standard from the main angle value of the collaborative subspace to obtain a numerical deviation; Executing one-way truncation verification on the numerical deviation, if the numerical deviation is greater than zero, judging that pathological compensation exists in the current action, and establishing the numerical deviation as the difference amplitude; And if the numerical deviation is smaller than or equal to zero, judging that the current action is in a physiological allowable range, and assigning the difference amplitude to zero.

Description

Fracture healing period rehabilitation guidance system based on multi-mode data acquisition Technical Field The invention relates to the technical field of intelligent medical treatment, in particular to a fracture healing period rehabilitation guidance system based on multi-mode data acquisition. Background The technical field of intelligent medical technology relates to the cross fusion of the Internet of things, artificial intelligence, big data analysis and biomedical engineering, and aims at remodelling a medical service flow by using digital sensing equipment and an information platform, wherein the key matters in the field are that various sensors and medical terminals are used for continuously collecting individual physiological parameters, pathological characteristics and behavior patterns, and a calculation model is used for carrying out mining analysis on massive heterogeneous data, so that remote health monitoring, auxiliary diagnosis and treatment decision making and personalized intervention schemes are provided, the medical patterns are pushed to evolve from disease treatment to full-period health management, wherein a traditional fracture healing period rehabilitation guidance system refers to an auxiliary means which is combined with single imaging examination by relying on manual periodic follow-up, the method mainly aims at the training planning and progress monitoring matters of the fracture postoperative household period of a patient, and is characterized in that the patient is required to go to a hospital according to a fixed period to shoot X-ray films to observe the growth form of poroma, a rehabilitation doctor orally formulates the training action of the next stage according to an image result and a general rehabilitation manual, the patient relies on subjective pain of the patient or uses a simple protractor to measure the joint activity degree during the household rehabilitation period, the daily training frequency and duration are recorded through a paper log, and then the doctor consults a handwriting record and visually observes the on-site limb activity condition of the patient to judge the recovery degree and adjust the follow-up plan. The prior art severely relies on intermittent imaging examination and subjective handwriting journals of patients to conduct rehabilitation process management, the discontinuous monitoring mode cannot capture dynamic changes of mechanical environments of fractured ends in real time, actual internal loads applied to fractured ends by deep muscle contraction are difficult to quantify simply by pain feedback or in-vitro joint angle measurement, targeted adaptation of rigidity nonlinear evolution characteristics in the callus growth process is lacking, implicit compensation behaviors in complex muscle cooperative motions cannot be accurately analyzed, axial stress favorable for healing and shearing stress causing displacement are difficult to distinguish, and the patients face risks of secondary injury caused by stress overload or limb function dissimilarity caused by compensation motions in home training. Disclosure of Invention In order to solve the technical problems in the prior art, the embodiment of the invention provides a fracture healing period rehabilitation guidance system based on multi-mode data acquisition. The technical scheme is as follows: in one aspect, there is provided a fracture healing period rehabilitation guidance system based on multimodal data acquisition, the system comprising: the data acquisition processing module is used for applying sweep frequency excitation to acquire acceleration response signals of the bone ends of the affected limbs, acquiring surface electromyographic signals of target muscle groups and compensatory muscle groups of the affected limbs, generating a muscle activation intensity envelope sequence, acquiring joint angle data of the limbs, converting the joint angle data into Euler angle data, and establishing a physiological characteristic data set; The healing state evaluation module is used for extracting natural frequency parameters of acceleration response signals in the physiological characteristic data set, calculating a slope increasing along with the amplitude of the surface electromyographic signals, defining the slope as a frequency drift gradient, comparing the frequency drift gradient with a linearity standard, judging a fracture healing stage and obtaining a mechanical tolerance boundary standard; The collaborative mode monitoring module is used for extracting the physiological characteristic data set, decomposing surface electromyographic signals to obtain a muscle collaborative characteristic matrix, orthogonalizing the muscle collaborative characteristic matrix to construct a geometrical projection subspace, calculating a geometrical projection subspace included angle as a collaborative subspace main angle, and obtaining an action compensation evaluation index; The mechanical vector