Pain sensing method and system based on multi-dimensional data analysis

Abstract

The invention relates to the technical field of data analysis, and discloses a pain sensing method and system based on multi-dimensional data analysis, and the method comprises the steps: generating a first pain score based on an existing pain scoring method; the method includes introducing the limb features to generate a second pain score, generating a predicted pain score based on the diagnostic features, and performing a pain intervention based on the predicted pain score and the second pain score. The system corresponds to the method. According to the method, limb features are introduced, and the authenticity of pain characterization is enhanced; by adjusting the weight of the multi-modal data, the data dependence on the face and audio features is reduced, and the evaluation efficiency is improved; a prediction model is constructed in combination with diagnosis data, combination of clinical experience and data analysis is realized, and evaluation reliability is improved; and confidence difference value judgment and dynamic correction are carried out, so that the problem of conflict between multi-modal data and diagnosis results is solved, dynamic adaptive adjustment of pain scores is realized, and the pertinence and effectiveness of a pain intervention scheme are effectively improved.

Description

Technical Field

[0001] The present application relates to the field of data analysis technology, and specifically to a pain perception method and system based on multidimensional data analysis. Background Art

[0002] Currently, pain assessments often rely on single-modality data (such as facial expressions or audio), which struggles to fully capture individual pain differences and fails to integrate clinical diagnostic information, resulting in inaccurate assessments. For example, traditional methods score solely based on facial expressions or vocal features, ignoring the importance of body movements (such as crouching and grasping) in pain representation. They also fail to leverage diagnostic data in electronic medical records (such as medical history and laboratory parameters) to improve assessment reliability.

[0003] In addition, existing technologies lack dynamic correction. When multimodal data conflict with diagnostic results, the scores cannot be adaptively adjusted, which can easily lead to deviations in intervention plans.

[0004] A Chinese invention patent with authorization publication number CN 117038055B discloses a pain assessment method, system, device, and medium based on a multi-expert model. This invention ignores the importance of limb movements in pain representation and fails to utilize diagnostic data in electronic medical records to improve assessment reliability.

[0005] Therefore, how to integrate multimodal features and diagnostic data of face, audio, and limbs to build a pain assessment system with dynamic correction capabilities is a technical problem that needs to be solved urgently.

[0006] In summary, there is an urgent need for a new technical solution for pain perception based on multidimensional data analysis. Summary of the Invention

[0007] The purpose of this application is to provide a pain perception method and system based on multidimensional data analysis to solve the technical problems raised in the above background technology.

[0008] To achieve the above objectives, this application discloses the following technical solutions: In a first aspect, the present application discloses a pain perception method based on multidimensional data analysis. The method performs a first pain perception based on the patient's facial features and audio features to generate a first pain score. After generating the first pain score, the method includes the following steps: S1: performing a second pain perception based on the patient's facial features, audio features, and limb features to generate a second pain score; wherein the limb features are used to characterize the patient's limb movements caused by pain; S2: Predicting pain based on the patient's diagnostic features to generate a predicted pain score; wherein the diagnostic features are used to characterize the patient's diagnostic condition; S3: Performing pain intervention based on the predicted pain score and the second pain score; wherein the pain intervention is used to relieve the patient's pain.

[0009] Preferably, the limb features are obtained by extracting features based on big data corresponding to the patient's limb movements caused by the features.

[0010] Preferably, when generating the second pain score, the amount of data collected of the facial features and the audio features is reduced, and the weights corresponding to the facial features and the audio features are adjusted.

[0011] Preferably, the diagnostic features are obtained by acquiring diagnostic feature data from an existing electronic medical record system and performing feature extraction, wherein the diagnostic feature data includes medical history records, laboratory test indicators and physician evaluation scale values.

[0012] Preferably, the generation of the predicted pain score comprises: A predictive pain model is constructed based on historical diagnostic feature data. The operation of the predictive pain model includes: Extracting corresponding indicator features based on the laboratory test indicators; extracting corresponding evaluation features based on the values ​​of the doctor evaluation scale; A predicted pain score is generated based on the indicator features and the assessment features.

[0013] Preferably, performing pain intervention based on the predicted pain score and the second pain score includes: generating a confidence difference value based on the predicted pain score and the second pain score, and determining whether the confidence difference value falls within a preset confidence difference value threshold range; wherein the confidence difference value threshold range is used to measure the confidence of the confidence difference value; If the confidence difference falls within the confidence difference threshold range, outputting the second pain score, and performing pain intervention based on the second pain score; If the confidence difference does not fall within the confidence difference threshold range, a predicted limb feature is generated based on the predicted pain score, a correction parameter is generated based on the predicted limb feature and the limb feature, the second pain score is corrected using the correction parameter to generate a third pain score, the third pain score is output, and pain intervention is performed based on the third pain score; wherein the correction parameter is used to characterize the difference between the predicted limb feature and the limb feature, and is also used to generate the third pain score.

[0014] Preferably, the generation of the predicted limb features comprises: A limb mapping relationship between the predicted pain score and the limb feature is constructed, and the predicted pain score is input to generate the corresponding predicted limb feature.

[0015] Preferably, the generation of the correction parameters includes: Calculating the similarity between the predicted limb feature and the limb feature, and defining the similarity as a feature difference value; Based on a preset parameter mapping relationship, the characteristic difference value is mapped to a correction parameter.

[0016] Preferably, the third pain score is generated based on the correction parameter and the second pain score.

[0017] In a second aspect, the present application discloses a pain perception system based on multidimensional data analysis, which is applicable to the pain perception method based on multidimensional data analysis as described above, and includes a second scoring module, a prediction scoring module, and a pain intervention module; The second scoring module is configured to: perform a second pain perception based on the patient's facial features, audio features, and limb features to generate a second pain score; wherein the limb features are used to characterize the patient's limb movements caused by pain; The prediction scoring module is configured to: perform pain prediction based on the patient's diagnostic characteristics to generate a predicted pain score; wherein the diagnostic characteristics are used to characterize the patient's diagnostic condition; The pain intervention module is configured to perform pain intervention based on the predicted pain score and the second pain score; wherein the pain intervention is used to relieve the patient's pain.

[0018] Beneficial effects: The pain perception method and system based on multidimensional data analysis of the present application significantly improve the comprehensiveness and accuracy of pain assessment by integrating multimodal features and diagnostic data of face, audio, and limbs; introduce limb features and use data such as the average amplitude of important joint movements of the human body to enhance the authenticity of pain representation, which is especially suitable for scenarios where children and others may conceal pain; by adjusting the weights of multimodal data, the data dependence on facial and audio features is reduced, thereby improving assessment efficiency; a prediction model is constructed in combination with diagnostic data to achieve the combination of clinical experience and data analysis, thereby improving assessment reliability; confidence difference judgment and dynamic correction generate correction parameters based on the difference between predicted limb features and actual features, resolve the conflict problem between multimodal data and diagnostic results, and achieve dynamic adaptive adjustment of pain scores, providing accurate basis for personalized pain intervention, and effectively improving the pertinence and effectiveness of pain intervention plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 A flowchart of a pain perception method based on multidimensional data analysis provided in an embodiment of the present application; Figure 2 This is a structural block diagram of the pain perception system based on multidimensional data analysis provided in an embodiment of the present application. DETAILED DESCRIPTION

[0021] The following is a clear and complete description of the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0022] In this document, the term "comprising" is intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising..." does not preclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0023] In clinical practice, many medical staff are unable to accurately assess the pain perception of children at different ages, so they are unable to make appropriate pain management or intervention measures for the pain. Furthermore, considering the psychological activities and expressions of children, there may be situations where the pain is not very severe but they act very painful in order to attract attention. Compared with facial features and audio features, limb features are more authentic, so limb features are introduced. In response to the above problems: The first aspect of this embodiment discloses Figure 1 A pain perception method based on multidimensional data analysis is shown. This method performs a first pain perception based on the patient's facial features and audio features to generate a first pain score. It should be noted that the generation of the first pain score in this embodiment can be any one of the existing methods for pain scoring based on facial features and audio features.

[0024] The pain perception method based on multidimensional data analysis comprises the following steps after generating the first pain score: S1: Performing a second pain perception based on the patient's facial features, audio features, and limb features to generate a second pain score; wherein the limb features are used to characterize the patient's limb movements caused by pain; S2: Predicting pain based on the patient's diagnostic features to generate a predicted pain score; wherein the diagnostic features are used to characterize the patient's diagnostic condition; S3: Perform pain intervention based on the predicted pain score and the second pain score; wherein the pain intervention is used to relieve the patient's pain.

[0025] In this embodiment, the second pain perception can be scored using the same pain scoring method as the first pain perception; however, due to the introduction of limb characteristics, the accuracy and authenticity of the second pain score are higher than the first pain score.

[0026] With the development of existing big data technology and the application of image recognition technology, it is feasible to capture and analyze the patient's limb movements based on a camera. In this embodiment: Specifically, the limb features are obtained by extracting the big data corresponding to the patient's limb movements caused by the features.

[0027] As a preferred implementation of this embodiment, the limb feature obtained by feature extraction is the average of the motion amplitudes of important joints on the human body during the monitoring period. The motion data of the patient's knee joint during a 5-minute monitoring period is collected through inertial measurement, and the knee joint flexion and extension angles are extracted using the bone key point detection algorithm. The arithmetic mean of the angle change values ​​of adjacent frames is calculated as the average motion amplitude, and the average motion amplitude is the limb feature.

[0028] Through the above, based on the precise extraction of limb features, a data basis is provided for the generation of the second pain score and pain intervention.

[0029] In existing technologies, pain perception based on facial and audio features requires a huge amount of collected data. To address this phenomenon: Specifically, when generating the second pain score, the amount of data of the collected facial features and audio features is reduced, and the weights corresponding to the facial features and audio features are adjusted.

[0030] It should be noted that in the prior art, facial and audio features are often combined for pain perception. Therefore, in this embodiment, a strategy combining facial, audio, and limb features for pain perception is adopted. In this strategy, the introduction of limb features reduces the reliance of pain perception on facial and audio features, thereby reducing the amount of facial and audio feature data collected.

[0031] As a preferred implementation of this embodiment, when adjusting the weights corresponding to facial features and audio features, the weights corresponding to facial features and audio features are adjusted to 0.3 and 0.3 respectively, and the weight of limb features is configured to 0.4. Based on this, the accuracy and authenticity of the second pain score are improved.

[0032] In the actual application of pain perception, medical staff can roughly estimate the patient's pain score based on the patient's diagnosis results based on experience. However, this experience cannot be applied to complete data analysis. To address this phenomenon: Specifically, the diagnostic features are obtained by obtaining diagnostic feature data from an existing electronic medical record system and performing feature extraction. The diagnostic feature data includes medical history records, laboratory test indicators, and physician evaluation scale values.

[0033] Through the above, the acquisition of diagnostic feature data provides a data basis for the extraction of diagnostic features.

[0034] For the acquired diagnostic feature data, the extraction method adopted in this embodiment is: Specifically, the generation of predicted pain scores includes: A predictive pain model is constructed based on historical diagnostic feature data. The operation of the predictive pain model includes: Extract corresponding indicator features based on laboratory examination indicators; extract corresponding evaluation features based on the values ​​of the doctor evaluation scale; Generate a predicted pain score based on the indicator features and assessment features.

[0035] In this embodiment, an existing deep learning model is used to construct a predictive pain model based on historical diagnostic feature data. After obtaining the indicator features and evaluation features, a weighted summation method is used to obtain a predicted pain score. It should be noted that during the weighted summation, the predicted pain score is normalized using existing normalization techniques to a value that can be directly compared with the second pain score.

[0036] In the practical application of pain perception, the ultimate goal is to provide patients with more precise pain intervention, thereby reducing their pain and improving their satisfaction with the medical treatment process. However, existing pain perception based on facial and audio features ignores diagnostic features. Universal pain perception based on big data corresponding to facial and audio features leads to pain perception for the sake of pain perception, failing to improve the targeted nature of pain perception. To address this issue: Specifically, pain intervention based on the predicted pain score and the second pain score includes: Based on the predicted pain score and second pain score Generate confidence margins , determine whether the confidence difference value belongs to the preset confidence difference value threshold range; wherein the confidence difference value threshold range is used to measure the confidence of the confidence difference value; In this embodiment, the confidence difference is calculated using a confidence difference calculation formula, which is based on the predicted pain score. and second pain score The absolute value of the difference is calculated, and the confidence difference calculation formula is:

[0037] Furthermore, in this embodiment, based on the mean of historical confidence differences, ±2 times the standard deviation of the mean is taken as the confidence difference threshold range.

[0038] If the confidence difference falls within the confidence difference threshold range, outputting a second pain score, and performing pain intervention based on the second pain score; If the confidence difference does not fall within the confidence difference threshold range, a predicted limb feature is generated based on the predicted pain score, a correction parameter is generated based on the predicted limb feature and the limb feature, the second pain score is corrected using the correction parameter to generate a third pain score, the third pain score is output, and pain intervention is performed based on the third pain score; wherein the correction parameter is used to characterize the difference between the predicted limb feature and the limb feature, and is also used to generate the third pain score.

[0039] It should be noted that the pain intervention in this embodiment is a method of pain relief known to those skilled in the art, for example, taking different doses of analgesics based on different pain levels (the pain level can be obtained by zoning the pain score).

[0040] Based on the above, this embodiment optimizes pain intervention based on the analysis of the predicted pain score and the second pain score, thereby improving the pertinence of pain perception.

[0041] In the aforementioned analysis of limb features, it was found that different pain scores produce different limb features. Based on this finding, this embodiment predicts the generation of limb features based on the existing generative adversarial network. Specifically: A limb mapping relationship between predicted pain scores and limb features is constructed, and the predicted pain scores are input to generate the corresponding predicted limb features.

[0042] Through the above, the generation of predicted limb characteristics provides a data basis for the correction of the second pain score, thereby providing a data reference for further improving the accuracy and authenticity of the pain score.

[0043] Specifically, the generation of correction parameters includes: The similarity between the predicted limb feature and the limb feature is calculated, and the similarity is defined as a feature difference value. In this embodiment, the similarity technology can be any existing similarity analysis method.

[0044] Based on the preset parameter mapping relationship, the feature difference value is mapped to the correction parameter.

[0045] As a preferred implementation of this embodiment, the parameter mapping relationship is to use the existing normalization technology to map the feature difference value into a numerical value that can be directly calculated with the second pain score.

[0046] Based on the obtained correction parameters, this embodiment performs the following operations: Specifically, the third pain score The generation of the correction parameters and second pain score conduct.

[0047] In this embodiment, the third pain score is calculated using a third pain score calculation formula, which is:

[0048] In a simple example, the predicted limb features (average knee joint range of motion of 110°) and the actual limb features (90°) reveal that the patient's pain is actually less than predicted. The cosine similarity is calculated to be 0.9, which is defined as the feature difference value. Using existing min-max normalization techniques, 0.9 is mapped to a correction parameter of 0.8. For example, if the second pain score is 8, the third pain score = 8 × 0.8 = 6.4, achieving dynamic correction of the pain score.

[0049] Based on the above, this embodiment generates a more accurate and authentic third pain score based on the correction parameters and the second pain score, providing a data basis for personalized pain intervention.

[0050] The second aspect of this embodiment discloses Figure 2 A pain perception system based on multidimensional data analysis is shown, which is applicable to the pain perception method based on multidimensional data analysis as described above, and includes a second scoring module, a prediction scoring module, and a pain intervention module; The second scoring module is configured to: generate a second pain score based on the patient's facial features, audio features, and limb features for a second pain perception; wherein the limb features are used to characterize the patient's limb movements caused by pain; The prediction scoring module is configured to: predict pain based on the patient's diagnostic characteristics and generate a predicted pain score; wherein the diagnostic characteristics are used to characterize the patient's diagnostic condition; The pain intervention module is configured to: perform pain intervention based on the predicted pain score and the second pain score; wherein the pain intervention is used to relieve the patient's pain.

[0051] It should be noted that the pain perception system based on multidimensional data analysis of this embodiment corresponds to the aforementioned pain perception system based on multidimensional data analysis. Therefore, the contents not specifically described in the pain perception system based on multidimensional data analysis of this embodiment may include but are not limited to functional definitions, working principles and technical effects, etc., and may all refer to the records in the aforementioned pain perception method based on multidimensional data analysis. This text will not elaborate on them here.

[0052] In summary, the pain perception method and system based on multidimensional data analysis of this embodiment significantly improve the comprehensiveness and accuracy of pain assessment by integrating multimodal features and diagnostic data of face, audio, and limbs; introduce limb features and use data such as the average amplitude of important joint movements of the human body to enhance the authenticity of pain representation, which is especially suitable for scenarios where children and others may conceal pain; by adjusting the weights of multimodal data, the data dependence on facial and audio features is reduced, thereby improving assessment efficiency; a prediction model is constructed in combination with diagnostic data to achieve the combination of clinical experience and data analysis, thereby improving assessment reliability; confidence difference judgment and dynamic correction generate correction parameters based on the difference between predicted limb features and actual features, resolve the conflict problem between multimodal data and diagnostic results, and achieve dynamic adaptive adjustment of pain scores, providing accurate basis for personalized pain intervention, and effectively improving the pertinence and effectiveness of pain intervention plans.

[0053] In the embodiments provided herein, it should be understood that the embodiments described herein can be implemented using hardware, software, firmware, middleware, code, or any appropriate combination thereof. For hardware implementation, the processor may be implemented in one or more of the following: an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field-programmable gate array (FPGA), a processor, a controller, a microcontroller, a microprocessor, or other electronic units designed to implement the functionality described herein, or any combination thereof. For software implementation, some or all of the processes of the embodiments may be performed by a computer program instructing the relevant hardware. During implementation, the program may be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein communication media includes any medium that facilitates the transmission of a computer program from one location to another. The storage medium may be any available medium that can be accessed by a computer. Computer-readable storage media may include, but are not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing the desired program code in the form of instructions or data structures and accessible by a computer.

[0054] Finally, it should be noted that the above is only a preferred embodiment of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent replacements for some of the technical features therein. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. A pain perception method based on multidimensional data analysis, wherein the method performs a first pain perception based on the patient's facial features and audio features and generates a first pain score, characterized in that: After generating the first pain score, the method comprises the following steps: S1: performing a second pain perception based on the patient's facial features, audio features, and limb features to generate a second pain score; wherein the limb features are used to characterize the patient's limb movements caused by pain; S2: Predicting pain based on the patient's diagnostic features to generate a predicted pain score; wherein the diagnostic features are used to characterize the patient's diagnostic condition; S3: Performing pain intervention based on the predicted pain score and the second pain score; wherein the pain intervention is used to relieve the patient's pain.

2. The pain perception method based on multidimensional data analysis according to claim 1, characterized in that: The limb features are obtained by extracting features based on big data corresponding to the patient's limb movements caused by the features.

3. The pain perception method based on multidimensional data analysis according to claim 1, characterized in that: When generating the second pain score, the amount of data collected of the facial features and the audio features is reduced, and the weights corresponding to the facial features and the audio features are adjusted.

4. The pain perception method based on multidimensional data analysis according to claim 1, characterized in that: The diagnostic features are obtained by extracting diagnostic feature data from an existing electronic medical record system, and the diagnostic feature data includes medical history records, laboratory test indicators and physician evaluation scale values.

5. The pain perception method based on multidimensional data analysis according to claim 4, characterized in that: The generation of the predicted pain score comprises: A predictive pain model is constructed based on historical diagnostic feature data. The operation of the predictive pain model includes: Extracting corresponding indicator features based on the laboratory test indicators; extracting corresponding evaluation features based on the values ​​of the doctor evaluation scale; A predicted pain score is generated based on the indicator features and the assessment features.

6. The pain perception method based on multidimensional data analysis according to claim 1, characterized in that: The performing pain intervention based on the predicted pain score and the second pain score includes: generating a confidence difference value based on the predicted pain score and the second pain score, and determining whether the confidence difference value falls within a preset confidence difference value threshold range; wherein the confidence difference value threshold range is used to measure the confidence of the confidence difference value; If the confidence difference falls within the confidence difference threshold range, outputting the second pain score, and performing pain intervention based on the second pain score; If the confidence difference does not fall within the confidence difference threshold range, a predicted limb feature is generated based on the predicted pain score, a correction parameter is generated based on the predicted limb feature and the limb feature, the second pain score is corrected using the correction parameter to generate a third pain score, the third pain score is output, and pain intervention is performed based on the third pain score; wherein the correction parameter is used to characterize the difference between the predicted limb feature and the limb feature, and is also used to generate the third pain score.

7. The pain perception method based on multidimensional data analysis according to claim 6, characterized in that: The generation of the predicted limb features includes: A limb mapping relationship between the predicted pain score and the limb feature is constructed, and the predicted pain score is input to generate the corresponding predicted limb feature.

8. The pain perception method based on multidimensional data analysis according to claim 6, characterized in that: The generation of the correction parameters includes: Calculating the similarity between the predicted limb feature and the limb feature, and defining the similarity as a feature difference value; Based on a preset parameter mapping relationship, the characteristic difference value is mapped to a correction parameter.

9. The pain perception method based on multidimensional data analysis according to claim 6, characterized in that: The third pain score is generated based on the correction parameter and the second pain score.

10. A pain perception system based on multidimensional data analysis, the system being applicable to the pain perception method based on multidimensional data analysis according to any one of claims 1 to 9, characterized in that: The system includes a second scoring module, a prediction scoring module, and a pain intervention module; The second scoring module is configured to: perform a second pain perception based on the patient's facial features, audio features, and limb features to generate a second pain score; wherein the limb features are used to characterize the patient's limb movements caused by pain; The prediction scoring module is configured to: perform pain prediction based on the patient's diagnostic characteristics to generate a predicted pain score; wherein the diagnostic characteristics are used to characterize the patient's diagnostic condition; The pain intervention module is configured to perform pain intervention based on the predicted pain score and the second pain score; wherein the pain intervention is used to relieve the patient's pain.

Images