Search

CN-122025139-A - Nerve-mediated syncope risk prediction method and system

CN122025139ACN 122025139 ACN122025139 ACN 122025139ACN-122025139-A

Abstract

The invention discloses a nerve-mediated syncope risk prediction method and a system, which relate to the technical field of medical evaluation, and the method comprises the steps of firstly dividing units according to the identity type of a patient and the two dimensions of a syncope scene, and constructing a scene-syncope association model according to historical syncope records; by multi-dimensional data acquisition, matching the cause characteristics corresponding to the identity types of the user, locking the associated suspicious syncope scene, predicting the syncope risk of the user, providing intervention measures for the user after the risk exists, monitoring the intervention condition, simultaneously carrying out corresponding feedback, the method can identify a plurality of syncope scenes and a plurality of identity types, the collaborative risk is quantized, the accuracy of risk level judgment is guaranteed, the prediction accuracy is greatly improved, early and high-accuracy risk prediction is realized, scene-identity-intervention association rules are established, individuation and pertinence of intervention measures are improved, the intervention effect is guaranteed, and therefore the safety of users is improved.

Inventors

  • ZHANG LI
  • WANG BIN
  • WU JIAWEI
  • ZHANG ROU
  • JIANG YUXUAN
  • ZHENG DONGZE

Assignees

  • 汕头大学医学院第一附属医院

Dates

Publication Date
20260512
Application Date
20260203

Claims (10)

  1. 1. A method of predicting the risk of neurological syncope, comprising the steps of: S1, establishing data, namely screening an effective syncope data set from a history syncope record, and constructing a scene-syncope database by utilizing the effective syncope data set; S2, risk prediction, namely when monitoring a user, firstly identifying scene characteristics of the user, monitoring the user based on the scene characteristics to obtain an associated data set of the user, predicting the nerve-mediated syncope risk of the user according to a scene-syncope database, the scene characteristics of the user and the associated data set, and executing S3 if the prediction result is that the risk exists; and S3, risk intervention, namely providing syncope intervention measures for the user based on a scene-syncope database, monitoring the user in real time, acquiring a related data set of the user intervention, analyzing the nerve-mediated syncope risk level of the user after the intervention, and then carrying out risk feedback.
  2. 2. The method for predicting the risk of neurological mediated syncope according to claim 1, wherein the specific process of S1 is: s1-1, screening each patient with complete recorded basic information, NMS (network system) confirmed diagnosis, scene characteristics of syncope attacks, associated data, intervention measures and intervention effective grades from historical syncope records as each effective patient, and taking the basic information, NMS confirmed diagnosis, scene characteristics of syncope attacks, cause characteristics, associated data, intervention measures and intervention effective grades of each effective patient as an effective syncope data set; S1-2, classifying each effective patient according to basic information to obtain patient identity types, and classifying scene features and incentive features of syncope attacks of each effective patient in each patient identity type through a K-means clustering algorithm to form various syncope scenes; S1-3, extracting associated data, intervention measures and intervention effective grades of each effective patient in each syncope scene corresponding to each patient identity type, taking a union set of each associated parameter in the associated data of each effective patient as an associated data set, and calculating the weight of each associated parameter in the associated data set; s1-4, forming a scene-syncope database by the associated data set of each patient identity type corresponding to each syncope scene, the weight of each associated parameter in the associated data set and the reference intervention effective level of each intervention measure.
  3. 3. A method of predicting the risk of neurological mediated syncope according to claim 2, wherein said identifying the reference intervention significance level for each intervention is by: Counting the number of effective patients to which each intervention measure is applied, calculating the application frequency of each intervention measure, simultaneously counting the application frequency of each intervention effective level of each intervention measure, dividing the application frequency of each intervention effective level of each intervention measure by the application frequency of each intervention measure, and obtaining the effective weight of each intervention effective level of each intervention measure; counting the application frequency of the intervention effective grade in each intervention measure which is larger than the preset intervention effective grade threshold, wherein the application frequency is used as the standard reaching frequency, and the effect standard reaching weight of each intervention measure is = (standard reaching frequency of each intervention measure/application frequency of each intervention measure) multiplied by 0.5+ (standard reaching frequency of each intervention measure/total standard reaching frequency of all intervention measures) multiplied by 0.5; Multiplying each intervention effective level of each intervention measure by the weight of the corresponding intervention effective level, multiplying the weight of the corresponding intervention effective level by the effect standard reaching weight of the corresponding intervention measure, and accumulating to obtain the confidence coefficient of each intervention measure; The confidence coefficient is larger than or equal to a preset confidence coefficient threshold value, namely, the intervention effective grade with the effective weight larger than 0.5 exists and is directly used as a reference intervention effective grade, and if the intervention effective grade with the maximum effective weight does not exist, the intervention effective grade with the maximum effective weight is selected as the reference intervention effective grade; the confidence coefficient is smaller than a preset confidence coefficient threshold value, namely selecting the intervention effective grade corresponding to the maximum effective weight as the reference intervention effective grade.
  4. 4. The method for predicting the risk of neurological mediated syncope according to claim 1, wherein the specific process of S2 is: s2-1, acquiring physiological data and activity intensity through intelligent wearable equipment worn by a user, acquiring a space type through positioning service, acquiring real-time environment data through an environment sensor module, then matching scene characteristics of the user, extracting basic information of the user, and acquiring an identity type and various suspicious syncope scenes; S2-2, acquiring associated data sets of various suspicious syncope scenes corresponding to the user identity types from a scene-syncope database, and acquiring the associated data sets of various suspicious syncope scene monitoring by using intelligent wearable equipment; s2-3, analyzing the syncope attack probability of the user in various suspicious syncope scenes by utilizing the associated data set monitored by the various suspicious syncope scenes, predicting the nerve-mediated syncope risk level of the user, judging that the predicted result is the risk if the nerve-mediated syncope risk level is more than or equal to level 2, and executing S3, otherwise, judging that the predicted result is safe.
  5. 5. The method for predicting the risk of neurological syncope according to claim 4, wherein the specific process of acquiring identity type and various suspicious syncope scenes is as follows: Dividing the basic information of the user into identity data and illness state data according to inherent attribute and health association in two dimensions, and forming an identity type by the identity data and the illness state type; acquiring various syncope scenes corresponding to the user identity types from a scene-syncope database, and acquiring corresponding incentive features from the various syncope scenes to serve as associated incentive features; the method comprises the steps of obtaining associated data intervals corresponding to associated incentive features from a data center, comparing physiological data, activity intensity and real-time environment data with the associated data intervals corresponding to the associated incentive features, obtaining the matching degree of a user and the associated incentive features, taking the associated incentive features with the matching degree larger than a preset matching degree threshold as the associated incentive features, and combining scene features of the user and the associated incentive features to obtain various suspicious syncope scenes.
  6. 6. The method for predicting the risk of neurological syncope according to claim 4, wherein the analyzing the syncope attack probability of the user in each type of suspicious syncope scene comprises the following steps: Extracting relevant data sets corresponding to various suspicious syncope scenes from a scene-syncope database, comparing the relevant data sets with the relevant data sets monitored by various suspicious syncope scenes, and if a certain relevant parameter in the relevant data sets monitored by certain suspicious syncope scenes is in an abnormal interval of the relevant parameter in the relevant data sets, judging the relevant parameter as abnormal, so as to count all relevant parameters corresponding to the abnormal conditions of various suspicious syncope scenes; Weights of all associated parameters of corresponding anomalies of all suspicious syncope scenes are extracted from the scene-syncope database, and then accumulated, and the results are used as syncope attack probabilities of all suspicious syncope scenes.
  7. 7. The method for predicting the risk of neurological mediated syncope according to claim 6, wherein the predicting the level of neurological mediated syncope risk for a user comprises: acquiring the nerve-mediated syncope risk level of each patient in each suspicious syncope scene corresponding to the user identity type from the history syncope record, and then carrying out weighted average calculation to serve as the average nerve-mediated syncope risk level of each suspicious syncope scene; Acquiring the intervention effective level of each patient in each suspicious syncope scene corresponding to the user identity type from the history syncope record, and then carrying out weighted average calculation to serve as the average intervention effective level of each suspicious syncope scene; Normalizing the faint attack probability, the average nerve-mediated faint risk level and the average intervention effective level of various suspicious faint scenes, setting the scene risk weights of the various suspicious faint scenes, multiplying the faint attack probability of the various suspicious faint scenes by the scene risk weights, and accumulating to obtain the result as a nerve-mediated faint risk characteristic value of the user; Comparing the nerve-mediated syncope risk characteristic value of the user with a preset characteristic value interval corresponding to each nerve-mediated syncope risk level, and taking the nerve-mediated syncope risk level of the characteristic value interval where the nerve-mediated syncope risk characteristic value is located as the nerve-mediated syncope risk level of the user.
  8. 8. The method for predicting the risk of neurological mediated syncope according to claim 7, wherein the step of providing the user with syncope intervention comprises the steps of: The method comprises the steps of taking various suspicious syncope scenes with syncope attack probability more than or equal to 50% as various high-risk scenes, selecting the intervention measures with the largest reference intervention effective level in the various high-risk scenes from a scene-syncope database as various intervention measures, taking various suspicious syncope scenes with syncope attack probability more than or equal to 30% as various medium-risk scenes, selecting the intervention measures with the reference intervention effective level as intermediate values in the various medium-risk scenes from the scene-syncope database as various intervention measures, and taking the intersection of the various intervention measures and the various intervention measures as syncope intervention measures.
  9. 9. The method for predicting risk of neurological mediated syncope according to claim 8, wherein the specific process of risk feedback is: Executing each type of intervention measures in the syncope intervention measures, executing each type of intervention measures, and then collecting the associated data set of each high-risk scene and the associated data set of each medium-risk scene; analyzing the syncope attack probability of each high-risk scene and each medium-risk scene by using the associated data set of each high-risk scene and the associated data set of each medium-risk scene, and then calculating the nerve-mediated syncope risk level of the user after intervention; the nerve-mediated syncope risk level=1 of a plurality of prognosis users, and judging that the syncope intervention of the users is successful, and meanwhile, judging that the intervention effective level of syncope intervention measures is 3; if the nerve-mediated syncope risk level of the user=3 grade, the nerve-mediated syncope risk level of the user=2 grade after intervention, judging that the syncope part of the user is successfully interfered, and simultaneously judging that the effective intervention grade of the syncope intervention measures is 2 grade; If the nerve-mediated syncope risk level of the user=3, the nerve-mediated syncope risk level of the user=3 after intervention, judging that the syncope intervention of the user fails, and judging that the effective intervention level of the syncope intervention measures is 1; If the nerve-mediated syncope risk level of the user=2, the nerve-mediated syncope risk level of the user=2 or 3 after intervention, judging that the syncope intervention of the user fails, judging that the intervention effective level of the syncope intervention measure is 1, and simultaneously sending the user syncope intervention result and the intervention effective level of the syncope intervention measure to a control terminal for syncope intervention result prompt.
  10. 10. A system for predicting the risk of a neurological-mediated syncope performed using the method for predicting the risk of a neurological-mediated syncope according to any one of claims 1 to 9, comprising the following modules: the data establishing module is used for screening out an effective syncope data set from the historical syncope records and constructing a scene-syncope database by utilizing the effective syncope data set; The risk prediction module is used for firstly identifying scene characteristics of a user when the user is monitored, monitoring the user based on the scene characteristics to obtain an associated data set of the user, predicting the nerve-mediated syncope risk of the user according to the scene-syncope database, the scene characteristics of the user and the associated data set, and executing the risk intervention module if the predicted result is that the risk exists; The risk intervention module is used for acquiring historical syncope intervention records of the user, providing syncope intervention measures for the user based on a scene-syncope database, monitoring the user in real time, acquiring a related data set of the user intervention, analyzing the nerve-mediated syncope risk level of the user after intervention, and then carrying out risk feedback.

Description

Nerve-mediated syncope risk prediction method and system Technical Field The invention relates to the technical field of medical evaluation, in particular to a method and a system for predicting the risk of nerve-mediated syncope. Background Nerve-mediated syncope (NMS) is a common syncope type, the onset of which is closely related to life style and environmental factors, single drug treatment effect is limited, long-term management needs to rely on incentive avoidance and behavior adjustment, risk prediction can accurately identify risks through long-term dynamic monitoring, personalized prevention suggestions are pushed, and accidental injury risks are reduced through pre-onset prevention. The prior art discloses a system and a method for automatically detecting vital signs by a handheld PDA (personal digital assistant) with publication number CN117850601A, wherein the system comprises a vital sign intelligent identification module, a threshold adjustment module, an energy efficiency optimization module, a depth insight analysis module, a self-adaptive decision support module, a health state prediction module and an intervention suggestion module. In the invention, the signal processing technology, the deep learning algorithm and the federal learning method are adopted, so that the processing precision and the intelligence level of data are improved, the combination of wavelet transformation and a convolutional neural network is adopted, the key space features in vital sign data are effectively extracted, the long-term and short-term memory network is adopted, the recognition capability of parameters such as heart rate, respiratory rate and the like is enhanced, the Q learning algorithm and the deep Q network are applied, the early warning threshold and the monitoring strategy can be automatically adjusted according to real-time data and environmental changes, and the random forest and the application of a support vector machine algorithm provide a solid foundation for formulating personalized intervention suggestions. The prior art discloses a disease prediction system as disclosed in the application with the publication number of CN117612698A, which comprises a data acquisition module, a calculation module, a judgment execution module and a judgment module, wherein the data acquisition module is used for acquiring individual attributes and clinical data samples of patient group members and carrying out normalization processing, the calculation module is used for calculating the difference degree of the individual attributes and the clinical data of the patient group members and judging whether the calculated difference degree of the individual attributes and the clinical data of the patient group members is smaller than a preset reference factor, the judgment execution module is used for constructing a consensus function decision model and carrying out group disease prediction based on the consensus function decision model if the calculated difference degree of the individual attributes and the clinical data of the patient group members is smaller than the preset reference factor, and the judgment module is used for utilizing the reinforcement learning method training data based on logistic regression to establish a fusion model and carrying out group disease prediction on the similarity of the individual attributes of the patient group members. The application can improve the prediction accuracy. Aiming at the scheme, the method has the following defects that 1, the nerve-mediated syncope is closely related to the state of a user and the space environment where the user is located, but the prior art lacks a clear environment scene recognition and a complete identity type construction mechanism, also lacks a scene-syncope association model, and cannot directly associate a user activity scene with a potential risk, so that a high-risk scene cannot be actively recognized, and scene-related physiological fluctuation and syncope precursor signals cannot be distinguished, so that the miss rate is remarkably increased. 2. The cause and risk level of NMS are strongly related to people, such as NMS induced by teenager high-altitude body position change and exercise, NMS induced by senior high-altitude basic diseases and long-standing, NMS easy to appear after meal hypotension due to blood volume change of pregnant women, lack of complete identity type construction, and incapability of adapting to risk characteristics of different people. 3. The scene and physiological data are dual cores for evaluating NMS risks, and lack of scene-syncope association analysis leads to lack of scientific basis for calculating the attack probability, so that the prediction accuracy is greatly reduced only by relying on a single index threshold, and meanwhile, the personalized causes cannot be identified without association analysis, only general prevention suggestions can be given, and risks cannot be avoided from the source. 4.