CN-122024987-A - Case analysis and prediction method for ear symptoms secondary to maxillofacial joint disorder

CN122024987ACN 122024987 ACN122024987 ACN 122024987ACN-122024987-A

Abstract

The invention relates to the technical field of medical health information, in particular to a case analysis and prediction method for ear symptoms secondary to maxillofacial joint disorder, which comprises the steps of obtaining original multi-modal data of a plurality of sample objects at different follow-up points; the method comprises the steps of obtaining an individual characteristic value set under each follow-up point according to each sample object, carrying out multi-mode fusion processing according to the individual characteristic value set under each follow-up point, determining the comprehensive state index of the sample object under each follow-up point, constructing an individual dynamic track fitting curve according to the comprehensive state index of the sample object under all the follow-up points, extracting a plurality of change characteristic values from the individual dynamic track fitting curve, and obtaining the change characteristic values of all the sample objects to be used as training input characteristic data of an analysis model. According to the invention, the individual characteristic values with different characteristic dimensions are subjected to multi-mode fusion analysis, so that the data quality and the representativeness of training data are effectively improved.

Inventors

Ren Jiaojie
LI JIACUN
LUO TIAN

Assignees

陕西省中医药研究院(陕西省中医医院陕西省中西医结合研究所)
中国人民解放军空军军医大学

Dates

Publication Date: 20260512
Application Date: 20260414

Claims (10)

1. A method for case analysis and prediction of ear symptoms secondary to maxillofacial joint disorders, comprising the steps of: acquiring original multi-mode data of a plurality of sample objects at different follow-up points, wherein the original multi-mode data at least comprises imaging data, functional biomechanical data and clinical subjective report data; for each sample object, carrying out data preprocessing and physiological characteristic extraction operation on original multi-mode data of the sample object under different follow-up points to obtain an individual characteristic value set under each follow-up point; Carrying out multi-mode fusion processing according to the individual characteristic value set under each follow-up point, and determining the comprehensive state index of the sample object under each follow-up point; constructing an individual dynamic track fitting curve according to the comprehensive state indexes of the sample object at all the follow-up points, and extracting a plurality of change characteristic values from the individual dynamic track fitting curve; and acquiring the change characteristic values of all sample objects as training input characteristic data of an analysis model, wherein the analysis model is used for evaluating the change trend of the state of the oral and jaw system.
2. The method for analyzing and predicting the cases of ear symptoms secondary to maxillofacial joint disorder according to claim 1, wherein the performing data preprocessing on the original multi-modal data of the sample object at different follow-up points at least comprises performing filtering and noise reduction, spatial registration and time sequence marking on the original multi-modal data to obtain preprocessed multi-modal data of the sample object at different follow-up points.
3. The method for case analysis and prediction of ear symptoms secondary to maxillofacial joint disorder according to claim 2, wherein the performing physiological feature extraction operation on the preprocessed multi-modal data of the sample object at different follow-up points to obtain an individual feature value set at each follow-up point comprises: determining a plurality of anatomical feature values according to the imaging data, wherein the anatomical feature values at least comprise a joint disc-condyle relation angle, a condyle bone density and a joint gap size; Determining a plurality of electromyographic signal characteristic values and functional biomechanical characteristic values according to the functional biomechanical data, wherein the electromyographic signal characteristic values at least comprise a resting-stage root mean square value, an active-stage integral electromyographic value, a median frequency and a slope, and the functional biomechanical characteristic values at least comprise a maximum opening degree, an opening and closing movement speed and a track deviation degree; determining a plurality of subjective clinical characteristic values according to clinical subjective report data, wherein the subjective clinical characteristic values at least comprise a maxillofacial joint disorder index and a subjective scoring value; And carrying out standardized treatment on the anatomical structure characteristic value, the electromyographic signal characteristic value, the functional biomechanical characteristic value and the subjective clinical characteristic value of the sample object at different follow-up points to obtain individual characteristic values.
4. The method for analyzing and predicting ear symptoms secondary to maxillofacial joint disorders according to claim 3, wherein said performing a multi-modal fusion process according to the set of individual eigenvalues at each follow-up point, determining the overall state index of the sample subject at each follow-up point comprises: Acquiring individual characteristic value combinations corresponding to different follow-up points of each sample object under the same characteristic dimension, and determining an ear symptom expression index of each characteristic dimension according to the individual characteristic value combinations of each sample object under each characteristic dimension; according to the ear symptom expression index and the associated weight of each feature dimension, weighting all individual feature values of the sample object under each follow-up point, and determining the comprehensive state index of the sample object under each follow-up point; The associated weights are weights of added value 1 set for imaging data, functional biomechanical data and clinical subjective report data according to historical experience.
5. The method for case analysis and prediction of ear symptoms secondary to maxillofacial joint disorders according to claim 4, wherein determining the ear symptom performance index for each feature dimension from individual feature value combinations for each sample object in each feature dimension comprises: According to the individual characteristic value combination of each sample object under each characteristic dimension, individual characteristic values of each characteristic dimension of a plurality of specific sample objects under the first characteristic abnormal node and the symptom occurrence node are selected; Determining a first symptom expression factor of a target feature dimension according to individual feature values and subjective scoring values of the target feature dimension of each specific sample object under the first feature abnormal node and the symptom occurrence node; Counting the number of specific sample objects with abnormal target feature dimensions and abnormal target feature dimensions, and determining a second symptom expression factor of the target feature dimensions according to the ratio of the number of the two sample objects; determining a third symptom expression factor of the target feature dimension according to the occurrence time point of the target feature dimension of each specific sample object under the first feature abnormal node and the symptom occurrence node; determining a performance weight for each symptom performance factor based on the coefficient of variation of the first, second, and third symptom performance factors for each feature dimension; And carrying out weighted summation on the first, second and third symptom expression factors of the target feature dimension by using the expression weight, and determining an ear symptom expression index of the target feature dimension.
6. The method for case analysis and prediction of ear symptoms secondary to maxillofacial joint disorders according to claim 5, wherein determining the first symptom-representing factor of the target feature dimension according to the individual feature value, the subjective score value of the target feature dimension of each specific sample object at the first feature abnormality node and the symptom occurrence node comprises: Calculating the difference between individual characteristic values of the target characteristic dimension of each specific sample object under the first characteristic abnormal node and the symptom occurrence node to form a characteristic change amplitude list under the target characteristic dimension; Calculating the difference between subjective scoring values of each specific sample object under the first characteristic abnormal node and the symptom occurrence node to form a synchronous ear symptom severity change amplitude list; And determining a first symptom expression factor of the target feature dimension according to the correlation between the two lists.
7. The method for case analysis and prediction of ear symptoms secondary to maxillofacial joint disorders according to claim 5, wherein determining a third symptom expression factor of the target feature dimension according to the occurrence time point of the target feature dimension of each specific sample object under the first feature abnormality node and the symptom occurrence node comprises: Calculating the time interval between the occurrence time points of the target feature dimension of each specific sample object under the first feature abnormal node and the symptom occurrence node, and calculating the average value of the time intervals of all the specific sample objects; And carrying out negative correlation normalization processing on the average value of all the time intervals, and taking the obtained normalized value as a third symptom expression factor of the target characteristic dimension.
8. The method for analyzing and predicting ear symptoms secondary to maxillofacial joint disorders according to claim 4, wherein the weighting all individual feature values of the sample object at each follow-up point according to the ear symptom performance index and the associated weight of each feature dimension to determine the comprehensive state index of the sample object at each follow-up point comprises: the individual characteristic values of each characteristic dimension under the corresponding characteristic types are weighted and averaged by utilizing the ear symptom expression index of each characteristic dimension under the same characteristic type, and the comprehensive state factor of each characteristic type of the sample object under the target follow-up point is determined; And weighting and summing the comprehensive state factors of each feature type by utilizing the association weight of each feature type to determine the comprehensive state index of the sample object under the target follow-up point, wherein the target follow-up point is any follow-up point.
9. The method for case analysis and prediction of ear symptoms secondary to maxillofacial joint disorders according to claim 1, wherein after performing multi-modal fusion processing according to the set of individual eigenvalues at each follow-up point, determining the comprehensive state index of the sample object at each follow-up point, further comprises: Acquiring a comprehensive state index of each sample object before a final follow-up point to form a comprehensive state index time sequence, wherein the final follow-up point is at least used for representing a follow-up point corresponding to a final event; Calculating a difference index between every two sample objects according to the comprehensive state index time sequence of each sample object; Clustering all sample objects by utilizing each difference index to obtain track form class labels of each cluster, and taking all track form class labels as training target labels of the analysis model so as to carry out supervised learning training on the analysis model.
10. The method for analyzing and predicting cases of ear symptoms secondary to maxillofacial joint disorders according to claim 1, wherein the variation eigenvalues comprise at least a slope, number of extreme points, number of inflection points, variance, coefficient of variation, and curvature.

Description

Case analysis and prediction method for ear symptoms secondary to maxillofacial joint disorder Technical Field The invention relates to the technical field of medical health information, in particular to a case analysis and prediction method for ear symptoms secondary to maxillofacial joint disorder. Background In the field of oromandibular functional analysis, subjects often exhibit a variety of complex symptoms including abnormal ear experiences, which are closely related to the functional state of the oromandibular system, including dynamic physiological signals such as muscle activity, joint movement, etc. Symptom analysis is currently based on physiological signals, often relying on static, isolated data collected at a single point in time. Because the dynamic evolution characteristics of the functions of the oral and jaw system cannot be effectively captured and quantized, the constructed analysis model is poor in quality and insufficient in representativeness of training data, and further the reliability and generalization capability of the analysis result of the model are low, so that stable and accurate references are difficult to provide in practical application. Disclosure of Invention In order to solve the above technical problems of poor quality and insufficient representativeness of the training data relied on for the analysis model for predicting the ear symptoms secondary to the maxillofacial joint disorder, the present invention aims to provide a case analysis and prediction method for predicting the ear symptoms secondary to the maxillofacial joint disorder, which adopts the following specific technical scheme: one embodiment of the present invention provides a method for case analysis and prediction of ear symptoms secondary to maxillofacial joint disorders, the method comprising the steps of: acquiring original multi-mode data of a plurality of sample objects at different follow-up points, wherein the original multi-mode data at least comprises imaging data, functional biomechanical data and clinical subjective report data; for each sample object, carrying out data preprocessing and physiological characteristic extraction operation on original multi-mode data of the sample object under different follow-up points to obtain an individual characteristic value set under each follow-up point; Carrying out multi-mode fusion processing according to the individual characteristic value set under each follow-up point, and determining the comprehensive state index of the sample object under each follow-up point; constructing an individual dynamic track fitting curve according to the comprehensive state indexes of the sample object at all the follow-up points, and extracting a plurality of change characteristic values from the individual dynamic track fitting curve; and acquiring the change characteristic values of all sample objects as training input characteristic data of an analysis model, wherein the analysis model is used for evaluating the change trend of the state of the oral and jaw system. Further, the data preprocessing of the original multi-mode data of the sample object under different follow-up points at least comprises the steps of filtering and denoising, spatial registration and time sequence marking of the original multi-mode data to obtain the preprocessed multi-mode data of the sample object under different follow-up points. Further, performing physiological feature extraction operation on the preprocessed multi-mode data of the sample object under different follow-up points to obtain an individual feature value set under each follow-up point, including: determining a plurality of anatomical feature values according to the imaging data, wherein the anatomical feature values at least comprise a joint disc-condyle relation angle, a condyle bone density and a joint gap size; Determining a plurality of electromyographic signal characteristic values and functional biomechanical characteristic values according to the functional biomechanical data, wherein the electromyographic signal characteristic values at least comprise a resting-stage root mean square value, an active-stage integral electromyographic value, a median frequency and a slope, and the functional biomechanical characteristic values at least comprise a maximum opening degree, an opening and closing movement speed and a track deviation degree; determining a plurality of subjective clinical characteristic values according to clinical subjective report data, wherein the subjective clinical characteristic values at least comprise a maxillofacial joint disorder index and a subjective scoring value; And carrying out standardized treatment on the anatomical structure characteristic value, the electromyographic signal characteristic value, the functional biomechanical characteristic value and the subjective clinical characteristic value of the sample object at different follow-up points to obtain individual characteristic values.