CN-121999976-A - Fracture infant rehabilitation exercise compliance change prediction method and system

CN121999976ACN 121999976 ACN121999976 ACN 121999976ACN-121999976-A

Abstract

The application relates to the technical field of medical information processing, and discloses a method and a system for predicting rehabilitation exercise compliance change of a fracture child patient, wherein the method comprises the steps of constructing a target child patient rehabilitation exercise compliance longitudinal time sequence at a preset follow-up time node; extracting stage transition characteristics and identifying compliance stage states, synchronously constructing a psychological-behavioral multi-index synchronous time sequence, wherein the indexes comprise pain, terrorism and self-efficacy scores, analyzing index change synchronicity and consistency, identifying psychological-behavioral coupling states, cooperatively modeling the two states as input characteristics, establishing an association relation, predicting the significant change risk of compliance in the next follow-up stage based on modeling results, and outputting the results. The method can accurately describe the compliance stepwise change, reveal the internal driving mechanism, has accurate and interpretable prediction result, and provides scientific basis for clinical rehabilitation intervention.

Inventors

ZHANG JINYUAN
HU JINGYAO
WANG XUEYAN

Assignees

内蒙古医科大学

Dates

Publication Date: 20260508
Application Date: 20260127

Claims (10)

1. The method for predicting the rehabilitation exercise compliance change of the children suffering from fracture is characterized by comprising the following steps of: collecting and constructing a rehabilitation exercise compliance longitudinal time sequence of a target child patient at a plurality of preset follow-up time nodes; Based on the rehabilitation exercise compliance longitudinal time sequence, extracting phase transition characteristics reflecting a compliance change process, and identifying a compliance phase state of the target infant currently according to a preset phase division rule; Synchronously collecting and constructing a psychological-behavioral multi-index synchronous time sequence of the target infant at a plurality of time points same as the follow-up time node, wherein the psychological-behavioral multi-index comprises pain scores, terrorism scores and self-efficacy scores; Analyzing the change synchronicity and consistency of each index in the psychological-behavioral multi-index synchronous time sequence in the time dimension, and identifying the current psychological-behavioral coupling state of the target infant according to a preset coupling unbalance judging rule; The compliance stage state and the psychological-behavioral coupling state are used as input features to carry out collaborative modeling, and an association relation between the compliance stage state and the psychological-behavioral coupling state is established; And predicting the risk of the target infant suffering from significant change of compliance in the next follow-up stage based on the result of the collaborative modeling, and outputting a prediction result.
2. The method for predicting the rehabilitation exercise compliance change of a patient suffering from fracture according to claim 1, wherein, The rehabilitation exercise compliance longitudinal time sequence is constructed by acquiring compliance score data acquired based on an orthopedic patient function exercise compliance scale at each follow-up time node, carrying out structural processing on the compliance score data to form a numerical sequence containing a compliance total score and a dimensionality score, and arranging the numerical sequences of all time points according to the sequence of the follow-up time nodes.
3. The method for predicting the rehabilitation exercise compliance change of a patient suffering from fracture according to claim 2, wherein, The stage transition feature comprises at least one of the following change amplitude of the compliance index between the adjacent follow-up nodes, the consistency of the change direction of the compliance index in the continuous multiple follow-up nodes, and the amplifying or convergence trend of the change amplitude between the continuous multiple adjacent follow-up nodes.
4. The method for predicting the rehabilitation exercise compliance change of a patient suffering from fracture according to claim 3, The method comprises the steps of identifying the compliance stage state according to a preset stage division rule, wherein the stage division rule divides the compliance stage state into a stable high compliance stage, a fluctuation transition stage and a high risk descent stage, the identification process comprises the steps of comparing the extracted stage transition characteristic with a threshold value or a condition set in the stage division rule, and determining the stage state of the target infant according to a comparison result.
5. The method for predicting the rehabilitation exercise compliance change of a patient suffering from fracture according to claim 1, wherein, The analysis of the synchronicity and consistency of the changes specifically includes at least one of analyzing whether the direction of the change in pain score is consistent with the direction of the change in terrorist score, analyzing whether the change in self-efficacy score remains synchronized with the improvement in function and the relief of pain, and analyzing whether there is a lag between the change in psychological index and the change in compliance that exceeds a preset time interval.
6. The method for predicting the rehabilitation exercise compliance with fracture infant according to claim 5, wherein, The method specifically comprises the steps of setting at least one judging condition of a coupling unbalance type according to a preset coupling unbalance judging rule, judging that a target infant is in a psychological-behavioral coupling unbalance state when the change of the psychological-behavioral multi-index synchronous time sequence meets any judging condition, recording the corresponding unbalance type, and judging that the infant is in a psychological-behavioral coupling balance state otherwise.
7. The method for predicting the rehabilitation exercise compliance change of a patient suffering from fracture according to claim 1, wherein, The collaborative modeling adopts a logistic regression model, wherein an input feature vector of the logistic regression model comprises the compliance stage state feature and the psychological-behavior coupling state feature, and the output of the logistic regression model is the prediction probability of the target infant to generate high-risk stage transition; the establishment of the association relation specifically comprises the step of determining the influence degree and direction of different types of psychological-behavioral coupling states on the transition of the compliance stage through the characteristic weight coefficient obtained through training of the logistic regression model.
8. A system for predicting compliance with rehabilitation exercise for children suffering from fracture, comprising: The compliance longitudinal time sequence construction module is used for acquiring and constructing a rehabilitation exercise compliance longitudinal time sequence of the target infant at a plurality of preset follow-up time nodes; the feature extraction and state identification module is used for extracting phase transition features reflecting the compliance change process based on the rehabilitation exercise compliance longitudinal time sequence and identifying the compliance phase state of the target infant according to a preset phase division rule; The data acquisition module is used for synchronously acquiring and constructing a psychological-behavioral multi-index synchronous time sequence of the target infant at a plurality of time points which are the same as the follow-up time node, wherein the psychological-behavioral multi-index comprises pain scores, terrorism scores and self-efficacy scores; The coupling state identification module is used for analyzing the change synchronicity and consistency of each index in the psychological-behavioral multi-index synchronous time sequence in the time dimension and identifying the current psychological-behavioral coupling state of the target infant according to a preset coupling unbalance judgment rule; the association relation establishing module is used for carrying out collaborative modeling by taking the compliance stage state and the psychological-behavioral coupling state as input characteristics, and establishing an association relation between the compliance stage state and the psychological-behavioral coupling state; And the result output module is used for predicting the risk of the target infant suffering from significant change of compliance in the next follow-up stage based on the collaborative modeling result and outputting a prediction result.
9. The system for predicting the compliance with rehabilitation exercise for children with fracture according to claim 8, wherein, The compliance longitudinal time sequence construction module is specifically used for acquiring compliance score data acquired based on an orthopedic patient functional exercise compliance scale at each follow-up time node, carrying out structural processing on the compliance score data to form a numerical sequence containing a compliance total score and a dimensionality score, and arranging the numerical sequences of all time points according to the sequence of the follow-up time nodes to construct the rehabilitation exercise compliance longitudinal time sequence.
10. The system for predicting the compliance with rehabilitation exercise for children with fracture according to claim 8, wherein, The stage transition feature comprises at least one of the following change amplitude of the compliance index between the adjacent follow-up nodes, the consistency of the change direction of the compliance index in the continuous multiple follow-up nodes, and the amplifying or convergence trend of the change amplitude between the continuous multiple adjacent follow-up nodes.

Description

Fracture infant rehabilitation exercise compliance change prediction method and system Technical Field The application relates to the technical field of medical information processing, in particular to a method and a system for predicting rehabilitation exercise compliance change of children suffering from fracture. Background At home recovery stage after the operation of the GARTLAND III-stage humeral epicondylitis fracture of the school-age child, the recovery exercise compliance is a key factor affecting the recovery effect of the elbow joint function, and is directly related to the occurrence risk of long-term complications. Clinical practice shows that the compliance of the rehabilitation exercise of the children is not changed linearly with time smoothly, but enters into key time windows such as active activity stage after removing gypsum, removing kirschner wires and relieving pain, and the like, and the compliance presents obvious stepwise fluctuation or sudden drop characteristics, thus bringing great challenges to rehabilitation management. In the prior art, the prediction method for the rehabilitation exercise compliance of the children patients has a plurality of defects. Most schemes consider compliance as continuous linear change indexes, a single time point scoring or simple longitudinal trend analysis is adopted, high risk nodes which are about to be remarkably reduced are difficult to identify, a part of the methods consider pain, self-efficacy, fear exercises and other influencing factors, but are mostly used as mutually independent variable analysis, the dynamic coupling relation of each factor in the time dimension is ignored, the clinical common problems that the pain is relieved but the compliance is reduced are not explained, and the like are difficult to provide clear and accurate basis for clinical nursing follow-up and rehabilitation intervention because the traditional prediction scheme stays at a result prediction layer, lacks the interpretation capability of an intrinsic driving mechanism of compliance stage mutation. Disclosure of Invention The application aims to provide a fracture infant rehabilitation exercise compliance change prediction method and system, which are used for solving the problems in the background technology. According to a first aspect of the present application, there is provided a method for predicting rehabilitation exercise compliance variation of a patient suffering from fracture, comprising the steps of: collecting and constructing a rehabilitation exercise compliance longitudinal time sequence of a target child patient at a plurality of preset follow-up time nodes; Based on the rehabilitation exercise compliance longitudinal time sequence, extracting phase transition characteristics reflecting a compliance change process, and identifying a compliance phase state of the target infant currently according to a preset phase division rule; Synchronously collecting and constructing a psychological-behavioral multi-index synchronous time sequence of the target infant at a plurality of time points same as the follow-up time node, wherein the psychological-behavioral multi-index comprises pain scores, terrorism scores and self-efficacy scores; Analyzing the change synchronicity and consistency of each index in the psychological-behavioral multi-index synchronous time sequence in the time dimension, and identifying the current psychological-behavioral coupling state of the target infant according to a preset coupling unbalance judging rule; The compliance stage state and the psychological-behavioral coupling state are used as input features to carry out collaborative modeling, and an association relation between the compliance stage state and the psychological-behavioral coupling state is established; And predicting the risk of the target infant suffering from significant change of compliance in the next follow-up stage based on the result of the collaborative modeling, and outputting a prediction result. The method comprises the steps of obtaining compliance score data collected based on an orthopedic patient functional exercise compliance scale at each follow-up time node, carrying out structural processing on the compliance score data to form a numerical sequence containing a compliance total score and a dimensionality score, and arranging the numerical sequences of all time points according to the sequence of the follow-up time nodes to construct the rehabilitation exercise compliance longitudinal time sequence. Preferably, the phase transition feature comprises at least one of the following change amplitude of the compliance indicators between adjacent follow-up nodes, the consistency of the change directions of the compliance indicators in a plurality of follow-up nodes, and the amplifying or converging trend of the change amplitude between the plurality of adjacent follow-up nodes. The method comprises the steps of identifying the compliance phase state