Traditional Chinese medicine diagnosis and treatment data analysis system and method based on digital human cloning technology

Abstract

The present invention relates to the field of electrical data processing technology, and specifically discloses a TCM diagnosis and treatment data analysis system and method based on digital human cloning technology, the system comprising: a TCM diagnosis and treatment data acquisition module, used to collect characterization data of cloned objects and digital humans; a clone similarity degree analysis module, used to comprehensively analyze and obtain similarity evaluation values ​​of cloned objects of digital humans, and perform similarity evaluation on digital humans according to the similarity evaluation values ​​of cloned objects of digital humans to obtain high-similarity digital humans; a behavior pattern accuracy evaluation module, used to perform accuracy evaluation on high-similarity digital humans and obtain evaluation results. By providing a TCM diagnosis and treatment data analysis system and method based on digital human cloning technology, the present invention can reveal possible errors and deviations of digital humans in the process of behavior simulation, thereby improving the overall reliability of digital human applications, reducing user confusion and dissatisfaction caused by incoherent or erroneous behaviors, and improving user interactive experience.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical data processing, and in particular to a system and method for analyzing TCM diagnosis and treatment data based on digital human cloning technology. Background Art

[0002] At present, digital human cloning technology is a very important part of the field of traditional Chinese medicine treatment technology. With the advancement of technology and the growth of market demand, providing more comprehensive and efficient traditional Chinese medicine diagnosis and treatment has become the norm. The application of high-quality and efficient digital human cloning technology is of great significance to improving the experience of traditional Chinese medicine diagnosis and treatment subjects and enhancing market competitiveness.

[0003] For example, the invention patent with announcement number CN118426593B is a method and system based on virtual digital human interaction, including a text recognition module, a speech recognition module, a decision module, an execution module, a human-computer interaction module, a sentiment analysis module, a multimodal perception unit, a visual recognition module and a visual assistance module; the text recognition module is used for users to communicate with virtual digital humans through text input, and to analyze and understand the user's text input content and text sentiment; the speech recognition module is used for users to communicate with virtual digital humans through voice input, and the system uses speech recognition technology and voiceprint recognition technology to convert the user's voice into computer-readable text.

[0004] For example, the invention patent with announcement number CN117234369B is a digital human interaction method and system, a computer-readable storage medium, and a digital human device. The method includes: receiving a user interaction video containing voice; separating the user interaction video into video frames and voice, converting the voice into text, and obtaining multimodal data; obtaining emotion perception results and perception codes based on video, voice, and text respectively according to the multimodal data; determining a control label vector according to the emotion perception results or perception codes of each modality; fusing the control label vector into the perception codes of each modality; generating generated content based on video, voice, and text respectively according to the fused perception codes of each modality; and synthesizing the generated content based on video, voice, and text to obtain a synthesized video.

[0005] However, in the process of implementing the technical solution of the invention in the embodiments of the present application, the present application discovered that the above technology has at least the following technical problems: the digital human clones currently used for TCM diagnosis and treatment may gradually deviate from the real characteristics of the physical person during long-term use, resulting in a decline in the interactive experience, and there are limitations on improving the interactive experience of TCM diagnosis and treatment subjects. Summary of the invention

[0006] In view of the deficiencies of the prior art, the present invention provides a method and system for analyzing TCM diagnosis and treatment data based on digital human cloning technology, which can effectively solve the problems involved in the above-mentioned background technology.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: In a first aspect, the present invention provides a TCM diagnosis and treatment data analysis system based on digital human cloning technology, comprising: a TCM diagnosis and treatment data acquisition module, used to collect representation data of cloned objects and digital humans, wherein the representation data includes appearance feature data and sound feature data.

[0008] The cloning similarity analysis module is used to obtain the appearance similarity evaluation index of the digital person based on the appearance feature data, process the voice feature data to obtain the voice similarity evaluation index of the digital person, conduct a comprehensive analysis to obtain the similarity evaluation value of the digital person's clone object, and perform similarity evaluation on the digital person based on the similarity evaluation value of the digital person's clone object to obtain a high-similarity digital person.

[0009] The behavior pattern accuracy assessment module is used to obtain the behavior pattern data of high-similarity digital humans, process the behavior pattern data to obtain the behavior pattern accuracy assessment index of high-similarity digital humans, conduct accuracy assessment on the high-similarity digital humans based on the behavior pattern accuracy assessment index of high-similarity digital humans and obtain the assessment results. If the assessment result is qualified, the diagnosis and treatment stage is entered. If the assessment result is unqualified, the behavior pattern of the high-similarity digital humans is fed back and corrected and evaluated again.

[0010] As a further method, the appearance similarity evaluation index of the digital human is obtained by processing the appearance feature data. The specific processing process is: the appearance feature data includes the facial image and facial aspect ratio of the cloned object and the digital human.

[0011] Deploy several key points of facial features, overlap the facial image of the digital human with the facial image of the cloned object, extract the distance between the key points of facial features on the facial image of the cloned object and the digital human, and mark them as the deviation distance of the key points of facial features of the digital human.

[0012] The pixel values ​​of the key positions of the facial features of the cloned object and the digital human are extracted and subtracted, and the absolute value of the difference result is marked as the deviation pixel value of the key positions of the facial features of the digital human.

[0013] The allowable deviation facial aspect ratio, critical facial feature key position point deviation distance and critical facial feature key position point deviation pixel value are extracted from the traditional Chinese medicine diagnosis and treatment database, and a comprehensive analysis is performed to obtain the digital human appearance similarity evaluation index.

[0014] As a further method, the sound feature data is processed to obtain the digital human voice similarity evaluation index. The specific analysis process is: obtaining test audio of the cloned object and the digital human, deploying several time monitoring points, and the sound feature data includes the average speaking speed, the frequency of each harmonic component of the spectrum and the pitch of each time monitoring point.

[0015] The allowable deviation pitch, allowable deviation spectral harmonic component frequency and allowable deviation average speaking speed are extracted from the traditional Chinese medicine diagnosis and treatment database, and the voice similarity evaluation index of the digital human is obtained through comprehensive analysis.

[0016] As a further method, the comprehensive analysis obtains the similarity evaluation value of the digital person's clone object. The specific analysis process is: based on the digital person's appearance similarity evaluation index and voice similarity evaluation index, the comprehensive analysis obtains the similarity evaluation value of the digital person's clone object.

[0017] As a further method, the digital human is evaluated for similarity based on the similarity evaluation value of the digital human's clone object. The specific evaluation process is: extracting the clone object similarity evaluation threshold from the traditional Chinese medicine diagnosis and treatment database, comparing the clone object similarity evaluation value of the digital human with the clone object similarity evaluation threshold, if the clone object similarity evaluation value of the digital human is greater than or equal to the clone object similarity evaluation threshold, then the digital human is evaluated as qualified and marked as a high-similarity digital human, if the clone object similarity evaluation value of the digital human is less than the clone object similarity evaluation threshold, then the digital human is evaluated as unqualified, and feedback correction is performed, and the corrected digital human is evaluated for similarity.

[0018] As a further method, the behavior pattern data is processed to obtain a behavior pattern accuracy evaluation index of a high-similarity digital person. The specific analysis process is: the behavior pattern data includes a consistency evaluation value of diagnosis and treatment recommendations and a matching degree of medication recommendations of a high-similarity digital person.

[0019] The critical consistency evaluation value of diagnosis and treatment recommendations and the critical matching degree of medication recommendations were extracted from the traditional Chinese medicine diagnosis and treatment database, and a comprehensive analysis was performed to obtain the behavioral pattern accuracy evaluation index of a high-similarity digital human.

[0020] As a further method, the behavioral pattern accuracy assessment index of the high-similarity digital human is a quantitative indicator obtained by analyzing the consistency assessment value of diagnosis and treatment recommendations, the matching degree of medication recommendations and the accuracy of disease understanding, and is used to quantify the accuracy of the behavioral patterns of digital humans and cloned objects.

[0021] As a further method, the digital human appearance similarity evaluation index has a specific numerical expression as follows:

[0022]

[0023] Among them, A represents the appearance similarity evaluation index of the digital human, e represents the natural constant, f represents the facial aspect ratio of the digital human, Indicates the facial aspect ratio of the cloned object. represents the allowable deviation of the facial aspect ratio, Indicates the pixel value of the deviation of the key position of the i-th facial features of the digital person, Indicates the deviation pixel value of the critical facial features key position point, represents the deviation distance of the key position point of the i-th facial features of the digital human, Indicates the deviation distance of the set critical facial features key position points, It represents the influencing factor of appearance similarity evaluation corresponding to the set facial aspect ratio, It represents the influencing factor of appearance similarity assessment corresponding to the deviation pixel value of the key position points of the facial features. Indicates the appearance similarity assessment impact factor corresponding to the set deviation distance of the key position points of the facial features, i represents the number of each key position point of the facial features, i=1,2,3,...,m, and m represents the total number of key position points of the facial features. As a further method, the digital human voice similarity assessment index is a quantitative index obtained by analyzing the pitch, the frequency of each harmonic component of the spectrum and the average speaking speed, which is used to quantify the voice similarity between the digital human and the cloned object.

[0024] The second aspect of the present invention provides a method for analyzing TCM diagnosis and treatment data based on digital human cloning technology, comprising: collecting characterization data of the cloned object and the digital human, wherein the characterization data comprises appearance feature data and sound feature data.

[0025] The appearance similarity evaluation index of the digital person is obtained by processing the appearance feature data, the voice similarity evaluation index of the digital person is obtained by processing the voice feature data, and the similarity evaluation value of the digital person's clone object is obtained by comprehensive analysis. The digital person is evaluated for similarity based on the similarity evaluation value of the digital person's clone object to obtain a high-similarity digital person.

[0026] Obtain behavioral pattern data of high-similarity digital humans, process the behavioral pattern data to obtain a behavioral pattern accuracy assessment index of high-similarity digital humans, conduct accuracy assessment on the high-similarity digital humans based on the behavioral pattern accuracy assessment index of high-similarity digital humans and obtain the assessment results. If the assessment result is qualified, enter the diagnosis and treatment stage; if the assessment result is unqualified, provide feedback and correct the behavioral pattern of the high-similarity digital humans and assess them again.

[0027] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0028] (1) The present invention provides a method and system for analyzing TCM diagnosis and treatment data based on digital human cloning technology, which can reveal possible errors and deviations in the behavior simulation process of digital humans and encourage developers to make targeted optimizations and improvements, thereby improving the overall reliability of digital human applications. Accurate behavior patterns make digital humans more natural and smooth when interacting with users, reducing user confusion and dissatisfaction caused by inconsistent or erroneous behavior, and improving user interaction experience.

[0029] (2) By evaluating the voice similarity evaluation index of a digital human, the present invention can improve the naturalness and immersion of users' interaction with the digital human, enhance the users' trust and satisfaction, meet the users' personalized needs, and help the users to more accurately understand the information or service content provided by the digital human. At the same time, it can ensure that the voice of the digital human is highly consistent with that of the reference object, thereby improving recognition accuracy, which helps to reduce recognition errors or confusion caused by voice differences.

[0030] (3) By evaluating the similarity evaluation value of the digital human's clone object, the present invention can more accurately measure the extent to which the digital human can simulate the appearance and voice of a real expert, thereby ensuring its effectiveness and reliability in various application scenarios, helping to discover and improve deficiencies in the technology, and thus promoting the continuous optimization and development of digital human technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The present invention is further described using the accompanying drawings, but the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative work.

[0032] Figure 1 It is a schematic diagram of system module connection of the present invention.

[0033] Figure 2 It is a schematic diagram of the method flow of the present invention.

[0034] Figure 3 It is a schematic diagram of the functional relationship between the similarity evaluation value of the cloned object and the appearance similarity evaluation index of the digital human according to the present invention.

[0035] Figure 4 It is a schematic diagram of the functional relationship between the behavior pattern accuracy evaluation index of the high-similarity digital human of the present invention and the consistency evaluation value of the diagnosis and treatment advice. DETAILED DESCRIPTION

[0036] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0037] Reference Figure 1 As shown, the first aspect of the present invention provides a TCM diagnosis and treatment data analysis system based on digital human cloning technology, including: a TCM diagnosis and treatment data acquisition module, used to collect characterization data of cloned objects and digital humans, and the characterization data includes appearance feature data and sound feature data.

[0038] The cloning similarity analysis module is used to obtain the appearance similarity evaluation index of the digital person based on the appearance feature data, process the voice feature data to obtain the voice similarity evaluation index of the digital person, conduct a comprehensive analysis to obtain the similarity evaluation value of the digital person's clone object, and perform similarity evaluation on the digital person based on the similarity evaluation value of the digital person's clone object to obtain a high-similarity digital person.

[0039] The behavior pattern accuracy assessment module is used to obtain the behavior pattern data of high-similarity digital humans, process the behavior pattern data to obtain the behavior pattern accuracy assessment index of high-similarity digital humans, conduct accuracy assessment on the high-similarity digital humans based on the behavior pattern accuracy assessment index of high-similarity digital humans and obtain the assessment results. If the assessment result is qualified, the diagnosis and treatment stage is entered. If the assessment result is unqualified, the behavior pattern of the high-similarity digital humans is fed back and corrected and evaluated again.

[0040] Specifically, the appearance similarity evaluation index of the digital human is obtained according to the appearance feature data processing, and the specific processing process is: the appearance feature data includes the facial image and facial aspect ratio of the cloned object and the digital human.

[0041] Deploy several key points of facial features, overlap the facial image of the digital human with the facial image of the cloned object, extract the distance between the key points of facial features on the facial image of the cloned object and the digital human, and mark them as the deviation distance of the key points of facial features of the digital human.

[0042] The pixel values ​​of the key positions of the facial features of the cloned object and the digital human are extracted and subtracted, and the absolute value of the difference result is marked as the deviation pixel value of the key positions of the facial features of the digital human.

[0043] The allowable deviation facial aspect ratio, critical facial feature key position point deviation distance and critical facial feature key position point deviation pixel value are extracted from the traditional Chinese medicine diagnosis and treatment database, and a comprehensive analysis is performed to obtain the digital human appearance similarity evaluation index.

[0044] In a specific embodiment, the cloned object refers to the object that needs to be digitally cloned, and the facial aspect ratio generally refers to the ratio of the width to the height of the face, which can be calculated by measuring the width and height of the facial area, where the width is generally the distance between the two outermost key points, such as two ears or two cheeks, and the height is generally the distance between the forehead and the chin. The key position points of the facial features are randomly deployed, and then an affine transformation can be applied to overlap the facial image of the digital person with the facial image of the cloned object to ensure that the size and contour of the image are consistent, and the pixel values ​​of the key position points of the facial features on the facial images of the cloned object and the digital person can be measured by image processing software, where the pixel value refers to the grayscale value.

[0045] Specifically, the digital human appearance similarity evaluation index is expressed as:

[0046] Among them, A represents the appearance similarity evaluation index of the digital human, e represents the natural constant, f represents the facial aspect ratio of the digital human, Indicates the facial aspect ratio of the cloned object. represents the allowable deviation of the facial aspect ratio, Indicates the pixel value of the deviation of the key position of the i-th facial features of the digital person, Indicates the deviation pixel value of the critical facial features key position point, represents the deviation distance of the key position point of the i-th facial features of the digital human, Indicates the deviation distance of the set critical facial features key position points, It represents the influencing factor of appearance similarity evaluation corresponding to the set facial aspect ratio, It represents the influencing factor of appearance similarity assessment corresponding to the deviation pixel value of the key position points of the facial features. It represents the influencing factor of appearance similarity assessment corresponding to the set deviation distance of the key position points of facial features, i represents the number of each key position point of facial features, i=1,2,3,...,m, m represents the total number of key position points of facial features.

[0047] The algorithm of this embodiment combines the facial aspect ratio, the deviation pixel value and the deviation distance of the key position points of the facial features of the digital human and the cloned object, and comprehensively analyzes to obtain the appearance similarity evaluation index. The change of the facial aspect ratio will directly affect the relative position of the facial features in the face, which may further lead to the change of the deviation pixel value and the deviation distance of the key position points of the facial features. For example, when the facial aspect ratio increases, the width of the face increases, which may cause the relative position of the facial features such as the eyes and nose in the horizontal direction to change, thereby generating deviations. At the same time, slight deviations in the positions of the facial features may also affect the overall aspect ratio of the face through cumulative effects. Comprehensive analysis can obtain a more comprehensive, accurate and in-depth appearance similarity evaluation index.

[0048] Table 1. Example of appearance similarity evaluation index data

[0049]

[0050] As shown in Table 1, the appearance similarity evaluation index is jointly determined by the facial aspect ratio, the deviation pixel value of the key position points of the facial features and the deviation distance of the key position points of the facial features. In a specific embodiment, m=1, the reference standard facial aspect ratio is 1.6, the allowable deviation facial aspect ratio is 0.5, the critical facial feature key position point deviation pixel value is 10 pixels, the critical facial feature key position point deviation distance is 10mm, the appearance similarity evaluation influence factor corresponding to the set facial aspect ratio is 0.3, the appearance similarity evaluation influence factor corresponding to the set facial feature key position point deviation pixel value is 0.3, and the appearance similarity evaluation influence factor corresponding to the set facial feature key position point deviation distance is 0.4.

[0051] It should be explained that three key factors are considered in this embodiment, namely, the facial aspect ratio of the digital human and the cloned object, the deviation pixel value of the key position points of the facial features, and the deviation distance of the key position points of the facial features, to ensure that the facial structure of the digital human is consistent with human physiological characteristics, avoid the sense of unreality caused by disproportion, and help to establish the user's trust in the digital human. At the same time, the precise control of the position of the facial features makes the facial expressions of the digital human and the cloned Chinese medicine expert more delicate and natural. Therefore, the accurate position of the facial features is crucial to showing the professional image of the Chinese medicine expert. At the same time, in scenarios such as education or remote diagnosis and treatment, users need to interact with the digital human. Considering these key factors can ensure that the expression changes and eye contact of the digital human during the communication process can be highly realistic, thereby improving the user's interactive experience and satisfaction. By standardizing the facial aspect ratio deviation, the deviation pixel value of the key position points of the facial features, and the deviation distance of the key position points of the facial features of the digital human and the cloned object, it is ensured that they are compared at the same level, which improves the fairness and comparability of the evaluation. and The setting of helps to determine whether there is a significant deviation in the position of the facial features. By weighting the influence of facial aspect ratio deviation, facial feature key position point deviation pixel value and facial feature key position point deviation distance, it reflects their relative importance in the evaluation index. The weights of different factors can be adjusted according to different needs, making the model highly adaptable. It is not difficult to see that the smaller the facial aspect ratio deviation or the smaller the facial feature key position point deviation pixel value or the smaller the facial feature key position point deviation distance, the greater the appearance similarity evaluation index. By evaluating the digital human appearance similarity evaluation index, the evaluation results can be made more accurate and objective, reducing subjective errors caused by personal aesthetic differences, observation angles and other factors. It not only helps to improve the quality and efficiency of digital human creation and enhance user interaction experience, but also helps to promote the continuous development and application expansion of related technologies.

[0052] It should be understood that in this embodiment is the appearance similarity assessment influencing factor corresponding to the facial aspect ratio preset in the TCM diagnosis and treatment database, which represents the numerical value of the degree of influence of the facial aspect ratio on the appearance similarity assessment index. When used, the appearance similarity assessment influencing factor corresponding to the facial aspect ratio can be directly obtained from the TCM diagnosis and treatment database, and the corresponding relationship can be a pre-set mapping relationship. For example, the facial aspect ratio deviation of the digital human and the cloned object and the appearance similarity assessment influencing factor corresponding to the facial aspect ratio preset in the TCM diagnosis and treatment database form a mapping set, and the real-time facial aspect ratio is input into the mapping set to obtain the appearance similarity assessment influencing factor corresponding to the facial aspect ratio. The mapping relationship can be one-to-one or many-to-one. In this example, its value range is between 0 and 1; It is the appearance similarity assessment influencing factor corresponding to the deviation pixel value of the key position points of facial features preset in the traditional Chinese medicine diagnosis and treatment database, which represents the numerical value of the degree of influence of the deviation pixel value of the key position points of facial features on the appearance similarity assessment index. When used, the appearance similarity assessment influencing factor corresponding to the deviation pixel value of the key position points of facial features can be directly obtained from the traditional Chinese medicine diagnosis and treatment database. The corresponding relationship can be a pre-set mapping relationship. For example, the deviation pixel values ​​of the key position points of facial features of the digital human and the cloned object and the appearance similarity assessment influencing factors corresponding to the deviation pixel values ​​of the key position points of facial features preset in the traditional Chinese medicine diagnosis and treatment database form a mapping set. The real-time deviation pixel values ​​of the key position points of facial features are input into the mapping set to obtain the appearance similarity assessment influencing factors corresponding to the deviation pixel values ​​of the key position points of facial features. The mapping relationship can be one-to-one or many-to-one. In this example, its value range is between 0 and 1; It is the appearance similarity assessment influencing factor corresponding to the deviation distance of the key positions of the facial features preset in the traditional Chinese medicine diagnosis and treatment database, which represents the numerical value of the degree of influence of the deviation distance of the key positions of the facial features on the appearance similarity assessment index. When used, the appearance similarity assessment influencing factor corresponding to the deviation distance of the key positions of the facial features can be directly obtained from the traditional Chinese medicine diagnosis and treatment database, and the corresponding relationship can be a pre-set mapping relationship. For example, the deviation distance of the key positions of the facial features of the digital human and the cloned object and the appearance similarity assessment influencing factor corresponding to the deviation distance of the key positions of the facial features preset in the traditional Chinese medicine diagnosis and treatment database form a mapping set, and the real-time deviation distance of the key positions of the facial features is input into the mapping set to obtain the appearance similarity assessment influencing factor corresponding to the deviation distance of the key positions of the facial features. The mapping relationship can be one-to-one or many-to-one. In this example, its value range is between 0 and 1.

[0053] It should be explained that the digital human appearance similarity evaluation index in this embodiment is a quantitative indicator obtained by analyzing the facial aspect ratio, the deviation pixel value of the key position points of the facial features and the deviation distance of the key position points of the facial features of the digital human and the cloned object, and is used to quantify the degree of appearance similarity between the digital human and the cloned object.

[0054] In a specific embodiment, the sound feature data is processed to obtain the digital human voice similarity evaluation index. The specific analysis process is: obtaining the test audio of the cloned object and the digital human, deploying a number of time monitoring points, and the sound feature data includes the average speaking speed, the frequency of each harmonic component of the spectrum and the pitch of each time monitoring point.

[0055] The allowable deviation pitch, allowable deviation spectral harmonic component frequency and allowable deviation average speaking speed are extracted from the traditional Chinese medicine diagnosis and treatment database, and the voice similarity evaluation index of the digital human is obtained through comprehensive analysis.

[0056] It needs to be explained that pitch refers to the value of fundamental frequency, which is the lowest frequency component in sound and determines the basic pitch of sound. Speech rate refers to the speed of speaking and is usually measured in words spoken per minute. The pitch of the cloned object and the digital person can be directly measured by a spectrum analyzer, and the speech rate of the cloned object and the digital person can be measured by a stopwatch or a recording device. The frequency of the harmonic component of the spectrum is an integer multiple of the fundamental frequency. The fundamental and the first few components of the harmonic (such as the second harmonic, the third harmonic, etc.) usually have an important influence on the timbre and sound quality of the audio signal, while higher-order harmonics may gradually weaken and have less impact on the overall timbre. In this embodiment, the first to fifth harmonics are mainly considered. The frequency of the harmonic component of the spectrum in audio is an important aspect of the frequency domain analysis of audio signals. It involves the energy distribution of the sound signal at different frequencies and determines the timbre of the sound. It can be directly measured by a spectrum analyzer.

[0057] Specifically, the digital human voice similarity evaluation index is expressed as:

[0058] in, represents the digital human voice similarity evaluation index, represents the pitch of the digital human at the i-th time node, Indicates the pitch of the cloned object at the i-th time node, Indicates the allowable deviation pitch, represents the average speaking speed of digital people, Indicates the average speaking speed of the cloned object. Indicates the allowable deviation from the average speaking speed, represents the frequency of the gth harmonic component of the spectrum of the digital human, Indicates the frequency of the gth harmonic component of the spectrum of the cloned object. Indicates the permissible deviation frequency of the harmonic components of the spectrum, Indicates the sound similarity evaluation impact factor corresponding to the set pitch deviation, Indicates the sound similarity evaluation impact factor corresponding to the set average speech speed deviation. Indicates the sound similarity assessment impact factor corresponding to the set frequency of the spectrum harmonic component, j represents the number of each time node, j=1,2,3,...,n, n represents the total number of time node numbers, g represents the number of each harmonic component of the spectrum, g=1,2,3,...,f, f represents the total number of spectrum harmonic components. In this embodiment, the gth harmonic component frequency represents the frequency of the g+1th harmonic, for example, the first harmonic component frequency represents the frequency of the second harmonic, the second harmonic component frequency represents the frequency of the third harmonic, and so on.

[0059] The algorithm of this embodiment combines the pitch, frequency of each harmonic component of the spectrum and speech speed of the digital human and the cloned object, and comprehensively analyzes to obtain the voice similarity evaluation index. In the process of expression, pitch and speech speed often affect each other, and changes in speech speed may also affect people's perception of pitch. When speaking quickly, the change of pitch may be more frequent and subtle, while when speaking slowly, the change of pitch may be more stable and obvious. At the same time, the pitch will affect the brightness and penetration of the sound, and then affect the performance of the frequency of each harmonic component of the spectrum. For example, high-pitched sounds are often brighter and crisper, while low-pitched sounds are deeper and thicker. In addition, changes in speech speed may affect the performance of the frequency of each harmonic component of the spectrum. Comprehensive analysis can obtain a more comprehensive, accurate and in-depth voice similarity evaluation index.

[0060] Table 2 Example of sound similarity evaluation index data

[0061]

[0062] As shown in Table 2, the voice similarity evaluation index is jointly determined by the pitch, the frequency of each harmonic component of the spectrum and the speaking speed. In a specific embodiment, n=1, f=1, the pitch of the cloned object is 180Hz, the average speaking speed of the cloned object is 120 words / minute, the frequency of the second harmonic component of the spectrum of the cloned object is 200Hz, the allowable deviation frequency of the spectrum harmonic component is 50Hz, the allowable deviation pitch is 20Hz, the allowable deviation average speaking speed is 10 words / minute, the sound similarity evaluation impact factor corresponding to the set pitch is 0.3, the sound similarity evaluation impact factor corresponding to the set average speaking speed is 0.3, and the sound similarity evaluation impact factor corresponding to the set spectrum harmonic component frequency is 0.4. This formula takes into account three key factors, namely pitch, spectrum harmonic component frequency and speaking speed. The digital human can present a more natural and fluent voice expression, reduce machine-generated traces, make users feel more comfortable and friendly when listening to advice or receiving diagnosis and treatment, and help enhance the communication effect between users and digital humans. By standardizing the pitch deviation, frequency deviation of each harmonic component of the spectrum, and speech speed deviation of the digital human and the Chinese medicine expert, it is ensured that they are compared at the same level, which improves the fairness and comparability of the evaluation. By weighting the influence of pitch deviation, frequency deviation of each harmonic component of the spectrum, and speech speed deviation, it reflects their relative importance in the evaluation index. The weights of different factors can be adjusted according to different needs, making the model highly adaptable. It is not difficult to see that the smaller the pitch deviation, the smaller the frequency deviation of each harmonic component of the spectrum, or the smaller the average speech speed deviation, the greater the sound similarity evaluation index. By evaluating the sound similarity evaluation index of the digital human, the naturalness and immersion of the user's interaction with the digital human can be improved, the user's trust and satisfaction can be enhanced, and the user can understand the information or service content provided by the digital human more accurately. At the same time, it can ensure that the voice of the digital human is highly consistent with the reference object, thereby improving the recognition accuracy, which helps to reduce recognition errors or confusion caused by sound differences.

[0063] It should be understood that in this embodiment It is a sound similarity assessment influencing factor corresponding to the pitch deviation preset in the TCM diagnosis and treatment database, which represents a numerical value of the degree of influence of the pitch deviation on the sound similarity assessment index. When used, the sound similarity assessment influencing factor corresponding to the pitch deviation can be directly obtained from the TCM diagnosis and treatment database, and the corresponding relationship can be a pre-set mapping relationship. For example, the pitch deviation of each time node and the sound similarity assessment influencing factor corresponding to the pitch deviation preset in the TCM diagnosis and treatment database form a mapping set, and the real-time pitch is input into the mapping set to obtain the sound similarity assessment influencing factor corresponding to the pitch deviation. The mapping relationship can be one-to-one or many-to-one. In this example, its value range is between 0 and 1; It is a sound similarity assessment influencing factor corresponding to the average speech speed deviation preset in the traditional Chinese medicine diagnosis and treatment database, which represents the numerical value of the influence degree of the average speech speed deviation on the sound similarity assessment index. When used, the sound similarity assessment influencing factor corresponding to the average speech speed deviation can be directly obtained from the traditional Chinese medicine diagnosis and treatment database, and the corresponding relationship can be a pre-set mapping relationship. For example, the average speech speed and the sound similarity assessment influencing factor corresponding to the average speech speed deviation preset in the traditional Chinese medicine diagnosis and treatment database form a mapping set, and the real-time average speech speed is input into the mapping set to obtain the sound similarity assessment influencing factor corresponding to the average speech speed deviation. The mapping relationship can be one-to-one or many-to-one. In this example, its value range is between 0 and 1; It is the sound similarity assessment influence factor corresponding to the spectral harmonic component frequency preset in the traditional Chinese medicine diagnosis and treatment database, which represents the numerical value of the influence degree of the spectral harmonic component frequency on the sound similarity assessment index. When used, the sound similarity assessment influence factor corresponding to the spectral harmonic component frequency can be directly obtained from the traditional Chinese medicine diagnosis and treatment database, and the corresponding relationship can be a preset mapping relationship. For example, the spectral harmonic component frequency and the sound similarity assessment influence factor corresponding to the spectral harmonic component frequency preset in the traditional Chinese medicine diagnosis and treatment database form a mapping set, and the real-time spectral harmonic component frequency is input into the mapping set to obtain the sound similarity assessment influence factor corresponding to the spectral harmonic component frequency. The mapping relationship can be one-to-one or many-to-one. In this example, its value range is between 0 and 1.

[0064] It should be explained that the voice similarity evaluation index of the digital human in this embodiment is a quantitative index obtained by analyzing the pitch, the frequency of each harmonic component of the spectrum and the average speaking speed, and is used to quantify the voice similarity between the digital human and the cloned object.

[0065] In a specific embodiment, a similarity evaluation value of the digital person's clone object is obtained through comprehensive analysis. The specific analysis process is: based on the digital person's appearance similarity evaluation index and voice similarity evaluation index, a similarity evaluation value of the digital person's clone object is obtained through comprehensive analysis.

[0066] Specifically, the similarity evaluation value of the digital human clone object is expressed as:

[0067] in, Represents the similarity evaluation value of the digital human clone object, Represents the appearance similarity evaluation index of digital humans, represents the digital human voice similarity evaluation index, Indicates the influence factor of the clone object similarity evaluation value corresponding to the set appearance similarity evaluation index, Indicates the influencing factor of the cloned object similarity evaluation value corresponding to the set sound similarity evaluation index.

[0068] The algorithm of this embodiment combines the appearance similarity evaluation index and the voice similarity evaluation index of the digital human, and comprehensively analyzes to obtain the similarity evaluation value of the cloned object. The appearance similarity evaluation index and the voice similarity evaluation index are both important indicators for measuring the realism of the digital human. When the digital human is highly similar to the cloned object in both appearance and voice, its overall realism will be greatly enhanced. Although appearance and voice are two independent dimensions, they can complement each other. For example, if a digital human with a highly restored appearance does not have a realistic voice, its overall realism may be reduced, and vice versa. Together, they constitute important measurement criteria for the realism and fidelity of the digital human, and influence each other in terms of technology dependence, application scenarios, etc. Comprehensive analysis can obtain a more comprehensive, accurate, and in-depth similarity evaluation value of the cloned object.

[0069] like Figure 3 As shown, in a specific embodiment, =0.5, =0.3. =0.1, the functional relationship between the similarity evaluation value of the digital human clone and the appearance similarity evaluation index is shown in curve a; when =0.5, the functional relationship between the similarity evaluation value of the digital human clone and the appearance similarity evaluation index is shown in curve b; when =1, the functional relationship between the similarity evaluation value of the digital human’s cloned object and the appearance similarity evaluation index is shown in curve c.

[0070] It should be explained that in this embodiment, two key factors, namely, the appearance similarity evaluation index and the voice similarity evaluation index of the digital person, are considered, and a comprehensive and integrated similarity evaluation can be performed on the cloned object to ensure the comprehensiveness and accuracy of the evaluation results. The weights of the appearance similarity evaluation index and the voice similarity evaluation index can be flexibly adjusted according to different application requirements and user preferences to achieve personalized customization and differentiated evaluation, which helps to find out where the cloned object is insufficient or needs to be improved, and provides strong support for subsequent optimization and improvement. By weighting the influence of the appearance similarity evaluation index and the voice similarity evaluation index of the digital person, it reflects their relative importance in the evaluation index, and the weights of different factors can be adjusted according to different requirements, so that the model has good adaptability. It is not difficult to see that the larger the appearance similarity evaluation index of the digital person or the larger the voice similarity evaluation index, the larger the similarity evaluation value of the cloned object of the digital person. By evaluating the similarity assessment value of the digital human's clone object, we can more accurately measure the extent to which the digital human can simulate the appearance and voice of a real expert, thereby ensuring its effectiveness and reliability in various application scenarios, helping to discover and improve deficiencies in the technology, and thus promoting the continuous optimization and development of digital human technology. This can not only improve the performance of digital humans in specific applications, but also further promote the development of related technologies so that they can better serve social needs.

[0071] It should be understood that in this embodiment It is the clone object similarity evaluation value influencing factor corresponding to the appearance similarity evaluation index preset in the traditional Chinese medicine diagnosis and treatment database, which represents the numerical value of the influence degree of the appearance similarity evaluation index on the clone object similarity evaluation value. When used, the clone object similarity evaluation value influencing factor corresponding to the appearance similarity evaluation index can be directly obtained from the traditional Chinese medicine diagnosis and treatment database, and the corresponding relationship can be a pre-set mapping relationship. For example, the appearance similarity evaluation index of the digital human and the clone object similarity evaluation value influencing factor corresponding to the appearance similarity evaluation index preset in the traditional Chinese medicine diagnosis and treatment database form a mapping set, and the real-time appearance similarity evaluation index is input into the mapping set to obtain the clone object similarity evaluation value influencing factor corresponding to the appearance similarity evaluation index. The mapping relationship can be one-to-one or many-to-one. In this example, its value range is between 0 and 1; It is the clone object similarity evaluation value influencing factor corresponding to the sound similarity evaluation index preset in the traditional Chinese medicine diagnosis and treatment database, which represents the numerical value of the influence degree of the sound similarity evaluation index on the clone object similarity evaluation value. When used, the clone object similarity evaluation value influencing factor corresponding to the sound similarity evaluation index can be directly obtained from the traditional Chinese medicine diagnosis and treatment database, and the corresponding relationship can be a pre-set mapping relationship. For example, the digital human's sound similarity evaluation index and the clone object similarity evaluation value influencing factor corresponding to the sound similarity evaluation index preset in the traditional Chinese medicine diagnosis and treatment database form a mapping set, and the real-time sound similarity evaluation index is input into the mapping set to obtain the clone object similarity evaluation value influencing factor corresponding to the sound similarity evaluation index. The mapping relationship can be one-to-one or many-to-one. In this example, its value range is between 0 and 1.

[0072] It should be explained that the similarity evaluation value of the cloned object of the digital person in this embodiment is a quantitative index obtained by analyzing the appearance similarity evaluation index and the voice similarity evaluation index, and is used to quantify the similarity between the digital person and the cloned object.

[0073] In a specific embodiment, the digital human is evaluated according to the cloned object similarity evaluation value of the digital human. The specific evaluation process is: extracting the cloned object similarity evaluation threshold from the traditional Chinese medicine diagnosis and treatment database, comparing the cloned object similarity evaluation value of the digital human with the cloned object similarity evaluation threshold, if the cloned object similarity evaluation value of the digital human is greater than or equal to the cloned object similarity evaluation threshold, the digital human is evaluated as qualified and marked as a high-similarity digital human, and then the behavior pattern accuracy of the high-similarity digital human is evaluated, if the cloned object similarity evaluation value of the digital human is less than the cloned object similarity evaluation threshold, the digital human is evaluated as unqualified, and feedback correction is performed, and the corrected digital human is evaluated for similarity.

[0074] It should be explained that in this embodiment, the behavior pattern data is processed to obtain the behavior pattern accuracy evaluation index of the high-similarity digital person. The specific analysis process is: the behavior pattern data includes the consistency evaluation value of the diagnosis and treatment recommendation and the matching degree of the medication recommendation.

[0075] The critical diagnosis and treatment recommendation consistency assessment value and critical medication recommendation matching degree were extracted from the traditional Chinese medicine diagnosis and treatment database, and the behavior pattern accuracy assessment index was obtained through comprehensive analysis.

[0076] In a specific embodiment, it is necessary to select TCM cases to ensure that these cases cover the common disease types and symptom manifestations of TCM. The number of cases should be large enough to reflect the diversity and complexity of TCM diagnosis and treatment. Reference standard diagnosis and treatment recommendations are extracted from the TCM diagnosis and treatment database, and the high-similarity digital human system is allowed to automatically diagnose and generate diagnosis and treatment recommendations for the same cases. Statistical tools are used to obtain the number of identical characters between the high-similarity digital human diagnosis and treatment recommendation text and the reference standard diagnosis and treatment recommendation text, and the ratio of the number of characters in the high-similarity digital human diagnosis and treatment recommendation text is calculated and marked as the diagnosis and treatment recommendation consistency evaluation value; at the same time, reference standard drug recommendations are extracted from the TCM diagnosis and treatment database, and the high-similarity digital human system is allowed to automatically diagnose and generate drug recommendations for the same cases. Statistical tools are used to obtain cases with consistent drug recommendations between the high-similarity digital human and the reference standard, and the ratio of the total cases is calculated. This ratio is the medication recommendation matching degree.

[0077] It should be explained that the critical diagnosis and treatment recommendation consistency evaluation value in this embodiment refers to the threshold value that will affect the performance of the high-similarity digital human when the diagnosis and treatment recommendations of the high-similarity digital human and the cloned object deviate to a certain extent, and the critical medication recommendation matching degree refers to the threshold value that will affect the performance of the high-similarity digital human when the drug recommendations of the high-similarity digital human and the cloned object deviate to a certain extent.

[0078] Specifically, the accuracy evaluation index of the behavior pattern of a high-similarity digital human is expressed as follows:

[0079] in, represents the accuracy evaluation index of the behavior pattern of a high-similarity digital person, e represents a natural constant, represents the consistency evaluation value of the diagnosis and treatment recommendations of the cth TCM case with high similarity digital person, represents the critical diagnosis and treatment recommendation consistency assessment value, Indicates the matching degree of medication recommendation for highly similar digital people. Indicates the critical medication recommendation matching degree, It represents the behavioral pattern accuracy assessment impact factor corresponding to the set diagnosis and treatment recommendation consistency assessment value. It represents the impact factor of the behavioral pattern accuracy assessment corresponding to the set medication recommendation matching degree, c represents the number of each TCM case, c=1,2,3,...,p, and p represents the total number of TCM cases.

[0080] The algorithm of this embodiment combines the consistency evaluation value of the diagnosis and treatment suggestions of the high-similarity digital human and the matching degree of the medication recommendation, and comprehensively analyzes to obtain the behavior pattern accuracy evaluation index. A high consistency evaluation value of the diagnosis and treatment suggestions often means that the digital human and the Chinese medicine expert have a relatively unified cognition in diagnosing the disease and analyzing the cause and pathogenesis. This consistency of cognition provides a solid foundation for subsequent medication recommendations. When the digital human and the Chinese medicine expert reach a consensus on the diagnosis and treatment suggestions, they are also more likely to reach a consensus in terms of selecting drugs, determining dosages and routes of administration, thereby improving the matching degree of medication recommendations. Comprehensive analysis can obtain a more comprehensive, accurate and in-depth behavior pattern accuracy evaluation index.

[0081] Table 3 Example of behavior pattern accuracy evaluation index data

[0082]

[0083] As shown in Table 3, the behavior pattern accuracy assessment index is jointly determined by the diagnosis and treatment recommendation consistency assessment value and the medication recommendation matching degree of the high-similarity digital person. In a specific embodiment, p=1, the critical diagnosis and treatment recommendation consistency assessment value is 60%, the critical medication recommendation matching degree is 70%, the behavior pattern accuracy assessment impact factor corresponding to the set diagnosis and treatment recommendation consistency assessment value is 0.4, and the behavior pattern accuracy assessment impact factor corresponding to the set medication recommendation matching degree is 0.3.

[0084] It should be explained that this embodiment takes into account two key factors, namely the consistency evaluation value of the diagnosis and treatment suggestions of high-similarity digital people and the matching degree of medication recommendations, which helps to ensure that users receive scientific and reasonable treatment plans, thereby improving medical quality and safety. It can ensure that the diagnosis and treatment suggestions of digital people are highly consistent with those of real doctors, thereby improving the accuracy of diagnosis and treatment. It can also ensure that the recommended medication plan matches the patient's actual condition, thereby improving the effectiveness and safety of medication, further enhancing the patient's trust in the medication plan, and thus improving patient satisfaction. By standardizing the consistency evaluation value of the diagnosis and treatment suggestions of high-similarity digital people and the matching degree of medication recommendations, ensuring that they are compared at the same level, the fairness and comparability of the evaluation are improved, and at the same time and The setting of helps to determine whether there is a significant deviation in the performance of the digital human. By weighting the impact of the consistency evaluation value of the diagnosis and treatment recommendations and the matching degree of the medication recommendations of the high-similarity digital human, it reflects their relative importance in the evaluation index. The weights of different factors can be adjusted according to different needs, making the model highly adaptable. It is not difficult to see that the greater the consistency evaluation value of the diagnosis and treatment recommendations of the high-similarity digital human or the greater the matching degree of the medication recommendations, the greater the accuracy evaluation index of the behavior pattern of the high-similarity digital human. By evaluating the accuracy evaluation index of the behavior pattern of the high-similarity digital human, it can reveal the possible errors and deviations of the high-similarity digital human in the behavior simulation process, and prompt developers to make targeted optimizations and improvements, thereby improving the overall reliability of the high-similarity digital human application. The accurate behavior pattern makes the digital human more natural and smooth when interacting with the user, reducing user confusion and dissatisfaction caused by incoherent or incorrect behavior, which helps to improve the professionalism of the high-similarity digital human and the user experience.

[0085] like Figure 4 As shown, in a specific embodiment, when M=0.5, the functional relationship between the behavioral pattern accuracy evaluation index of a high-similarity digital human and the consistency evaluation value of the diagnosis and treatment recommendations is shown as curve a; when M=0.8, the functional relationship between the similarity evaluation value of the digital human's cloned object and the appearance similarity evaluation index is shown as curve b; when M=1, the functional relationship between the similarity evaluation value of the digital human's cloned object and the appearance similarity evaluation index is shown as curve c.

[0086] It should be understood that in this embodiment It is the behavior pattern accuracy assessment impact factor corresponding to the diagnosis and treatment recommendation consistency assessment value preset in the traditional Chinese medicine diagnosis and treatment database, which represents the numerical value of the degree of influence of the diagnosis and treatment recommendation consistency assessment value on the behavior pattern accuracy assessment index. When used, the behavior pattern accuracy assessment impact factor corresponding to the diagnosis and treatment recommendation consistency assessment value can be directly obtained from the traditional Chinese medicine diagnosis and treatment database, and the corresponding relationship can be a pre-set mapping relationship. For example, the diagnosis and treatment recommendation consistency assessment value of the high-similarity digital person and the behavior pattern accuracy assessment impact factor corresponding to the diagnosis and treatment recommendation consistency assessment value preset in the traditional Chinese medicine diagnosis and treatment database form a mapping set, and the real-time diagnosis and treatment recommendation consistency assessment value is input into the mapping set to obtain the behavior pattern accuracy assessment impact factor corresponding to the diagnosis and treatment recommendation consistency assessment value. The mapping relationship can be one-to-one or many-to-one. In this example, its value range is between 0 and 1; It is the behavior pattern accuracy assessment influencing factor corresponding to the medication recommendation matching degree preset in the traditional Chinese medicine diagnosis and treatment database, which represents the numerical value of the influence degree of the medication recommendation matching degree on the behavior pattern accuracy assessment index. When used, the behavior pattern accuracy assessment influencing factor corresponding to the medication recommendation matching degree can be directly obtained from the traditional Chinese medicine diagnosis and treatment database, and the corresponding relationship can be a pre-set mapping relationship. For example, the medication recommendation matching degree of a high-similarity digital person and the behavior pattern accuracy assessment influencing factor corresponding to the medication recommendation matching degree preset in the traditional Chinese medicine diagnosis and treatment database form a mapping set, and the real-time medication recommendation matching degree is input into the mapping set to obtain the behavior pattern accuracy assessment influencing factor corresponding to the medication recommendation matching degree. The mapping relationship can be one-to-one or many-to-one. In this example, its value range is between 0 and 1.

[0087] It should be explained that the behavioral pattern accuracy assessment index of the high-similarity digital human in this embodiment is a quantitative indicator obtained by analyzing the consistency assessment value of the diagnosis and treatment recommendations and the matching degree of the medication recommendations, and is used to quantify the accuracy of the behavioral patterns of the digital human and the cloned object.

[0088] Furthermore, by evaluating the accuracy of the behavioral patterns of high-similarity digital humans and cloned objects, it is possible to verify whether their diagnostic capabilities have reached or are close to the level of experts, which helps to quickly respond to patients' needs, provide preliminary diagnostic suggestions and treatment plans, and ensure the effectiveness and reliability of these suggestions and treatment plans, thereby improving the efficiency of overall medical services.

[0089] Specifically, feedback correction is performed, and the specific feedback process is: extracting the appearance similarity evaluation index threshold and the voice similarity evaluation index threshold of the digital person and the behavior pattern accuracy evaluation index threshold of the high-similarity digital person from the traditional Chinese medicine diagnosis and treatment database; if the appearance similarity evaluation index of the digital person is less than the appearance similarity evaluation index threshold, then adjusting the digital person's facial aspect ratio, the deviation pixel value of the key position points of the facial features, and the deviation distance of the key position points of the facial features to make it qualified; if the voice similarity evaluation index of the digital person is less than the voice similarity evaluation index threshold, then adjusting the digital person's pitch, the frequency of each harmonic component of the spectrum, and the speaking speed to make it qualified; if the behavior pattern accuracy evaluation index of the high-similarity digital person is less than the behavior pattern accuracy evaluation index threshold, then adjusting the consistency evaluation value of the diagnosis and treatment recommendations of the high-similarity digital person, the matching degree of the medication recommendation, and the accuracy of the condition understanding of each condition description text to make it qualified.

[0090] Reference Figure 2 As shown, the second aspect of the present invention provides a method for analyzing TCM diagnosis and treatment data based on digital human cloning technology, including: collecting characterization data of the cloned object and the digital human, wherein the characterization data includes appearance feature data and sound feature data.

[0091] The appearance similarity evaluation index of the digital person is obtained by processing the appearance feature data, the voice similarity evaluation index of the digital person is obtained by processing the voice feature data, and the similarity evaluation value of the digital person's clone object is obtained by comprehensive analysis. The digital person is evaluated for similarity based on the similarity evaluation value of the digital person's clone object to obtain a high-similarity digital person.

[0092] Obtain behavioral pattern data of high-similarity digital humans, process the behavioral pattern data to obtain a behavioral pattern accuracy assessment index of high-similarity digital humans, conduct accuracy assessment on the high-similarity digital humans based on the behavioral pattern accuracy assessment index of high-similarity digital humans and obtain the assessment results. If the assessment result is qualified, enter the diagnosis and treatment stage; if the assessment result is unqualified, provide feedback and correct the behavioral pattern of the high-similarity digital humans and assess them again.

[0093] The traditional Chinese medicine diagnosis and treatment database is used to store data related to traditional Chinese medicine diagnosis and treatment, including the allowable deviation tone, the allowable deviation speech speed, the appearance similarity assessment influencing factor corresponding to the set facial aspect ratio, the appearance similarity assessment influencing factor corresponding to the set deviation pixel value of the key position points of the facial features, the appearance similarity assessment influencing factor corresponding to the set deviation distance of the key position points of the facial features, the critical diagnosis and treatment recommendation consistency assessment value, the critical condition understanding accuracy, the clone object similarity assessment value influencing factor corresponding to the set appearance similarity assessment index, the clone object similarity assessment value influencing factor corresponding to the set voice similarity assessment index and the clone object similarity assessment threshold. The data in the traditional Chinese medicine diagnosis and treatment database are obtained by collecting data from multiple traditional Chinese medicine diagnosis and treatment practices.

[0094] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the present invention, they should all fall within the protection scope of the present invention.

Claims

1. A TCM diagnosis and treatment data analysis system based on digital human cloning technology, characterized in that: include: A TCM diagnosis and treatment data collection module, used to collect characterization data of cloned objects and digital humans, the characterization data including appearance feature data and voice feature data; The cloning similarity analysis module is used to obtain the appearance similarity evaluation index of the digital person according to the appearance feature data, to obtain the voice similarity evaluation index of the digital person according to the voice feature data, to obtain the similarity evaluation value of the digital person's clone object through comprehensive analysis, and to perform similarity evaluation on the digital person according to the similarity evaluation value of the digital person's clone object to obtain a digital person with high similarity; The behavior pattern accuracy assessment module is used to obtain the behavior pattern data of the high-similarity digital human, process the behavior pattern data to obtain the behavior pattern accuracy assessment index of the high-similarity digital human, conduct accuracy assessment on the high-similarity digital human according to the behavior pattern accuracy assessment index of the high-similarity digital human and obtain the assessment result. If the assessment result is qualified, the diagnosis and treatment stage is entered; if the assessment result is unqualified, feedback correction is made to the behavior pattern of the high-similarity digital human and the assessment is conducted again; The appearance feature data includes facial images and facial aspect ratios of the cloned object and the digital human; Deploy several key points of facial features, overlap the facial image of the digital person with the facial image of the cloned object, extract the distance between the key points of facial features on the facial image of the cloned object and the digital person, and mark them as the deviation distance of the key points of facial features on the digital person; Extracting pixel values ​​of key positions of facial features of the cloned object and the digital human's facial images and performing a difference operation, marking the absolute value of the difference result as the deviation pixel value of key positions of facial features of the digital human; The allowable deviation facial aspect ratio, critical facial features key position point deviation distance and critical facial features key position point deviation pixel value are extracted from the traditional Chinese medicine diagnosis and treatment database, and the digital human appearance similarity evaluation index is obtained through comprehensive analysis. The digital human appearance similarity evaluation index, the specific numerical expression is: in, represents the appearance similarity evaluation index of digital people, e represents a natural constant, represents the facial aspect ratio of the digital human, Indicates the facial aspect ratio of the cloned object. represents the allowable deviation of the facial aspect ratio, Indicates the pixel value of the deviation of the key position of the i-th facial features of the digital person, Indicates the deviation pixel value of the critical facial features key position point, represents the deviation distance of the key position point of the i-th facial features of the digital human, Indicates the deviation distance of the set critical facial features key position points, It represents the influencing factor of appearance similarity evaluation corresponding to the set facial aspect ratio, It represents the influencing factor of appearance similarity assessment corresponding to the deviation pixel value of the key position points of the facial features. It represents the influencing factor of appearance similarity assessment corresponding to the set deviation distance of the key position points of facial features, i represents the number of each key position point of facial features, i=1,2,3,...,m, m represents the total number of key position points of facial features.

2. According to claim 1, the TCM diagnosis and treatment data analysis system based on digital human cloning technology is characterized by: The sound feature data is processed to obtain the digital human voice similarity evaluation index, and the specific analysis process is as follows: Acquire test audio of the cloned object and the digital human, and deploy several time monitoring points, wherein the sound feature data includes average speech speed, frequency of each harmonic component of the spectrum, and pitch of each time monitoring point; The allowable deviation pitch, allowable deviation spectral harmonic component frequency and allowable deviation average speaking speed are extracted from the traditional Chinese medicine diagnosis and treatment database, and the voice similarity evaluation index of the digital human is obtained through comprehensive analysis.

3. According to claim 2, the TCM diagnosis and treatment data analysis system based on digital human cloning technology is characterized by: The comprehensive analysis obtains the similarity evaluation value of the digital human clone object, and the specific analysis process is as follows: Based on the appearance similarity evaluation index and voice similarity evaluation index of the digital person, a comprehensive analysis is performed to obtain the similarity evaluation value of the digital person's clone object.

4. According to claim 3, the TCM diagnosis and treatment data analysis system based on digital human cloning technology is characterized by: The similarity evaluation of the digital human is performed according to the similarity evaluation value of the digital human's clone object. The specific evaluation process is as follows: A clone object similarity assessment threshold is extracted from the traditional Chinese medicine diagnosis and treatment database, and the clone object similarity assessment value of the digital human is compared with the clone object similarity assessment threshold. If the clone object similarity assessment value of the digital human is greater than or equal to the clone object similarity assessment threshold, the digital human is assessed as qualified and marked as a high-similarity digital human. If the clone object similarity assessment value of the digital human is less than the clone object similarity assessment threshold, the digital human is assessed as unqualified, and feedback correction is performed, and the corrected digital human is evaluated for similarity.

5. According to claim 1, the TCM diagnosis and treatment data analysis system based on digital human cloning technology is characterized by: The behavior pattern data is processed to obtain the behavior pattern accuracy evaluation index of the high-similarity digital human. The specific analysis process is as follows: The behavior pattern data includes the consistency evaluation value of the diagnosis and treatment recommendations and the matching degree of the medication recommendations of the highly similar digital person; The critical consistency evaluation value of diagnosis and treatment recommendations and the critical matching degree of medication recommendations were extracted from the traditional Chinese medicine diagnosis and treatment database, and a comprehensive analysis was performed to obtain the behavioral pattern accuracy evaluation index of a high-similarity digital human.

6. According to claim 5, the TCM diagnosis and treatment data analysis system based on digital human cloning technology is characterized by: The behavioral pattern accuracy assessment index of the high-similarity digital human is a quantitative indicator obtained by analyzing the consistency assessment value of the diagnosis and treatment recommendations and the matching degree of the medication recommendations, and is used to quantify the accuracy of the behavioral patterns of the digital human and the cloned object.

7. According to claim 2, the TCM diagnosis and treatment data analysis system based on digital human cloning technology is characterized by: The digital human voice similarity evaluation index is a quantitative index obtained by analyzing the pitch, the frequency of each harmonic component of the spectrum and the average speaking speed, and is used to quantify the voice similarity between the digital human and the cloned object.

8. A method for applying the TCM diagnosis and treatment data analysis system based on digital human cloning technology as described in any one of claims 1 to 7, characterized in that: include: Collecting characterization data of the cloned object and the digital human, wherein the characterization data includes appearance feature data and voice feature data; The digital person's appearance similarity evaluation index is obtained by processing the appearance feature data, the digital person's voice similarity evaluation index is obtained by processing the voice feature data, and the digital person's cloned object similarity evaluation value is obtained by comprehensive analysis. The digital person is evaluated for similarity based on the cloned object similarity evaluation value to obtain a high-similarity digital person; Obtain behavioral pattern data of high-similarity digital humans, process the behavioral pattern data to obtain a behavioral pattern accuracy assessment index of high-similarity digital humans, conduct accuracy assessment on the high-similarity digital humans based on the behavioral pattern accuracy assessment index of high-similarity digital humans and obtain the assessment results. If the assessment result is qualified, enter the diagnosis and treatment stage; if the assessment result is unqualified, provide feedback and correct the behavioral pattern of the high-similarity digital humans and assess them again.

Images