A pain visualization identification method and system based on multidimensional data analysis

Abstract

The present invention proposes a pain visualization identification method and system based on multidimensional data analysis. The system uses EEG data, musculoskeletal ultrasound images and pain levels to construct a multidimensional pain visualization identification model. The constructed multidimensional pain visualization identification model generates the user's current pain level based on the real-time EEG data collected by the server and the musculoskeletal ultrasound images of a specific time period, and visualizes the user's pain area based on the user's current pain level. The present invention can meet the doctor's real-time monitoring of the patient's pain area, realizes an automated and intelligent pain identification effect, and can ensure that the doctor treats the patient in a targeted manner.

Description

Technical Field

[0001] The present invention belongs to the field of neural engineering technology, and in particular relates to a pain visualization recognition method and system based on multidimensional data analysis. Background Art

[0002] Pain is an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or an experience similar to this. This experience is subjective and is affected by many factors such as biology, psychology, and social environment. According to the definition of the International Association for the Study of Pain (IASP), pain is not just a feeling, but also includes painful experiences in emotional, cognitive, and social dimensions. Visual recognition technology for pain has always been a research hotspot in the field of neuroengineering medicine.

[0003] Multidimensional data analysis is a method of processing and analyzing data sets from multiple dimensions or attributes. In the medical field, pain management often involves multidimensional data such as patients' physiological indicators (such as heart rate, blood pressure), pain assessment scale scores, behavioral patterns (such as sleeping habits), and psychological factors (such as anxiety and depression). Multidimensional data analysis technology can integrate these complex data sets and reveal the inherent connections and potential laws between data through statistical methods, data mining and machine learning algorithms.

[0004] Visualization technology is a technology that converts complex data sets into intuitive forms such as graphics, images or animations so that people can understand and analyze them more easily. In the field of pain management, visualization technology can help medical staff and patients more intuitively understand the dynamic changes of pain, the relationship between pain and physiological indicators, and the effects of different treatment plans. Multidimensional data visualization technology is particularly suitable for displaying complex relationships between multiple variables, such as force-directed graphs, heat maps, scatter plots, etc.

[0005] The field of pain management faces many challenges, such as the subjectivity of pain assessment, the personalized needs of pain treatment, and the complexity and diversity of pain data. Methods for pain identification using musculoskeletal ultrasound have not been well utilized.

[0006] Based on the above problems, the present invention provides more comprehensive, accurate and effective pain management services for medical staff and patients by integrating multidimensional data, using visualization technology to display the dynamic changes of pain, and providing personalized treatment plan suggestions. The present invention specifically uses EEG data, musculoskeletal ultrasound images and pain levels to construct a multidimensional pain visualization recognition model. The constructed multidimensional pain visualization recognition model generates the user's current pain level based on the real-time EEG data collected by the server and the musculoskeletal ultrasound images of a specific time period, and visualizes the user's pain area based on the user's current pain level. The present invention can meet the doctor's real-time monitoring of the patient's pain area, realizes automated and intelligent pain recognition effects, and can ensure that the doctor treats the patient in a targeted manner. Summary of the invention

[0007] In order to solve the above technical problems, the present invention proposes a pain visualization identification method and system based on multidimensional data analysis.

[0008] In a first aspect of the present invention, a method for visually identifying pain based on multidimensional data analysis is provided, wherein the method comprises:

[0009] Obtain patients’ EEG data, musculoskeletal ultrasound images, and regional pain ratings;

[0010] Generate pain visualization vector features using the EEG data and the musculoskeletal ultrasound image;

[0011] Constructing a multi-dimensional pain visualization recognition model according to the pain visualization vector features and the regional pain level;

[0012] Receive the patient's current EEG data and previous examination musculoskeletal ultrasound images, generate current pain visualization vector features based on the current EEG data and the previous examination musculoskeletal ultrasound images, use the multi-dimensional pain visualization recognition model to generate the current regional pain level based on the current pain visualization vector features, and visualize the patient's pain area based on the current regional pain level.

[0013] Furthermore, the EEG data is obtained by processing historical resting-state EEG data and historical pain EEG data, and the musculoskeletal ultrasound image includes a historical musculoskeletal ultrasound image taken during the patient examination;

[0014] Furthermore, the EEG data is collected using a wireless EEG acquisition device, including a NeuSen W series wireless EEG acquisition system or a NeuroSc i wireless EEG acquisition device, which can collect brain signals from 64 pathways.

[0015] Furthermore, specific brain acquisition areas of the EEG signals of the EEG data are O1, O2, Oz, POZ, PO3, PO4, CZ, CPZ, P3 and P4.

[0016] Furthermore, before using the multi-dimensional pain visualization recognition model, the musculoskeletal ultrasound image is processed to separate the target pain site from other ultrasound image backgrounds, and the interpolar second-order moment S between the target pain site and other ultrasound image backgrounds is defined.

[0017] Furthermore, the EEG data and the feature vectors of the muscle-bone separation ultrasound image are concatenated to obtain the pain visualization vector feature.

[0018] Furthermore, the multi-dimensional pain visualization recognition model adopts an improved algorithm classifier based on the Fischer criterion.

[0019] A pain visualization recognition system based on multidimensional data analysis is also provided, the system comprising a patient multidimensional data acquisition module, a multidimensional physiological data storage module, a multidimensional pain visualization recognition model construction module, and a pain visualization recognition module, characterized in that:

[0020] The patient multi-dimensional data acquisition module is used to collect the patient's EEG data, musculoskeletal ultrasound images and regional pain levels; it is also used to collect the patient's current EEG data and pre-examination musculoskeletal ultrasound images;

[0021] The EEG data is obtained by processing historical resting EEG data and historical pain EEG data, and the musculoskeletal ultrasound image includes historical musculoskeletal ultrasound images taken during patient examinations;

[0022] The multidimensional physiological data storage module is used to store the patient's EEG data, musculoskeletal ultrasound images and regional pain levels, generate pain visualization vector features using the EEG data and the musculoskeletal ultrasound images, and generate current pain visualization vector features using the current EEG data and the pre-examination musculoskeletal ultrasound images;

[0023] The multi-dimensional pain visualization recognition model construction module is used to accept the pain visualization vector feature and the regional pain level to construct the multi-dimensional pain visualization recognition model;

[0024] The pain visualization recognition module is used to receive the current pain visualization vector features in the multidimensional physiological data storage module, and is connected to the multidimensional pain visualization recognition model construction module, input the current pain visualization vector features into the multidimensional pain visualization recognition model construction module, use the multidimensional pain visualization recognition model to generate the current area pain level, and visualize the pain area according to the current area pain level, and the visualization adopts WebGIS.

[0025] Furthermore, specific brain acquisition areas of the EEG signals of the EEG data are O1, O2, Oz, POZ, PO3, PO4, CZ, CPZ, P3 and P4.

[0026] Based on the above problems, the present invention provides more comprehensive, accurate and effective pain management services for medical staff and patients by integrating multidimensional data, using visualization technology to display the dynamic changes of pain, and providing personalized treatment plan suggestions. The present invention specifically uses EEG data, musculoskeletal ultrasound images and pain levels to construct a multidimensional pain visualization recognition model. The constructed multidimensional pain visualization recognition model generates the user's current pain level based on the real-time EEG data collected by the server and the musculoskeletal ultrasound images of a specific time period, and visualizes the user's pain area based on the user's current pain level. The present invention can meet the doctor's real-time monitoring of the patient's pain area, realize the automated and intelligent pain recognition effect, and at the same time, visualize the pain area to ensure that the doctor treats the patient in a targeted manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a flow chart of a pain visualization identification method based on multidimensional data analysis of the present invention;

[0028] Figure 2 It is a schematic diagram of the structure of a pain visualization recognition system based on multidimensional data analysis of the present invention;

[0029] Figure 3 It is a classification principle diagram of the classifier based on Fisher criterion in the present invention;

[0030] Figure 4 This is a muscle-bone ultrasound image of the painful area in an embodiment of the present invention;

[0031] Figure 5 This is a visualization image of thigh pain in an embodiment of the present invention. DETAILED DESCRIPTION

[0032] The invention is further described below in conjunction with the accompanying drawings and specific implementation methods.

[0033] In order to solve the above technical problems, the present invention proposes a pain visualization identification method and system based on multidimensional data analysis.

[0034] In a first aspect of the present invention, a method for visually identifying pain based on multidimensional data analysis is provided, wherein the method comprises:

[0035] Obtain patients’ EEG data, musculoskeletal ultrasound images, and regional pain ratings;

[0036] Generate pain visualization vector features using the EEG data and the musculoskeletal ultrasound image;

[0037] Constructing a multi-dimensional pain visualization recognition model according to the pain visualization vector features and the regional pain level;

[0038] In this embodiment, the regional pain levels can be divided into some human body areas such as thighs, calves, waist, back, neck, etc., and are set by technical personnel in this field according to custom.

[0039] Receive the patient's current EEG data and previous examination musculoskeletal ultrasound images, generate current pain visualization vector features based on the current EEG data and the previous examination musculoskeletal ultrasound images, use the multi-dimensional pain visualization recognition model to generate the current regional pain level based on the current pain visualization vector features, and visualize the patient's pain area based on the current regional pain level.

[0040] The pre-examination musculoskeletal ultrasound image in this example is a musculoskeletal ultrasound image taken before a patient undergoes a certain operation or a physical examination.

[0041] Furthermore, the EEG data is obtained by processing historical resting-state EEG data and historical pain EEG data, and the musculoskeletal ultrasound image includes a historical musculoskeletal ultrasound image taken during the patient examination;

[0042] In this embodiment, the EEG data is obtained by standardizing the historical pain EEG data and the historical resting EEG data based on the historical resting EEG data. The standardization process is:

[0043]

[0044] In the formula, A f is the EEG data, A p is the historical pain EEG data, A b is the historical resting-state EEG data, g is the standardization coefficient, which is usually set to 0.1-0.15.

[0045] Furthermore, the EEG data is collected using a wireless EEG acquisition device, including a NeuSen W series wireless EEG acquisition system or a NeuroSc i wireless EEG acquisition device, which can collect brain signals from 64 pathways.

[0046] Further, the specific brain acquisition areas of the EEG signals of the EEG data are O1, O2, Oz, POZ, PO3, PO4, CZ, CPZ, P3 and P4;

[0047] Furthermore, the calculation formula of the EEG data is as follows:

[0048]

[0049] In the formula, A f is the EEG data, k is the difference coefficient, N is the number of EEG channels that can be collected by the wireless device, M is the total number of channels, and B 1 , B 2 , ..., B 10 It is the amplitude of EEG signals measured corresponding to the 10 EEG pathways, namely O1, O2, Oz, POZ, PO3, PO4, CZ, CPZ, P3 and P4, in μV.

[0050] In this example, the optimal range of the difference coefficient is 0.1-0.3, and the EEG pathways can also select FZ, F3 and F4. These EEG pathways are related to the human body's pain perception and belong to the pain perception brain area.

[0051] Furthermore, before using the multi-dimensional pain visualization recognition model, the musculoskeletal ultrasound image is processed to separate the target pain site from other ultrasound image backgrounds. The interpolar second-order moment S between the target pain site and other ultrasound image backgrounds is defined as:

[0052] S=U 0 (G 0 -G) 2 +U 1 (G 1 -G) 2

[0053] Where U 0 is the ratio of the characteristic pixel value belonging to the target pain site to the musculoskeletal ultrasound image, G is the total average grayscale of the musculoskeletal ultrasound image, and U 1 is the ratio of the characteristic pixel values ​​belonging to other musculoskeletal ultrasound image background to the musculoskeletal ultrasound image, G 0 is the average grayscale of the target pain area, G 1is the average grayscale of other musculoskeletal ultrasound image backgrounds, and the interpolar second-order moment S between the target pain site and other ultrasound image backgrounds is greater than a certain threshold as the musculoskeletal ultrasound image.

[0054] Furthermore, the EEG data and the feature vectors of the muscle-bone separation ultrasound image are concatenated to obtain the pain visualization vector feature.

[0055] Furthermore, the multi-dimensional pain visualization recognition model adopts an improved algorithm classifier based on the Fisher criterion, and the calculation formula is as follows:

[0056]

[0057] H is the pain visualization vector feature, L(S) is the output regional pain level, W T is the normal vector perpendicular to the hyperplane, N is the number of EEG channels actually collected, M is the number of all channels of the EEG cap, and h m is the average grayscale of the target pain area in the musculoskeletal ultrasound image, h U is the average grayscale of other background parts of the musculoskeletal ultrasound image. In this example, N is 10 and M is 64 or 128.

[0058] A pain visualization recognition system based on multidimensional data analysis is also provided, the system comprising a patient multidimensional data acquisition module, a multidimensional physiological data storage module, a multidimensional pain visualization recognition model construction module, and a pain visualization recognition module, characterized in that:

[0059] The patient multi-dimensional data acquisition module is used to collect the patient's EEG data, musculoskeletal ultrasound images and regional pain levels; it is also used to collect the patient's current EEG data and pre-examination musculoskeletal ultrasound images;

[0060] The EEG data is obtained by processing historical resting EEG data and historical pain EEG data, and the musculoskeletal ultrasound image includes historical musculoskeletal ultrasound images taken during patient examinations;

[0061] The multidimensional physiological data storage module is used to store the patient's EEG data, musculoskeletal ultrasound images and regional pain levels, generate pain visualization vector features using the EEG data and the musculoskeletal ultrasound images, and generate current pain visualization vector features using the current EEG data and the pre-examination musculoskeletal ultrasound images;

[0062] The multi-dimensional pain visualization recognition model construction module is used to accept the pain visualization vector feature and the regional pain level to construct the multi-dimensional pain visualization recognition model;

[0063] The pain visualization recognition module is used to receive the current pain visualization vector features in the multidimensional physiological data storage module, and is connected to the multidimensional pain visualization recognition model construction module, input the current pain visualization vector features into the multidimensional pain visualization recognition model construction module, use the multidimensional pain visualization recognition model to generate the current area pain level, and visualize the pain area according to the current area pain level, and the visualization adopts WebGIS.

[0064] See also Figure 5 As shown, the pain site in this example may be the thigh, and the thigh pain identification visualization may be differentiated into three color pain levels of red, yellow and green through the pain visualization identification system based on multidimensional data analysis of the present invention. The pain level may be automatically generated based on the multidimensional pain visualization identification model, and automatically generated in real time along with the musculoskeletal ultrasound images collected during the patient's physical examination and the EEG data collected by the wireless EEG device.

[0065] Further, the specific brain acquisition areas of the EEG signals of the EEG data are O1, O2, Oz, POZ, PO3, PO4, CZ, CPZ, P3 and P4;

[0066] The calculation formula of the EEG data is as follows:

[0067]

[0068] In the formula, A f is the EEG data, k is the difference coefficient, N is the number of EEG channels that can be collected by the wireless device, M is the total number of channels, and B 1 , B 2 , ..., B 10 It is the amplitude of EEG signals measured corresponding to the 10 EEG pathways, namely O1, O2, Oz, POZ, PO3, PO4, CZ, CPZ, P3 and P4, in μV.

[0069] Based on the above problems, the present invention provides more comprehensive, accurate and effective pain management services for medical staff and patients by integrating multidimensional data, using visualization technology to display the dynamic changes of pain, and providing personalized treatment plan suggestions. The present invention specifically uses EEG data, musculoskeletal ultrasound images and pain levels to construct a multidimensional pain visualization recognition model. The constructed multidimensional pain visualization recognition model generates the user's current pain level based on the real-time EEG data collected by the server and the musculoskeletal ultrasound images of a specific time period, and visualizes the user's pain area based on the user's current pain level. The present invention can meet the doctor's real-time monitoring of the patient's pain area, realize the automated and intelligent pain recognition effect, and at the same time, visualize the pain area to ensure that the doctor treats the patient in a targeted manner.

[0070] A combination of multiple embodiments of the present invention can achieve all of the above effects, but it is not required that each embodiment of the present invention achieve all of the above advantages and effects, because each embodiment of the present invention can constitute an independent technical solution and make one or more contributions to the prior art.

[0071] The present invention does not particularly specify the structure of some modules, which shall be subject to the contents recorded in the prior art. The prior art mentioned in the above background technology section and the specific embodiment section of the present invention can be used as part of the present invention to understand the meaning of some technical features or parameters. The scope of protection of the present invention shall be subject to the contents actually recorded in the claims.

Claims

1. A pain visualization identification method based on multidimensional data analysis, characterized in that: The method comprises: Obtain patients’ EEG data, musculoskeletal ultrasound images, and regional pain ratings; Generate pain visualization vector features using the EEG data and the musculoskeletal ultrasound image; Constructing a multi-dimensional pain visualization recognition model according to the pain visualization vector features and the regional pain level; Receive the patient's current EEG data and a pre-examination musculoskeletal ultrasound image, generate a current pain visualization vector feature based on the current EEG data and the pre-examination musculoskeletal ultrasound image, generate a current regional pain level based on the current pain visualization vector feature using the multi-dimensional pain visualization recognition model, and visualize the patient's pain area based on the current regional pain level; The EEG data is obtained by processing historical resting EEG data and historical pain EEG data, and the musculoskeletal ultrasound image includes historical musculoskeletal ultrasound images taken during patient examinations; The EEG data is collected using a wireless EEG acquisition device, including a NeuSen W series wireless EEG acquisition system or a NeuroSci wireless EEG acquisition device, which can collect brain signals from 64 pathways; The specific brain areas for EEG data collection are O1, O2, Oz, POZ, PO3, PO4, CZ, CPZ, P3, and P4; The calculation formula for EEG data is as follows: : In the formula, is the EEG data, k is the difference coefficient, N is the number of EEG channels that can be collected by the wireless device, M is the total number of channels, The unit is the EEG signal amplitude measured corresponding to the 10 EEG pathways O1, O2, Oz, POZ, PO3, PO4, CZ, CPZ, P3 and P4. ; Before using the multi-dimensional pain visualization recognition model, the musculoskeletal ultrasound image is processed to separate the target pain site from other ultrasound image backgrounds. The interpolar second-order moment S between the target pain site and other ultrasound image backgrounds is defined as: In the formula, is the ratio of the characteristic pixel values ​​belonging to the target pain site to the musculoskeletal ultrasound image, G is the total average grayscale of the musculoskeletal ultrasound image, is the ratio of the characteristic pixel values ​​belonging to other musculoskeletal ultrasound image backgrounds to the musculoskeletal ultrasound image, is the average grayscale of the target pain area, is the average grayscale of other musculoskeletal ultrasound image backgrounds, and the interpolar second-order moment S between the target pain site and other ultrasound image backgrounds is greater than a certain threshold as the musculoskeletal ultrasound image; The pain visualization vector feature is obtained by splicing the EEG data and the feature vector of the muscle-bone splitting ultrasound image; The multi-dimensional pain visualization recognition model adopts an improved algorithm classifier based on the Fisher criterion, and the calculation formula is as follows: , H is the pain visualization vector feature, L(S) is the output regional pain level, W T is the normal vector perpendicular to the hyperplane, N is the number of EEG channels actually collected, M is the number of all channels of the EEG cap, is the average grayscale of the target painful area of ​​the musculoskeletal ultrasound image, is the average grayscale of other background parts of the musculoskeletal ultrasound image.

2. A pain visualization recognition system based on multidimensional data analysis, the system comprising a patient multidimensional data acquisition module, a multidimensional physiological data storage module, a multidimensional pain visualization recognition model construction module, and a pain visualization recognition module, characterized in that: The patient multi-dimensional data acquisition module is used to collect the patient's EEG data, musculoskeletal ultrasound images and regional pain levels; it is also used to collect the patient's current EEG data and pre-examination musculoskeletal ultrasound images; The EEG data is obtained by processing historical resting EEG data and historical pain EEG data, and the musculoskeletal ultrasound image includes historical musculoskeletal ultrasound images taken during patient examinations; The EEG data is obtained by processing historical resting EEG data and historical pain EEG data, and the musculoskeletal ultrasound image includes historical musculoskeletal ultrasound images taken during patient examinations; The EEG data is collected using a wireless EEG acquisition device, including a NeuSen W series wireless EEG acquisition system or a NeuroSci wireless EEG acquisition device, which can collect brain signals from 64 pathways; The specific brain acquisition areas of the EEG signals of the EEG data are O1, O2, Oz, POZ, PO3, PO4, CZ, CPZ, P3 and P4; The calculation formula of the EEG data is as follows: : In the formula, is the EEG data, k is the difference coefficient, N is the number of EEG channels that can be collected by the wireless device, M is the total number of channels, The unit is the EEG signal amplitude measured corresponding to the 10 EEG pathways O1, O2, Oz, POZ, PO3, PO4, CZ, CPZ, P3 and P4. ; The multidimensional physiological data storage module is used to store the patient's EEG data, musculoskeletal ultrasound images and regional pain levels, generate pain visualization vector features using the EEG data and the musculoskeletal ultrasound images, and generate current pain visualization vector features using the current EEG data and the pre-examination musculoskeletal ultrasound images; The multi-dimensional pain visualization recognition model construction module is used to accept the pain visualization vector feature and the regional pain level to construct the multi-dimensional pain visualization recognition model; The pain visualization recognition module is used to receive the current pain visualization vector feature in the multidimensional physiological data storage module, and is connected to the multidimensional pain visualization recognition model construction module, input the current pain visualization vector feature into the multidimensional pain visualization recognition model construction module, use the multidimensional pain visualization recognition model to generate the current regional pain level, and visualize the pain area according to the current regional pain level, wherein the visualization adopts WebGIS; Before using the multi-dimensional pain visualization recognition model, the musculoskeletal ultrasound image is processed to separate the target pain site from other ultrasound image backgrounds. The interpolar second-order moment S between the target pain site and other ultrasound image backgrounds is defined as: , In the formula, is the ratio of the characteristic pixel values ​​belonging to the target pain site to the musculoskeletal ultrasound image, G is the total average grayscale of the musculoskeletal ultrasound image, is the ratio of the characteristic pixel values ​​belonging to other musculoskeletal ultrasound image backgrounds to the musculoskeletal ultrasound image, is the average grayscale of the target pain area, is the average grayscale of other musculoskeletal ultrasound image backgrounds, and the interpolar second-order moment S between the target pain site and other ultrasound image backgrounds is greater than a certain threshold as the musculoskeletal ultrasound image; The pain visualization vector feature is obtained by splicing the EEG data and the feature vector of the muscle-bone splitting ultrasound image; The multi-dimensional pain visualization recognition model adopts an improved algorithm classifier based on the Fisher criterion, and the calculation formula is as follows: , H is the pain visualization vector feature, L(S) is the output regional pain level, W T is the normal vector perpendicular to the hyperplane, N is the number of EEG channels actually collected, M is the number of all channels of the EEG cap, is the average grayscale of the target painful area of ​​the musculoskeletal ultrasound image, is the average grayscale of other background parts of the musculoskeletal ultrasound image.

Images